Cross Reference: /src/usr.bin/indent

History log of /src/usr.bin/indent
Revision	Date	Author	Comments
1.7	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.6	16-May-2023	rillig	indent: allow comments in column 1 to be formatted
1.5	15-May-2023	rillig	indent: clean up local indentation profile The -eei option now works, the type hints are no longer necessary.
1.4	15-May-2023	rillig	indent: let indent format its own code With manual corrections, as indent does not properly indent multi-line '?:' expressions nor multi-line controlling expressions.
1.3	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.2	05-Oct-2021	rillig	indent: run indent on lexi.c, with manual corrections The variables 'keywords' and 'typenames' were indented using 8 spaces, even though -di0 was in effect, which should result in a single space, and -ut was in effect, which should result in a single tab instead of 8 spaces. The option -eei does not work as advertised, the controlling expressions are only indented by the normal amount, which easily leads to confusion as to whether the code belongs to the condition or the following statement.
1.1	26-Sep-2021	rillig	indent: add .indent.pro that almost matches the source code One might expect that the code of indent is properly indented according to its own capabilities, but that's not the case, there are many deviations. This indentation profile comes close to the existing code. Maybe someday indent's own source code can be formatted using this profile, but before attempting that, its remaining bugs have to be fixed. Development of indent has essentially stopped somewhere around 1990, as demonstrated by the wrong formatting of '...' that has only been fixed a few minutes ago. The '...' is an invention of C90. Indent's parser still considers '...' as consisting of the 3 tokens period-period-period, but that's OK since the effect is the same. Another feature that had been missing for a long time were C99 comments that span from '//' to the next newline. Before March 2021, these were parsed as a binary operator, which produced lots of funny side effects. Since indent's code makes use of several C99 features, as soon as it can properly indent its own code, the worst of these bugs will have been fixed.
1.15	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.14	07-Oct-2021	rillig	indent: raise WARNS from the default 5 up to 6
1.13	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.12	26-Mar-2021	rillig	indent: remove workaround for array initialization bug in lint The bug has been fixed in init.c 1.133 from 2021-03-25.
1.11	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.10	12-Mar-2021	rillig	indent: add helper functions for doing the actual output This allows to add debug logging to these few functions instead of all other places that might output something. Reducing the possible output formats to a few primitives makes dump_line simpler, especially the fprintf calls. It also removes the non-constant printf string. The call to output_int may be meant for debugging, as the character 0x80 is unlikely to appear in any real-world code. No functional change.
1.9	08-Mar-2021	rillig	indent: make it easy to compile indent in debug mode
1.8	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.7	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.6	08-Oct-2006	peter	branches: 1.6.82; WFORMAT is no more...
1.5	11-Oct-2000	is	More format string cleanup by sommerfeld.
1.4	18-Oct-1997	mrg	branches: 1.4.4; 1.4.12; merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	31-Jul-1993	mycroft	Add RCS indentifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.4.12.1	18-Oct-2000	tv	Pullup usr.bin string format fixes [is]. See "cvs log" for explicit revision numbers per file, from sommerfeld.
1.4.4.1	19-Oct-2000	he	Pull up revision 1.5 (requested by he): Format string cleanup.
1.6.82.1	10-Jun-2019	christos	Sync with HEAD
1.4	18-Nov-2021	rillig	indent: replace old license discussion with a brief history section
1.3	26-Mar-2021	rillig	indent: don't claim that indent is "the nicest C pretty printer around" That statement may have been true in 1993, but definitely is not true anymore, as of 2021. The part about "needs to be completely redone" is still true though since indent cannot even format its own source code in an acceptable way.
1.2	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.1	09-Apr-1993	cgd	branches: 1.1.1; 1.1.120; added, from net/2 (patch 124).
1.1.120.1	10-Jun-2019	christos	Sync with HEAD
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.3	04-Jun-2023	rillig	indent: remove trailing whitespace from README
1.2	27-Nov-2021	rillig	indent: reword README
1.1	18-Nov-2021	rillig	indent: replace old license discussion with a brief history section
1.88	12-Dec-2024	rillig	indent: add error handling for I/O errors Suggested by lint2.
1.87	10-Dec-2023	rillig	branches: 1.87.2; indent: be strict about options from profile files Previously, the "option" 'xdi0' was treated the same as '-xdi0'.
1.86	03-Dec-2023	rillig	indent: inline input-related macros No binary change.
1.85	15-Jun-2023	rillig	indent: miscellaneous cleanups, more tests for edge cases
1.84	14-Jun-2023	rillig	indent: reduce number of relocations Since all command line options modify a member of struct options, there is no need to encode that relocation 38 times. No functional change.
1.83	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.82	05-Jun-2023	rillig	indent: do not report broken lines, report configuration on stderr
1.81	05-Jun-2023	rillig	indent: rename variables, clean up comments No binary change.
1.80	18-May-2023	rillig	indent: rename a few functions No functional change.
1.79	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.78	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.77	14-May-2023	rillig	indent: remove foreign RCS IDs
1.76	14-May-2023	rillig	indent: miscellaneous cleanups
1.75	13-May-2023	rillig	indent: improve names of option variables No functional change.
1.74	13-May-2023	rillig	indent: don't try to read from the file '(null)/.indent.pro'
1.73	20-Jan-2023	rillig	indent: fix misleading comment
1.72	25-Nov-2021	rillig	indent: make error message for missing command line arguments clearer
1.71	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.70	07-Nov-2021	rillig	indent: parse special options strictly
1.69	05-Nov-2021	rillig	indent: clean up argument parsing In struct pro, place the dependent member below its dependency. In load_profile, consistently use -1 when outside a comment. No functional change.
1.68	31-Oct-2021	rillig	indent: clean up Initialize buffers in reading order, make comments more expressive, rename add_typename to register_typename, remove unused macro. No functional change.
1.67	29-Oct-2021	rillig	indent: parse options in a platform-independent way Previously, on an ILP32 platform, the option '-ts30000000000000000' resulted in the error message 'must be an integer', on LP64 platforms it resulted in the error message 'must be between 1 and 80'. Remove this unnecessary difference.
1.66	28-Oct-2021	rillig	indent: clean up indentation, comments, reduce No functional change.
1.65	28-Oct-2021	rillig	indent: fix error message for buffer overflow during option parsing At this early time, the input file has not been opened yet, so there is no reason to output either the input file name or the line number.
1.64	28-Oct-2021	rillig	indent: make error messages for option parsing more precise
1.63	28-Oct-2021	rillig	indent: parse option '-cli' strictly
1.62	28-Oct-2021	rillig	indent: topologically sort functions No functional change.
1.61	28-Oct-2021	rillig	indent: change product name, update version number NetBSD's indent has deviated enough from FreeBSD's indent to warrant a different product name. When indent was copied from FreeBSD in 2019, that update introduced several new bugs, some of which have been fixed in the NetBSD version. NetBSD indent, unlike FreeBSD indent, supports C99 comments and C99 initializer designators.
1.60	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.59	24-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards. Indent still does not get extra_expr_indent correctly, it also indents global variables after tagged declarations too deep. No functional change.
1.58	24-Oct-2021	rillig	indent: rename nitems to array_length
1.57	24-Oct-2021	rillig	indent: sort includes
1.56	17-Oct-2021	rillig	indent: parse int command line options strictly On i386 and other platforms where LONG_MAX == INT_MAX, the test t_errors/option_tabsize_very_large failed since the behavior on integer overflow differs between ILP32 and LP64 platforms. Noticed by gson@. Avoid this unintended difference by adding reasonable limits for each of the integer options and by replacing atoi with strtol.
1.55	13-Oct-2021	rillig	indent: check command line options stricter Previously, bool options were allowed to have trailing garbage. For example, the option '-bacc' could be spelled '-bacchus' as well. Check that the exact option name is given in the command line, to prevent typos in the configuration files and to reduce surprises just in case a future option is a prefix of an existing option, or vice versa. Add a new test program for error handling. Most of these tests are so simple that it would be overkill to create 3 files for each test.
1.54	08-Oct-2021	rillig	indent: clean up argument handling Sort the macros, remove redundancy from comment. Remove redundant lint comment. Lint still does not recognize __attribute__((__noreturn__)), but it also doesn't perform advanced control flow analysis, so there is no point in having the comment, as it doesn't suppress any warnings. No functional change.
1.53	08-Oct-2021	rillig	indent: unexport add_typedefs_from_file No functional change.
1.52	08-Oct-2021	rillig	indent: run indent on indent.h The formatting looks mostly OK. Some struct members had excessively long names, leaving no space for their corresponding comments. Renamed some of them using well-known abbreviations. The formatting for debug_vis_range is messed up, no idea why. It is clearly a function declaration, not a function definition, so there is no need to place the function name in column 1. No functional change.
1.51	07-Oct-2021	rillig	indent: rename opt.btype_2 to brace_same_line No functional change.
1.50	07-Oct-2021	rillig	indent: fix wrong or outdated comments No functional change.
1.49	07-Oct-2021	rillig	indent: remove global variable option_source It is only needed at startup, while parsing the options. The string "?" was not needed at all. No functional change.
1.48	07-Oct-2021	rillig	indent: raise WARNS from the default 5 up to 6
1.47	07-Oct-2021	rillig	tests/indent: test parsing of command line options in profile file
1.46	07-Oct-2021	rillig	indent: complain if the profile from the command line does not exist
1.45	07-Oct-2021	rillig	indent: allow long comments in profile files When reading a comment in a profile file, don't store the characters of the comment in the buffer, just skip them. This allows for long comments without triggering overflow errors.
1.44	07-Oct-2021	rillig	indent: prevent buffer overflow when reading profile
1.43	03-Oct-2021	rillig	indent: clean up load_profile No functional change.
1.42	03-Oct-2021	rillig	indent: reduce duplicate code in load_profiles No functional change.
1.41	03-Oct-2021	rillig	indent: rename functions There was no good reason for using the different verbs 'scan' and 'set' for two functions that essentially do the same. No functional change.
1.40	03-Oct-2021	rillig	indent: fix content of profile_name Previously, profile_name included the leading "-P", which was confusing.
1.39	26-Sep-2021	rillig	indent: unexport keyword table, clean up No functional change.
1.38	26-Sep-2021	rillig	indent: force all option variables to be in struct options No functional change.
1.37	26-Sep-2021	rillig	indent: reduce memory usage of the options table Almost all boolean options are negatable, so model this directly instead of saving each option twice. This saves memory, is faster and more directly models reality. No functional change.
1.36	26-Sep-2021	rillig	indent: list options in the same order as in the manual page No functional change.
1.35	26-Sep-2021	rillig	indent: reduce code for listing the options After this change, the few options that do not follow the standard scheme become more visible. They are '-bl', '-br' and '-ta'. No functional change.
1.34	26-Sep-2021	rillig	indent: negate and rename option.leave_comma The old name did not mirror the description in the manual page, and it was the only option that is negated. Inverting it allows the options table to be compressed.
1.33	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.32	26-Sep-2021	rillig	indent: handle special options separately Handling the special options separately removes the need for several macro definitions. It saves a bit of memory since without the option '--version', the option names are shorter. No functional change.
1.31	25-Sep-2021	rillig	indent: reduce abstraction layer for defining boolean options When initializing a boolean option, the most natural values are true and false. Replace the previous values ON and OFF with them. No functional change.
1.30	25-Sep-2021	rillig	indent: clean up argument handling No functional change.
1.29	25-Sep-2021	rillig	indent: clean up argument handling No functional change.
1.28	25-Sep-2021	rillig	indent: reduce binary size No functional change.
1.27	25-Sep-2021	rillig	indent: rename option variable to be more expressive No functional change.
1.26	25-Sep-2021	rillig	indent: convert options from ibool to bool No functional change intended.
1.25	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.24	25-Sep-2021	rillig	indent: clean up initialization of options The default values in 'struct pro' were redundant but all consistent, even with the commented defaults in main_parse_command_line. No functional change.
1.23	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.22	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.21	13-Mar-2021	rillig	indent: distinguish between 'column' and 'indentation' column == 1 + indentation. In addition, indentation is a relative distance while column is an absolute position. Therefore, don't confuse these two concepts, to prevent off-by-one errors. No functional change.
1.20	13-Mar-2021	rillig	indent: fix confusing variable names The word 'col' should only be used for the 1-based column number. This name is completely inappropriate for a line length since that provokes off-by-one errors. The name 'cols' would be acceptable although confusing since it sounds so similar to 'col'. Therefore, rename variables that are related to the maximum line length to 'line_length' since that makes for obvious code and nicely relates to the description of the option in the manual page. No functional change.
1.19	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.18	09-Mar-2021	rillig	indent: manually indent comments It's strange that indent's own code is not formatted by indent itself, which would be a good demonstration of its capabilities. In its current state, I don't trust indent to get even the tokenization correct, therefore the only safe way is to format the code manually.
1.17	07-Mar-2021	rillig	indent: sprinkle a few const No functional change.
1.16	07-Mar-2021	rillig	indent: remove redundant parentheses around return value No functional change.
1.15	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.14	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.13	22-Feb-2016	ginsbach	branches: 1.13.16; Stray '\n' in errx(3) format.
1.12	22-Feb-2016	ginsbach	Use errx(3).
1.11	04-Sep-2014	mrg	port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.10	12-Apr-2009	lukem	branches: 1.10.24; Fix WARNS=4 issues (-Wshadow -Wcast-qual -Wsign-compare)
1.9	07-Aug-2003	agc	branches: 1.9.42; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.8	14-Jul-2003	itojun	use bounded string op
1.7	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.6	19-Dec-1998	christos	char -> unsigned char, braces for gcc-2.8.1
1.5	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.4	18-Oct-1997	mrg	merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.9.42.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.10.24.1	21-Sep-2014	snj	Pull up following revision(s) (requested by mrg in ticket #110): usr.bin/indent/io.c: revision 1.15 usr.bin/indent/indent_globs.h: revision 1.10 usr.bin/indent/args.c: revision 1.11 usr.bin/indent/indent.1: revision 1.23 usr.bin/indent/indent.c: revision 1.19 port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.13.16.1	10-Jun-2019	christos	Sync with HEAD
1.87.2.1	02-Aug-2025	perseant	Sync with HEAD
1.74	04-Jan-2025	rillig	indent: make debug log more uniform
1.73	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.72	03-Jan-2025	rillig	indent: fix line breaks in else-if sequences The flag ps.want_newline did not adequately model the conditions under which a line break should be inserted, thus the redesign. A welcome side effect is that in statements like 'if (cond);', the semicolon is now placed on a separate line, thus becoming more visible.
1.71	12-Dec-2024	rillig	indent: add error handling for I/O errors Suggested by lint2.
1.70	27-Jun-2023	rillig	branches: 1.70.2; indent: fix 'blank line above first statement in function body'
1.69	26-Jun-2023	rillig	indent: implement 'blank line above first statement in function body'
1.68	23-Jun-2023	rillig	indent: properly store parser state in debug mode The stacks in the parser state are allocated now and need to be copied individually. The test whether two paren stacks are equal was broken since 2023-06-14 14:11:28.
1.67	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.66	16-Jun-2023	rillig	indent: merge lexer symbols for type in/outside parentheses
1.65	16-Jun-2023	rillig	indent: add debug output for typedef declarations
1.64	16-Jun-2023	rillig	indent: don't force a blank line between '}' and preprocessing line
1.63	16-Jun-2023	rillig	indent: rename a field of the parser state The previous name 'comment_in_first_line' was misleading, as it could mean that there was a comment in the first line of the file. No functional change.
1.62	15-Jun-2023	rillig	indent: rename state variable to be more accurate No binary change.
1.61	14-Jun-2023	rillig	indent: clean up the code, add a few tests
1.60	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.59	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.58	14-Jun-2023	rillig	indent: clean up debugging code
1.57	14-Jun-2023	rillig	indent: clean up handling of comments One less moving part in the parser state. No functional change.
1.56	14-Jun-2023	rillig	indent: remove another flag from parser state When processing a comment, the flag ps.next_col_1 was not used for the next token, but for a line within a comment. As its scope was limited to a single comment, there is no need to store it any longer than that No functional change.
1.55	14-Jun-2023	rillig	indent: remove a redundant flag from the parser state No functional change.
1.54	14-Jun-2023	rillig	indent: merge parser symbols for stmt and stmt_list They were handled in exactly the same way.
1.53	10-Jun-2023	rillig	indent: rename misleading variable The name started with 'line_start', but the value is not always the value from the beginning of the line. No functional change.
1.52	10-Jun-2023	rillig	indent: fix debug output When the parser state was first printed, there were unintended diff markers. Treat the previous lexer symbol like the other parts of the parser state, as omitting it from the diff output is confusing.
1.51	10-Jun-2023	rillig	indent: fix line break between semicolon and brace
1.50	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.49	10-Jun-2023	rillig	indent: clean up function names, fix blank lines in debug output
1.48	10-Jun-2023	rillig	indent: distinguish blank lines from newline characters
1.47	10-Jun-2023	rillig	indent: clean up debug output In diff mode, don't print a diff of the very first parser state, instead print its full state. Don't print headings for empty sections of the parser state.
1.46	10-Jun-2023	rillig	indent: clean up function and variable names
1.45	10-Jun-2023	rillig	indent: explain right-aligned code
1.44	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.43	09-Jun-2023	rillig	indent: sync debug information for lexer symbols
1.42	09-Jun-2023	rillig	indent: don't treat function call expressions as cast expressions
1.41	09-Jun-2023	rillig	indent: when an indentation is ambiguous, indent one level further The '-eei' mode now applies whenever the indentation from a multi-line expression could be confused with a following statement.
1.40	08-Jun-2023	rillig	indent: remove fragile heuristic for detecting cast expressions The assumption that in an expression of the form '(a * anything)', the '*' marks a pointer type was too simple-minded. For now, fix the obvious cases and leave the others for later. If needed, they can be worked around using the '-T' option.
1.39	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.38	07-Jun-2023	rillig	indent: send all debug output to the same stream
1.37	06-Jun-2023	rillig	indent: sort functions in call order No functional change.
1.36	05-Jun-2023	rillig	indent: improve layout of debug output
1.35	05-Jun-2023	rillig	indent: sync debug output with parser state
1.34	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.33	04-Jun-2023	rillig	indent: track the kind of '{' on the parser stack
1.32	04-Jun-2023	rillig	indent: fix debug output of the parser symbol stack Even though the stack always contains a stmt_list as first element, print it nevertheless to avoid confusion about starting at index 1, and to provide the full picture.
1.31	04-Jun-2023	rillig	indent: rename struct field, for better symmetry No binary change outside debug mode.
1.30	04-Jun-2023	rillig	lint: use separate lexer symbols for 'case' and 'default' It's not strictly necessary since these tokens behave in the same way, still, the code is more straight-forward when there are separate tokens.
1.29	04-Jun-2023	rillig	indent: classify 'inline' as a modifier rather than a word
1.28	04-Jun-2023	rillig	indent: use separate lexer symbols for the different kinds of ':'
1.27	04-Jun-2023	rillig	indent: handle the indentation of 'case' in a simpler way
1.26	04-Jun-2023	rillig	indent: separate code for handling parentheses and brackets Handling parentheses is more complicated than for brackets.
1.25	02-Jun-2023	rillig	indent: clean up Only print the 'token' buffer in debug mode if it is interesting, group the blocks in handling of '(' tokens by topic, remove obsolete comment from test.
1.24	02-Jun-2023	rillig	indent: fix formatting of declarations with preprocessing lines
1.23	23-May-2023	rillig	indent: fix indentation of struct declarations
1.22	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.21	23-May-2023	rillig	indent: fix spacing in declarations in for loops
1.20	22-May-2023	rillig	indent: implement suppressing optional blank lines
1.19	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.18	20-May-2023	rillig	indent: implement blank line above block comment
1.17	20-May-2023	rillig	indent: implement blank line after function body
1.16	20-May-2023	rillig	indent: implement blank lines around conditional compilation
1.15	20-May-2023	rillig	indent: add debug logging for brace indentation No functional change outside debug mode, as the initialization of di_stack[0] was redundant.
1.14	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.13	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.12	17-May-2023	rillig	indent: fix indentation in preprocessor line No binary change.
1.11	16-May-2023	rillig	indent: allow comments in column 1 to be formatted
1.10	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.9	15-May-2023	rillig	indent: clean up detection of whether parentheses form a cast No functional change.
1.8	15-May-2023	rillig	indent: indent multi-line conditions No functional change.
1.7	15-May-2023	rillig	indent: document feature toggle for debugging output
1.6	15-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.5	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.4	13-May-2023	rillig	indent: implement 'blank after declarations'
1.3	13-May-2023	rillig	indent: use enum instead of magic numbers for tracking declarations No functional change.
1.2	13-May-2023	rillig	indent: add debug logging for enum token classification
1.1	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.70.2.1	02-Aug-2025	perseant	Sync with HEAD
1.33	09-Jun-2023	rillig	indent: indent multi-line expressions according to parentheses This reverts the FreeBSD change from 2004-02-12 that had been imported on 2019-04-04.
1.32	05-Jun-2023	rillig	indent: do not report broken lines, report configuration on stderr
1.31	12-May-2023	rillig	indent: sync manual page with recent changes
1.30	26-Sep-2021	rillig	indent: fix definition of -cli in manual page See io.c, compute_label_indent.
1.29	06-Mar-2021	rillig	indent.1: sort options alphabetically
1.28	04-Apr-2019	wiz	New sentence, new line. Whitespace fixes.
1.27	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.26	25-Feb-2016	wiz	branches: 1.26.16; Remove trailing whitespace.
1.25	24-Feb-2016	ginsbach	Remove double space before [.,:] in macro arguments.
1.24	24-Feb-2016	ginsbach	Add the [n]ei and [n]eei options to the synopsis; already documented in description.
1.23	04-Sep-2014	mrg	port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.22	13-Oct-2012	njoly	branches: 1.22.8; Remove a few unneeded Pp macros.
1.21	08-Apr-2012	wiz	branches: 1.21.2; Remove unnecessary Bk/Ek pairs from SYNOPSIS. No effective change except where I used the opportunity to sort options and/or option descriptions.
1.20	12-Jan-2011	wiz	branches: 1.20.6; Spell out parenthesis. From Ryo HAYASAKA in PR 44372.
1.19	24-Mar-2009	joerg	Remove physical markup.
1.18	11-Sep-2005	wiz	branches: 1.18.30; When marking up "C", use .Tn consisntently. From YOMURA Masanori in private mail.
1.17	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.16	25-Feb-2003	wiz	.Nm does not need a dummy argument ("") before punctuation or for correct formatting of the SYNOPSIS any longer.
1.15	01-Dec-2001	wiz	.Pp not necessary before or after .Ss/.Sh.
1.14	20-Jul-2001	kristerw	Correct a minor nroff nit. This closes PR bin/9220.
1.13	23-Mar-2001	fair	Correct one typo in the patch from PR 9220.
1.12	16-Mar-2001	fair	Commit patch from PR 9220 to document all options, and consistently document defaults. Also, clean up nroff nits.
1.11	24-Mar-1999	mycroft	Remove a spurious .ne.
1.10	22-Mar-1999	garbled	More and more .Os cleanups. .Os is defined in the tmac.doc-common file, so we shouldn't override it with versions in the manpages. Many more to come.
1.9	07-Mar-1999	mycroft	Clean up SYNOPSIS formatting.
1.8	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.7	18-Oct-1997	mrg	merge lite-2.
1.6	09-Jan-1997	tls	RCS ID police
1.5	27-Sep-1995	jtc	fix formatting of example; PR #1535
1.4	11-Jan-1994	jtc	Fix spelling errors.
1.3	07-Aug-1993	cgd	do block commenting, if comment begins with slash-star-newline.
1.2	01-Aug-1993	mycroft	Add RCS indentifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	30-Dec-1993	mrg	4.4BSD-Lite2
1.18.30.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.20.6.2	30-Oct-2012	yamt	sync with head
1.20.6.1	17-Apr-2012	yamt	sync with head
1.21.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.22.8.1	21-Sep-2014	snj	Pull up following revision(s) (requested by mrg in ticket #110): usr.bin/indent/io.c: revision 1.15 usr.bin/indent/indent_globs.h: revision 1.10 usr.bin/indent/args.c: revision 1.11 usr.bin/indent/indent.1: revision 1.23 usr.bin/indent/indent.c: revision 1.19 port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.26.16.1	10-Jun-2019	christos	Sync with HEAD
1.396	07-Jan-2025	rillig	indent: condense and simplify parsing code
1.395	04-Jan-2025	rillig	indent: fix indentation of adjacent multi-line initializers The main topic of this change is parse.c:66, which makes the indentation of statements uniform with the indentation of other parser symbols. That change had the side effect of messing up the indentation of files whose first line does not start in column 1, such as in ps_ind_level.c. To fix this side effect, the initial indentation must be determined before pushing the placeholder token psym_stmt during initialization.
1.394	04-Jan-2025	rillig	indent: make debug log more uniform
1.393	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.392	03-Jan-2025	rillig	indent: fix line breaks in else-if sequences The flag ps.want_newline did not adequately model the conditions under which a line break should be inserted, thus the redesign. A welcome side effect is that in statements like 'if (cond);', the semicolon is now placed on a separate line, thus becoming more visible.
1.391	12-Dec-2024	rillig	indent: add error handling for I/O errors Suggested by lint2.
1.390	03-Dec-2023	rillig	branches: 1.390.2; indent: inline input-related macros No binary change.
1.389	03-Dec-2023	rillig	indent: group input-related variables into a struct No functional change.
1.388	03-Dec-2023	rillig	indent: use line number of the token start in diagnostics Previously, the line number of the end of the token was used, which was confusing in debug mode.
1.387	27-Jun-2023	rillig	indent: fix 'blank line above first statement in function body'
1.386	26-Jun-2023	rillig	indent: implement 'blank line above first statement in function body'
1.385	26-Jun-2023	rillig	indent: in -bad mode, don't add a blank line above a comment or '}'
1.384	25-Jun-2023	rillig	indent: move cast detection from the lexer to the main processor It is not the job of the lexer to modify the parser state.
1.383	25-Jun-2023	rillig	indent: fix formatting of parenthesized name in function definition
1.382	23-Jun-2023	rillig	indent: properly store parser state in debug mode The stacks in the parser state are allocated now and need to be copied individually. The test whether two paren stacks are equal was broken since 2023-06-14 14:11:28.
1.381	18-Jun-2023	rillig	indent: remove support for backspace in code and comments The C code in the whole tree does not contain a single literal backspace.
1.380	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.379	16-Jun-2023	rillig	indent: merge lexer symbols for type in/outside parentheses
1.378	16-Jun-2023	rillig	indent: fix spacing between postfix operator and left parenthesis
1.377	16-Jun-2023	rillig	indent: improve heuristics for cast expressions
1.376	16-Jun-2023	rillig	indent: improve heuristics for cast expressions
1.375	16-Jun-2023	rillig	indent: improve heuristics for casts
1.374	16-Jun-2023	rillig	indent: fix indentation and linebreaks in typedef declarations
1.373	16-Jun-2023	rillig	indent: don't force a blank line between '}' and preprocessing line
1.372	15-Jun-2023	rillig	indent: consolidate handling of statement continuations
1.371	15-Jun-2023	rillig	indent: rename state variable to be more accurate No binary change.
1.370	15-Jun-2023	rillig	indent: fix indentation of multi-line enum constant initializers
1.369	15-Jun-2023	rillig	indent: miscellaneous cleanups, more tests for edge cases
1.368	15-Jun-2023	rillig	indent: fix alignment of multi-line declarations
1.367	14-Jun-2023	rillig	indent: clean up the code, add a few tests
1.366	14-Jun-2023	rillig	indent: allow more than 128 brace levels
1.365	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.364	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.363	14-Jun-2023	rillig	indent: merge duplicate code
1.362	14-Jun-2023	rillig	indent: fix formatting of comment after 'switch (expr)'
1.361	14-Jun-2023	rillig	indent: use correct preprocessing directive in error message
1.360	14-Jun-2023	rillig	indent: allow more than 5 levels of #if/#endif
1.359	14-Jun-2023	rillig	indent: remove another flag from parser state When processing a comment, the flag ps.next_col_1 was not used for the next token, but for a line within a comment. As its scope was limited to a single comment, there is no need to store it any longer than that No functional change.
1.358	14-Jun-2023	rillig	indent: merge parser symbols for stmt and stmt_list They were handled in exactly the same way.
1.357	10-Jun-2023	rillig	indent: rename misleading variable The name started with 'line_start', but the value is not always the value from the beginning of the line. No functional change.
1.356	10-Jun-2023	rillig	indent: fix debug output When the parser state was first printed, there were unintended diff markers. Treat the previous lexer symbol like the other parts of the parser state, as omitting it from the diff output is confusing.
1.355	10-Jun-2023	rillig	indent: fix line break between semicolon and brace
1.354	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.353	10-Jun-2023	rillig	indent: in debug mode, null-terminate buffers
1.352	10-Jun-2023	rillig	indent: fix indentation of continuation lines in initializers
1.351	10-Jun-2023	rillig	indent: clean up function and variable names
1.350	10-Jun-2023	rillig	indent: fix token classification in declarations As a side effect, indent handles _Generic from C11 properly now, at least in -nlp mode.
1.349	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.348	09-Jun-2023	rillig	indent: trim trailing blank lines
1.347	09-Jun-2023	rillig	indent: group lexer symbols by topic, sort processing functions No functional change.
1.346	09-Jun-2023	rillig	indent: support C99 compound literals
1.345	09-Jun-2023	rillig	indent: don't treat function call expressions as cast expressions
1.344	09-Jun-2023	rillig	indent: eliminate unused variable No functional change.
1.343	09-Jun-2023	rillig	indent: when an indentation is ambiguous, indent one level further The '-eei' mode now applies whenever the indentation from a multi-line expression could be confused with a following statement.
1.342	09-Jun-2023	rillig	indent: format its own code
1.341	08-Jun-2023	rillig	indent: remove fragile heuristic for detecting cast expressions The assumption that in an expression of the form '(a * anything)', the '*' marks a pointer type was too simple-minded. For now, fix the obvious cases and leave the others for later. If needed, they can be worked around using the '-T' option.
1.340	08-Jun-2023	rillig	indent: fix indentation of initializer lists with designators
1.339	08-Jun-2023	rillig	indent: clean up and condense code No functional change.
1.338	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.337	06-Jun-2023	rillig	indent: compute indentation of 'case' labels on-demand One less moving part to keep track of. No functional change.
1.336	05-Jun-2023	rillig	indent: in 'if (expr)', the parentheses do not form a cast expression No functional change. When stepping through the code in debug mode, it was just too confusing that indent would log an 'unknown cast' in this situation.
1.335	05-Jun-2023	rillig	indent: format own source code
1.334	05-Jun-2023	rillig	indent: don't remove blank line after 'if (expr) {'
1.333	05-Jun-2023	rillig	indent: do not report broken lines, report configuration on stderr
1.332	05-Jun-2023	rillig	indent: fix formatting of 'do' statements
1.331	05-Jun-2023	rillig	indent: make heuristics for '*' pointer types simpler Previously, a '}' token did not reset the state machine, but it should.
1.330	05-Jun-2023	rillig	indent: fix trailing whitespace after comment
1.329	05-Jun-2023	rillig	indent: rename variables, clean up comments No binary change.
1.328	05-Jun-2023	rillig	indent: clean up handling of whitespace No functional change.
1.327	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.326	04-Jun-2023	rillig	indent: track the kind of '{' on the parser stack
1.325	04-Jun-2023	rillig	indent: ensure that the 'block init level' never goes negative No functional change.
1.324	04-Jun-2023	rillig	indent: rename struct field, for better symmetry No binary change outside debug mode.
1.323	04-Jun-2023	rillig	indent: fix formatting of compound expressions, at least partially
1.322	04-Jun-2023	rillig	lint: use separate lexer symbols for 'case' and 'default' It's not strictly necessary since these tokens behave in the same way, still, the code is more straight-forward when there are separate tokens.
1.321	04-Jun-2023	rillig	indent: classify 'inline' as a modifier rather than a word
1.320	04-Jun-2023	rillig	indent: use separate lexer symbols for the different kinds of ':'
1.319	04-Jun-2023	rillig	indent: handle the indentation of 'case' in a simpler way
1.318	04-Jun-2023	rillig	indent: separate code for handling parentheses and brackets Handling parentheses is more complicated than for brackets.
1.317	03-Jun-2023	rillig	indent: fix indentation of adjacent '{'
1.316	03-Jun-2023	rillig	indent: clean up handling of brace indentation No functional change.
1.315	02-Jun-2023	rillig	indent: force each statement on a new line Previously, '{} while (cond)' was kept on a single line, even though the 'while' was independent of the '{}'.
1.314	02-Jun-2023	rillig	indent: remove newline between 'switch' and '{'
1.313	02-Jun-2023	rillig	indent: improve heuristics of classifying '*' as pointer or operator
1.312	02-Jun-2023	rillig	indent: clean up Only print the 'token' buffer in debug mode if it is interesting, group the blocks in handling of '(' tokens by topic, remove obsolete comment from test.
1.311	02-Jun-2023	rillig	indent: fix formatting of declarations with preprocessing lines
1.310	23-May-2023	rillig	indent: separate code for handling enums from the lexer The lexer's responsibility is to generate tokens, it's not supposed to update the parser state. Centralize the state transitions that control indentation of enum constants to keep the lexer code clean. Skip comments, newlines and preprocessing lines when updating the parser state for enum constants and for '*' in declarations.
1.309	23-May-2023	rillig	indent: fix indentation of struct declarations
1.308	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.307	23-May-2023	rillig	indent: extract processing of a single token to separate function No functional change.
1.306	23-May-2023	rillig	indent: fix spacing around '*' in declarations
1.305	23-May-2023	rillig	indent: fix spacing in declarations in for loops
1.304	22-May-2023	rillig	indent: fix spacing between block braces
1.303	22-May-2023	rillig	indent: implement suppressing optional blank lines
1.302	21-May-2023	rillig	indent: don't read out-of-bounds memory in preprocessing lines (Since a few minutes.) If a line '#if 0' was followed by an unlikely line '#', the second line was interpreted as '#if' as well. To detect this bug automatically, a dynamic analysis tool would need to know that only the memory between lab.mem and lab.mem + lab.len has defined content. This constraint, in turn, would throw up at the bottom of copy_comment_wrap, which for a brief moment intentionally violates this constraint.
1.301	21-May-2023	rillig	indent: don't error out on unrecognized preprocessor directives This allows indent to be used on the GCC preprocessor output.
1.300	20-May-2023	rillig	indent: remove redundant checks in processing of '}' No functional change.
1.299	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.298	20-May-2023	rillig	indent: implement blank line after function body
1.297	20-May-2023	rillig	indent: implement blank lines around conditional compilation
1.296	20-May-2023	rillig	indent: add debug logging for brace indentation No functional change outside debug mode, as the initialization of di_stack[0] was redundant.
1.295	18-May-2023	rillig	indent: remove detailed rules for blank before comment
1.294	18-May-2023	rillig	indent: rename a few functions No functional change.
1.293	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.292	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.291	18-May-2023	rillig	indent: remove unnecessary variable size optimization Due to the enum that follows in the struct, the short variable was padded to 4 bytes anyway. No functional change.
1.290	16-May-2023	rillig	indent: directly access the input buffer No functional change.
1.289	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.288	16-May-2023	rillig	indent: remove blank between comment and parentheses or brackets Finally, indent formats its own source code without messing up the layout.
1.287	16-May-2023	rillig	indent: fix handling of INDENT OFF/ON comments Previously, the 'INDENT OFF' comments were interpreted when the newline token from the line above the comment was processed, which was earlier than could be reasonably expected. The 'INDENT ON' comments were interpreted equally early, which led to the situation that the 'INDENT OFF' comments were preserved literally but the 'INDENT ON' comments weren't.
1.286	15-May-2023	rillig	indent: clean up detection of whether parentheses form a cast No functional change.
1.285	15-May-2023	rillig	indent: fix cast detection In process_lparen_or_lbracket, ps.paren[...].maybe_cast was not initialized, which may have been the cause for seemingly random spacing around binary operators. While here, clean up the code by reducing the number of accesses to the parser state.
1.284	15-May-2023	rillig	indent: fix detection of casts A word followed by a '(' does not start a cast expression.
1.283	15-May-2023	rillig	indent: fix type cast in function definition
1.282	15-May-2023	rillig	indent: fix duplicate space between comment and binary operator
1.281	15-May-2023	rillig	indent: format its own code, extend some comments With manual corrections, as there are still some bugs left. No functional change.
1.280	15-May-2023	rillig	indent: improve type guessing, fix formatting of declarations
1.279	15-May-2023	rillig	indent: fix spacing between function prototype and attributes
1.278	15-May-2023	rillig	indent: fix indentation of struct member names
1.277	15-May-2023	rillig	indent: indent multi-line conditions No functional change.
1.276	15-May-2023	rillig	indent: fix indentation of statements after controlling expression
1.275	15-May-2023	rillig	indent: fix indentation of expressions in -nlp -eei mode
1.274	15-May-2023	rillig	indent: fix indentation of multi-line '?:' expressions in functions
1.273	15-May-2023	rillig	indent: let indent format its own code With manual corrections, as indent does not properly indent multi-line '?:' expressions nor multi-line controlling expressions.
1.272	15-May-2023	rillig	indent: fix spacing in for loop with declaration (since 2022-02-13)
1.271	15-May-2023	rillig	indent: remove redundant include lines
1.270	15-May-2023	rillig	indent: clean up memory allocation No functional change.
1.269	15-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.268	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.267	14-May-2023	rillig	indent: only null-terminate the buffers if necessary The only case where a buffer is used as a C-style string is when looking up a keyword. No functional change.
1.266	14-May-2023	rillig	indent: remove foreign RCS IDs
1.265	14-May-2023	rillig	indent: miscellaneous cleanups
1.264	13-May-2023	rillig	indent: prevent undefined behavior on unbalanced parentheses
1.263	13-May-2023	rillig	indent: do not add a blank at the beginning of a line Most calls to output_line did already reset the variable. There may be some untested edge cases in or after comments, but these should be fine as well.
1.262	13-May-2023	rillig	indent: do not add a space before a comment that starts a line
1.261	13-May-2023	rillig	indent: replace __dead functions with return statements No functional change.
1.260	13-May-2023	rillig	indent: use enum instead of magic numbers for tracking declarations No functional change.
1.259	13-May-2023	rillig	indent: rename struct fields for buffers No binary change except for assertion line numbers.
1.258	13-May-2023	rillig	indent: clean up a condition, add comments No functional change.
1.257	13-May-2023	rillig	indent: preserve indentation of preprocessor directives
1.256	12-May-2023	rillig	indent: rename placeholder symbol for parser stack No functional change outside debug mode.
1.255	12-May-2023	rillig	indent: remove code for parsing declarations without semicolon The statement from the comment that declarations do not need semicolons is wrong. A possible input that matched this rule is 'void f(void) { int a }'.
1.254	12-May-2023	rillig	indent: remove statistics The numbers from the statistics were wrong.
1.253	12-May-2023	rillig	indent: condense code for handling spaced expressions No functional change outside debug mode.
1.252	11-May-2023	rillig	indent: don't touch comments in preprocessing lines The indentation of multi-line comments was wrong, and the code for handling them was too complicated.
1.251	11-May-2023	rillig	indent: remove broken code for handling blank lines This fixes several bugs where blank lines were erroneously added or removed, treating these old bugs for new bugs in different places. These new bugs are expected to be easier to fix, as the old bugs will not interfere anymore.
1.250	11-May-2023	rillig	indent: move parser state variables to the parser_state struct Include the variables in the debug output.
1.249	11-May-2023	rillig	indent: eliminate a local variable for else-if handling No functional change intended.
1.248	11-May-2023	rillig	indent: move force_nl into the parser state This way, it is included in the debug output. No functional change.
1.247	11-May-2023	rillig	indent: remove unnecessary assignments to last_else No functional change intended.
1.246	11-May-2023	rillig	indent: remove buggy code for swapping tokens It is not the job of an indenter to swap tokens, even if it's only about placing comments elsewhere. The code that swapped the tokens was complicated, buggy and impossible to understand. In -br (brace right) mode, indent no longer moves a '{' from the beginning of a line to the end of the previous line, as that was handled by the token swapping code as well. This change is unintended, but it will be easier to re-add that now that the code is simpler.
1.245	09-May-2022	rillig	indent: clean up control flow, remove Capsicum No functional change.
1.244	23-Apr-2022	rillig	indent: group global variables related to output control No functional change.
1.243	23-Apr-2022	rillig	indent: remove Capsicum support NetBSD doesn't have Capsicum.
1.242	13-Feb-2022	rillig	indent: rename parser_state.p_l_follow and paren_level The previous variable names were misleading. Paren_level is not the current level of parentheses but the one from the beginning of the current output line. For better accuracy, rename it to line_start_paren_level. P_l_follow is not the level of parentheses that will be active at some point in the future, as the previous name suggested. Instead, it is the level of parentheses right now. For better accuracy, rename it to nparen. This nicely matches its main usage, which is as index to the parser_state.paren array. No binary change.
1.241	13-Feb-2022	rillig	indent: replace bitmasking code with struct The struct directly represents the properties of a pair of parentheses, without forcing the human reader to decode any bitset. This makes it easier to find the remaining bugs in the heuristic for determining the kind of parentheses. No functional change outside debug mode.
1.240	13-Feb-2022	rillig	indent: change parser_state.cast_mask to 0-based indexing Having 1-based indexing was completely unexpected, and it didn't match the 0-based indexing of parser_state.paren_indents. No functional change.
1.239	28-Nov-2021	rillig	indent: treat L"string" as a single token There is never whitespace between the 'L' and the string literal or the character constant. There might be a backslash-newline between them, but that case was not handled before either. No functional change.
1.238	28-Nov-2021	rillig	indent: clean up and document input handling The transformation of moving comments from after an 'if (expr)' after the following brace has a large implementation cost (about 300 lines of code) and makes input handling quite complicated. Document the overall idea to save future readers some time. No functional change.
1.237	27-Nov-2021	rillig	indent: accept a few formatting suggestions from indent The remaining issues are still that the conditions look ambiguous even with -eei, and that __attribute__ is broken into a separate line. No functional change.
1.236	27-Nov-2021	rillig	indent: rename dump functions to output No functional change.
1.235	27-Nov-2021	rillig	indent: inline switch_buffer The function name was not accurate all the time. Now that inp_from_comment is a separate function, it doesn't make sense anymore to offload the 3 simple statements to a separate function. No functional change.
1.234	26-Nov-2021	rillig	indent: add buf_add_range for adding characters to a buffer No functional change.
1.233	26-Nov-2021	rillig	indent: move ind_add from io.c to indent.c It's a general-purpose function that is not directly related to input or output.
1.232	25-Nov-2021	rillig	indent: rename ps.in_function_parameters to match reality This flag is only set while parsing the parameters of a function definition, but not for a function declaration. See buffer_add in the test fmt_decl. No functional change.
1.231	25-Nov-2021	rillig	indent: rename ps.in_stmt to in_stmt_or_decl The previous name didn't match reality. No functional change.
1.230	25-Nov-2021	rillig	indent: rename ps.ind_stmt to in_stmt_cont This makes a comment redundant. No functional change.
1.229	25-Nov-2021	rillig	indent: clean up style No functional change.
1.228	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.227	19-Nov-2021	rillig	indent: fix included headers
1.226	19-Nov-2021	rillig	indent: replace ps.procname with ps.is_function_definition Omly the first character of ps.procname was ever read, and it was only compared to '\0'. Using a bool for this means simpler code, less memory and fewer wasted CPU cycles due to the removed strncpy. No functional change.
1.225	19-Nov-2021	rillig	indent: remove all references to inbuf from indent.c No functional change.
1.224	19-Nov-2021	rillig	indent: move character input handling from indent.c to io.c No functional change.
1.223	19-Nov-2021	rillig	indent: move character input from indent.c to io.c No functional change.
1.222	19-Nov-2021	rillig	indent: replace direct access to the input buffer This is a preparation for abstracting away all the low-level details of handling the input. The goal is to fix the current bugs regarding line number counting, out of bounds memory access, and generally unreadable code. No functional change.
1.221	19-Nov-2021	rillig	indent: add debug logging for input buffer handling
1.220	19-Nov-2021	rillig	indent: rename input buffer variables From reading the names 'save_com' and 'sc_end', it was not obvious enough that these two variables are the limits of the same buffer, the names were just too unrelated. No functional change.
1.219	19-Nov-2021	rillig	indent: group variables for input handling No functional change.
1.218	07-Nov-2021	rillig	indent: fix handling of C99 comments after 'if (expr)'
1.217	07-Nov-2021	rillig	indent: demonstrate disappearing form feed
1.216	07-Nov-2021	rillig	indent: various cleanups Make several comments more precise. Rename process_end_of_file to process_eof to match the token name. Change the order of assignments in analyze_comment to keep the com_ind computations closer together. In copy_comment_wrap, use pointer difference instead of pointer addition to stay away from undefined behavior. No functional change.
1.215	07-Nov-2021	rillig	indent: rename ps.decl_nest to decl_level This better matches the comment. No functional change.
1.214	07-Nov-2021	rillig	indent: reduce negations in process_else, clean up comments No functional change.
1.213	07-Nov-2021	rillig	indent: only access buffer data in the range [buf.s, buf.e) No functional change.
1.212	07-Nov-2021	rillig	indent: rename type_at_paren_level_0 to type_outside_parentheses For symmetry with type_in_parentheses. No functional change.
1.211	07-Nov-2021	rillig	indent: distinguish between typename in parentheses and other words This gets rid of two members of parser_state. No functional change for well-formed programs. The sequence of '++int' or '--size_t' may be formatted differently than before, but no program is expected to contain that sequence. Rename lsym_ident to lsym_word since 'ident' was too specific. This token type is used for constants and string literals as well. Strictly speaking, a string literal is not a word, but at least it's better than before.
1.210	07-Nov-2021	rillig	indent: rename 'inbuf' functions to 'inp' The variable 'inp' used to be named 'inbuf'. Make the function names correspond to the variable name again. No functional change.
1.209	05-Nov-2021	rillig	indent: rename process_keyword_do to process_do, same for 'else' Before the symbols from the tokenizer had the prefix 'lsym', the symbols could not be simply called 'else' and 'do'. The functions for processing the tokens followed that naming scheme. When the prefix 'lsym' was introduced, the word 'keyword' was no longer needed, neither in the constants nor in the function names. No functional change.
1.208	05-Nov-2021	rillig	indent: rename ps.curr_newline to next_col_1 For symmetry with ps.curr_col_1. No functional change.
1.207	04-Nov-2021	rillig	indent: split process_comment_in_code into separate functions No functional change.
1.206	04-Nov-2021	rillig	indent: fix joining of adjacent unary '+' operators
1.205	03-Nov-2021	rillig	indent: inline indentation_after, shorten function name to ind_add There were only few calls to indentation_after, so inlining it spares the need to look at yet another function definition. Another effect is that code.s and code.e appear in the code as a pair now, instead of a single code.s, making the scope of the function call obvious. In ind_add, there is no need to check for '\0' anymore since none of the buffers can ever contain a null character, these are filtered out by inbuf_read_line. No functional change.
1.204	01-Nov-2021	rillig	indent: fix missing blank after 'return' (since 2021-10-31) In indent.c 1.200 from 2021-10-31, the subtypes of identifier tokens were removed since they were redundant. An unintended side effect was that a parenthesized expression after 'return' was no longer separated by a blank. Before that change, 'return' was tokenized as an lsym_ident with subtype kw_other, and want_space_before_lparen handled this case in the last line. After the change, 'return' was treated as an ordinary identifier, and unless the option '-pcs' (blank after function call) was given, the blank was removed. The other keywords that had kw_other are not affected since they do not expect a '(' afterwards. These keywords are 'break', 'continue', 'goto', 'inline' and 'restrict'. Curiously, there was not a single test case that covered 'return(expr)'. While here, remove the trailing ',' from the enum lexer_symbol, which is not allowed in standard C, it is a GNU extension. Lint doesn't complain about this since the default LINTFLAGS include '-g' for GCC mode.
1.203	31-Oct-2021	rillig	indent: clean up Initialize buffers in reading order, make comments more expressive, rename add_typename to register_typename, remove unused macro. No functional change.
1.202	31-Oct-2021	rillig	indent: for '-pcs', add blank between function and '(' Before indent-2021.09.30.21.48.12, the blank had always been added, even in '-npcs' mode. Since then, the blank had never been added. Now, add the blank in '-pcs' mode and omit it in '-npcs' mode.
1.201	31-Oct-2021	rillig	indent: replace kw_tag with lsym_tag This leaves only one special type of token, which is lsym_ident, which in some cases represents a type name and in other cases an identifier, constant or string literal. No functional change.
1.200	31-Oct-2021	rillig	indent: replace simple cases of keyword_kind with lexer_symbol The remaining keyword kinds 'tag' and 'type' require a bit more thought, so do them in a separate step. No functional change.
1.199	31-Oct-2021	rillig	indent: rename lsym_type to better reflect reality Type names that occur in parentheses are parsed as lsym_ident having the subtype kw_type instead. No functional change.
1.198	31-Oct-2021	rillig	indent: add separate lexer symbol for offsetof No functional change.
1.197	31-Oct-2021	rillig	indent: add separate lexer symbol for sizeof The plan is to get rid of the type keyword_kind, which largely overlaps with lexer_symbol. No functional change.
1.196	30-Oct-2021	rillig	indent: push down variable comment_buffered No functional change.
1.195	30-Oct-2021	rillig	indent: rename prev_newline and prev_col_1 to curr These two flags describe the token that is currently processed. In process_binary_op, curr_newline can never be true since newline is not a binary operator, so remove that condition. No functional change.
1.194	30-Oct-2021	rillig	indent: reorder assignments in switch_buffer No functional change.
1.193	30-Oct-2021	rillig	indent: move buffer functions further up No functional change.
1.192	30-Oct-2021	rillig	indent: group variables by topic No functional change.
1.191	30-Oct-2021	rillig	indent: prevent buffer overflow in search_stmt_comment printf '{ if (%010000d) /comment/ ; }' '0' \| indent
1.190	30-Oct-2021	rillig	indent: add debug logging for save_com This will help in finding the proper fix for the assertion failure in search_stmt_comment. Add an assertion in search_stmt_lbrace to prevent the previous, incomplete fix from being applied again.
1.189	30-Oct-2021	rillig	indent: prevent buffer overflows in 'if (expr) ... stmt'
1.188	30-Oct-2021	rillig	indent: revert previous fix of assertion failure The strange code with the out of bounds memory access is needed to transform 'if (expr) /* comment / {' to 'if (expr) { / comment */', that is, to move the comment to the right. Add a test that prevents "repairing" this code again.
1.187	30-Oct-2021	rillig	indent: fix assertion failure in search_stmt_comment I have no idea why the code was written in such a convoluted way before. By removing all the code that didn't make sense, everything just works as expected, and the existing tests all pass, especially those in token_comment.c that mention search_stmt_comment.
1.186	30-Oct-2021	rillig	indent: replace tabsize with hardcoded 8 in process_comma On 2018-07-25, FreeBSD added the option '-ts' to make the tabulator size configurable, replacing several constants 7, 8, 9 with tabsize. The 8 in the expression 'max_col - 8' was not related to the tabulator size but instead represents the typical width of a variable name. Subtracting a tab from the right margin doesn't make sense since the right margin need not be aligned on a tabstop. See the test fmt_decl.c, where the declaration 'struct s0 a,b;' is split into several lines because the estimate for the variable name following the comma is too high. There would have been plenty of space to the right to keep the whole declaration in a single line. No functional change.
1.185	30-Oct-2021	rillig	indent: don't risk a buffer overflow in code_add_decl_indent The buffers have a safety margin of 5 characters, so the bounds check is not strictly necessary. It makes the code more uniform though. No functional change.
1.184	30-Oct-2021	rillig	indent: clean up code_add_decl_indent In layout computations, it is helpful for human readers to list the summands in logical order. In this case, the expression 'code_len + base_ind' was rather confusing, so replace it with 'base_ind + code_len'. This makes the code straight-forward enough that it doesn't need any comments anymore. No functional change.
1.183	30-Oct-2021	rillig	indent: remove confusing modulo from code_add_decl_indent The only effects of the modulo operation was to make indent slower and to confuse human readers. During the computation of the indentation, the main focus is on the difference between the current indentation, as computed from the base indentation and the current code, and the target indentation. All these computations take opt.tabsize into account. When looking only at the difference, whether or not a multiple of opt.tabsize is added does not matter. No functional change.
1.182	30-Oct-2021	rillig	indent: inline bloated call to 'parse' during initialization No functional change.
1.181	30-Oct-2021	rillig	indent: condense code for parsing command line arguments Previously, the cascade of 'if' statements suggested that there were 6 different cases to be handled when in reality there are only 3: no arguments, 1 argument, 2 arguments. Let the code express this directly. No functional change.
1.180	30-Oct-2021	rillig	indent: extract main_load_profiles from main_parse_command_line No functional change.
1.179	29-Oct-2021	rillig	indent: remove redundant comments, remove punctuation from debug log The comment about 'null stmt' between braces probably meant 'no statements between braces'. The comments at psym_switch_expr only repeated what the code says or had been outdated 29 years ago already since opt.case_indent does not have to be 'one level down'. In the debug log, the quotes around the symbol names are not necessary after a ':'. The parse stack also does not need this much punctuation. Reducing a do-while loop to nothing instead of a statement saves a few CPU cycles. It works because after each lbrace, a stmt is pushed to the parser stack. This stmt can only ever be reduced to a stmt_list but never be removed.
1.178	29-Oct-2021	rillig	indent: fix missing blank before binary operator
1.177	29-Oct-2021	rillig	indent: merge isblank and is_hspace into ch_isblank No functional change.
1.176	29-Oct-2021	rillig	indent: replace segmentation fault with assertion
1.175	29-Oct-2021	rillig	indent: initialize 'ps' via code This saves 3 kB of binary size since the parser state is rather large and only very few members are initialized to non-zero values. No functional change.
1.174	29-Oct-2021	rillig	indent: clean up main_init_globals No functional change.
1.173	29-Oct-2021	rillig	indent: fix undefined behavior in buffer handling Adding an arbitrary integer to a pointer may result in an out of bounds pointer, so replace the addition with a pointer subtraction. In the buffer handling functions, handle 'buf' and 'l' before 's' and 'e', since they are pairs. In inbuf_read_line, use 's' instead of 'buf' to make the code easier to understand for human readers. No functional change.
1.172	29-Oct-2021	rillig	indent: mark obviously broken code
1.171	29-Oct-2021	rillig	indent: use prev/curr/next to refer to the current token The word 'last' just didn't match with 'next'. No functional change.
1.170	29-Oct-2021	rillig	indent: rename ps.dumped_decl_indent and indent_declaration The word 'dump' in 'ps.dumped_decl_indent' was too close to dump_line, which led to confusion since the variable controls whether the indentation has been added to the code buffer, which happens way before actually dumping the current line to the output file. The function name 'indent_declaration' was too unspecific, it did not reveal where the indentation of the declaration actually happened. No functional change.
1.169	29-Oct-2021	rillig	indent: keep p_l_follow nonnegative, use consistent comparison No functional change.
1.168	29-Oct-2021	rillig	indent: spell 'parentheses' properly in messages and comments
1.167	28-Oct-2021	rillig	indent: clean up indentation, comments, reduce No functional change.
1.166	28-Oct-2021	rillig	indent: remove unused local variable in lexi Since the previous commit, lexi is always called with the same argument, so remove that parameter. The previous commit broke the debug logging by not printing "transient state" anymore. Replace this with "rolled back parser state" at the caller's site. No functional change.
1.165	28-Oct-2021	rillig	indent: reduce negations in search_stmt_lookahead No functional change.
1.164	28-Oct-2021	rillig	indent: clean up comments and function names Having accurate names for the lexer symbols and the parser symbols makes most of the comments redundant. Remove these. Rename process_decl to process_type, to match the name of the corresponding lexer symbol. In this phase, it's just a single type token, not a whole declaration. No functional change.
1.163	28-Oct-2021	rillig	indent: make error messages for option parsing more precise
1.162	26-Oct-2021	rillig	indent: clean up process_comment There is no undefined behavior since the compared characters are always from the basic execution character set. All other cases are covered by the condition above for now_len. Fix debug logging for non-ASCII characters, previously a character was output as \xffffffc3.
1.161	26-Oct-2021	rillig	indent: make ps.keyword easier to understand Previously, ps.keyword did not have any documentation and was not straight-forward. In some cases it was reset to kw_0, in others it was set to an interesting value. The idea behind it was to remember the kind of word of the previous token, to decide whether to have a space between sizeof or offsetof and a following '('. No functional change.
1.160	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.159	25-Oct-2021	rillig	indent: improve debug logging Output the various details in chronological order.
1.158	25-Oct-2021	rillig	indent: rename search_brace to search_stmt No functional change.
1.157	25-Oct-2021	rillig	indent: rename local variable sp_sw to spaced_expr The 'sp' probably meant 'space-enclosed'; no idea what 'sw' was meant to mean. Maybe 'switch', but that would have been rather ambiguous when talking about control flow statements. No functional change.
1.156	25-Oct-2021	rillig	indent: split type token_type into 3 separate types Previously, token_type was used for 3 different purposes: 1. symbol types from the lexer 2. symbol types on the parser stack 3. kind of control statement for 'if (expr)' and similar statements Splitting the 41 constants into separate types makes it immediately clear that the parser stack never handles comments, preprocessing lines, newlines, form feeds, the inner structure of expressions. Previously, the constant switch_expr was especially confusing since it was used for 3 different purposes: when returned from lexi, it represented the keyword 'switch', in the parser stack it represented 'switch (expr)', and it was used for a statement head as well. The only overlap between the lexer symbols and the parser symbols are '{' and '}', and the keywords 'do' and 'else'. To increase confusion, the constants of the previous token_type were in apparently random order and before 2021, they had cryptic, highly abbreviated names. No functional change.
1.155	24-Oct-2021	rillig	indent: rename form_feed to tt_lex_form_feed No functional change.
1.154	24-Oct-2021	rillig	indent: split kw_for_or_if_or_while into separate constants No functional change.
1.153	24-Oct-2021	rillig	indent: split kw_do_or_else into separate constants It was unnecessarily confusing to have the token types keyword_do_else, keyword_do and keyword_else at the same time, without any hint in what they differed. Some of the token types seem to be used by the lexer while others are used in the parse stack. Maybe all token types can be partitioned into these groups, which would suggest to use two different types for them. And if not, it's still clearer to have this distinction in the names of the constants. No functional change.
1.152	24-Oct-2021	rillig	indent: rename seen_quest to quest_level The new name aligns with other similar variables like ind_level, case_ind_level and ifdef_level. The old name 'seen' is mainly used for bool variables. No functional change.
1.151	24-Oct-2021	rillig	indent: fix indentation of ad-hoc tagged variables Seen among others in usr.bin/indent/lexi.c, variable 'keywords'.
1.150	24-Oct-2021	rillig	indent: initialize variables in main_loop in declaration No functional change.
1.149	24-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards. Indent still does not get extra_expr_indent correctly, it also indents global variables after tagged declarations too deep. No functional change.
1.148	24-Oct-2021	rillig	indent: clean up format of warnings and errors Previously, warnings and errors had the form of C block comments. Before NetBSD io.c 1.20 from 2019-10-19, this format made sense because the diagnostics could end up in the same output stream as the formatted output. Since NetBSD io.c 1.20 from 2019-10-19, all diagnostics are redirected to stderr. This change was not mentioned in the commit message back then, it makes sense nevertheless. Since stdout and stderr now are properly separated, there is no need anymore to keep the weird format for warnings and errors. Switch to the standard 'error: file:line' format. Move the function 'diag' to indent.c to have access to the name of the current input file.
1.147	24-Oct-2021	rillig	indent: fix line number counting at beginning of function body
1.146	24-Oct-2021	rillig	indent: rename nitems to array_length
1.145	24-Oct-2021	rillig	indent: replace global variable use_ff with function parameter
1.144	20-Oct-2021	rillig	indent: rename ps.last_u_d to match its comment No functional change.
1.143	20-Oct-2021	rillig	indent: rename parser stack variables No functional change.
1.142	20-Oct-2021	rillig	indent: rename blankline_requested variables The words 'prefix' and 'postfix' sounded too much like horizontal concepts, like in operators. The actual purpose of these variables is to add blank lines before and after the current line, so use the same wording as in the command line options. No functional change.
1.141	20-Oct-2021	rillig	indent: invert condition in process_newline It's hard to follow a condition that combines many negated terms with '\|\|'. Group the conditions by their origin. The condition '!opt.break_after_comma && break_comma' still sounds like a contradition, more investigations to follow. No functional change.
1.140	20-Oct-2021	rillig	indent: rename next_blank_lines to blank_lines_to_output The previous name was already an improvement over the name before that (n_real_blanklines), but didn't express the intended purpose clearly enough, so try another name. No functional change.
1.139	19-Oct-2021	rillig	indent: if a file ends with indent off, don't add space-newline
1.138	17-Oct-2021	rillig	indent: parse int command line options strictly On i386 and other platforms where LONG_MAX == INT_MAX, the test t_errors/option_tabsize_very_large failed since the behavior on integer overflow differs between ILP32 and LP64 platforms. Noticed by gson@. Avoid this unintended difference by adding reasonable limits for each of the integer options and by replacing atoi with strtol.
1.137	09-Oct-2021	rillig	indent: condense code for calculating indentations No functional change.
1.136	09-Oct-2021	rillig	indent: extract common code for advancing a single tab No functional change.
1.135	08-Oct-2021	rillig	indent: clean up comments, parentheses, debug messages, boolean operator No functional change.
1.134	08-Oct-2021	rillig	indent: rename in_or_st to init_or_struct This makes a few comments redundant. No functional change.
1.133	08-Oct-2021	rillig	indent: rename fill_buffer to inbuf_read_line No functional change.
1.132	08-Oct-2021	rillig	indent: clean up process_decl, replace unnecessary strlen No functional change.
1.131	08-Oct-2021	rillig	indent: remove unnecessary forward declarations No functional change.
1.130	08-Oct-2021	rillig	indent: reduce negations in main_loop No functional change.
1.129	08-Oct-2021	rillig	indent: fix parsing of preprocessor lines with comments and strings
1.128	08-Oct-2021	rillig	indent: run indent on indent.h The formatting looks mostly OK. Some struct members had excessively long names, leaving no space for their corresponding comments. Renamed some of them using well-known abbreviations. The formatting for debug_vis_range is messed up, no idea why. It is clearly a function declaration, not a function definition, so there is no need to place the function name in column 1. No functional change.
1.127	08-Oct-2021	rillig	indent: split process_keyword_do_else into separate functions No functional change.
1.126	08-Oct-2021	rillig	indent: rename tokens lparen and rparen to be more precise No functional change.
1.125	07-Oct-2021	rillig	indent: rename bp_save to saved_inp_s, be_save to saved_inp_e Using the same naming convention makes it easier to relate the variables. No functional change.
1.124	07-Oct-2021	rillig	indent: group variables for the input buffer The input buffer follows the same concept as the intermediate buffers for label, code, comment and token, so use the same type for it. No functional change.
1.123	07-Oct-2021	rillig	indent: move definition of bufsize from header to implementation No functional change.
1.122	07-Oct-2021	rillig	indent: rename opt.btype_2 to brace_same_line No functional change.
1.121	07-Oct-2021	rillig	indent: clean up code, remove outdated wrong comments No functional change.
1.120	07-Oct-2021	rillig	indent: use braces around multi-line statements No functional change.
1.119	07-Oct-2021	rillig	indent: let the code breathe a bit by inserting empty lines No functional change.
1.118	07-Oct-2021	rillig	indent: clean up comments No functional change.
1.117	07-Oct-2021	rillig	indent: fix wrong or outdated comments No functional change.
1.116	07-Oct-2021	rillig	indent: remove redundant comments No functional change.
1.115	07-Oct-2021	rillig	indent: reduce indentation No functional change.
1.114	07-Oct-2021	rillig	indent: remove global variable option_source It is only needed at startup, while parsing the options. The string "?" was not needed at all. No functional change.
1.113	07-Oct-2021	rillig	indent: clean up colon handling No functional change.
1.112	07-Oct-2021	rillig	indent: add high-level API for working with buffers This makes the code more boring to read, which is actually good. Less fiddling with memcpy and pointer arithmetics. Since indent is not a high-performance tool used for bulk operations on terabytes of source code, there is no need to squeeze out every possible CPU cycle. No functional change.
1.111	07-Oct-2021	rillig	indent: rename copy_id to copy_token No functional change.
1.110	07-Oct-2021	rillig	indent: raise WARNS from the default 5 up to 6
1.109	07-Oct-2021	rillig	indent: prevent division by zero
1.108	05-Oct-2021	rillig	indent: rename n_real_blanklines The word 'n' was not as helpful as possible, the word 'real' did not give any clue at all about the variable's purpose. No functional change.
1.107	05-Oct-2021	rillig	indent: fix off-by-one error for indented first line
1.106	05-Oct-2021	rillig	indent: make off-by-one error in main_prepare_parsing more visible No functional change.
1.105	05-Oct-2021	rillig	indent: make variable names more expressive The abbreviation 'dec' looked too much like 'decimal' instead of the intended 'declaration'. No functional change.
1.104	05-Oct-2021	rillig	indent: remove variable name prefix 'inout_' This makes the variable names more readable. The prefix is not actually needed to understand the code, it is rather distracting. The compiler and lint will guard against any accidental mismatch between pointer, integer and bool. No functional change.
1.103	05-Oct-2021	rillig	indent: fix Clang-Tidy warnings, clean up bakcopy The comment above and inside bakcopy had been outdated for at least the last 28 years, the backup file is named "%s.BAK", not ".B%s". Prevent buffer overflow for very long filenames (sprintf -> snprintf).
1.102	05-Oct-2021	rillig	indent: fix spelling in comments
1.101	05-Oct-2021	rillig	indent: merge duplicate code into is_hspace No functional change.
1.100	05-Oct-2021	rillig	indent: clean up code for appending to buffers Use *e++ for appending and e[-1] for testing the previously appended character, like in other places in the code. No functional change.
1.99	05-Oct-2021	rillig	indent: merge duplicate code for reading from input buffer No functional change.
1.98	03-Oct-2021	rillig	indent: rename functions There was no good reason for using the different verbs 'scan' and 'set' for two functions that essentially do the same. No functional change.
1.97	03-Oct-2021	rillig	indent: fix content of profile_name Previously, profile_name included the leading "-P", which was confusing.
1.96	30-Sep-2021	rillig	indent: remove space between ')' and '(' in declarations
1.95	30-Sep-2021	rillig	indent: untangle want_blank_before_lparen No functional change.
1.94	30-Sep-2021	rillig	indent: extract want_blank_before_lparen No functional change.
1.93	30-Sep-2021	rillig	indent: add space between ',' and '[' in C99 initializations
1.92	27-Sep-2021	rillig	indent: let indent format the comments after previous refactoring Before this refactoring, I had skipped this section of the code from formatting since the 'default:' branch was enclosed in a block of its own, and that block would have been indented one more level to the right. Extracting that code into a separate function got rid of the extra braces. No functional change.
1.91	27-Sep-2021	rillig	indent: split search_brace into smaller functions No functional change.
1.90	27-Sep-2021	rillig	indent: use binary instead of linear search when adding types No functional change.
1.89	27-Sep-2021	rillig	indent: rename rwcode to keyword_kind, various cleanup No idea what the 'rw' in 'rwcode' meant, it had been imported that way 28 years ago. Since rwcode specifies the kind of a keyword, the prefix 'kw_' makes sense. No functional change.
1.88	26-Sep-2021	rillig	indent: unexport global variables The variable match_state was write-only and was thus removed. No functional change.
1.87	26-Sep-2021	rillig	indent: negate and rename option.leave_comma The old name did not mirror the description in the manual page, and it was the only option that is negated. Inverting it allows the options table to be compressed.
1.86	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.85	26-Sep-2021	rillig	indent: fix missing space between comma and ellipsis According to lint's C grammar, in standard C an ellipsis only occurs after a comma. There are GCC extensions that allow an ellipsis as the only function parameter, as well as in 'case a ... b', but these are rare.
1.84	25-Sep-2021	rillig	indent: misc cleanup No functional change.
1.83	25-Sep-2021	rillig	indent: convert found_err to bool That variable had slipped through the migration since it consequently used int for the declaration, the definition and all assignments. No functional change.
1.82	25-Sep-2021	rillig	indent: use strlen instead of own implementation The two loops looks similar but differ in a crucial detail that makes up for a '+ 1'. No functional change.
1.81	25-Sep-2021	rillig	indent: merge duplicate code for token buffers No functional change.
1.80	25-Sep-2021	rillig	indent: clean up argument handling No functional change.
1.79	25-Sep-2021	rillig	indent: un-abbreviate a few parser_state members, clean up comments No functional change.
1.78	25-Sep-2021	rillig	indent: remove dead code for printing comments after empty lines This code has been commented out for at least 29 years. No functional change.
1.77	25-Sep-2021	rillig	indent: reduce code and data size for lexing of numbers Instead of having a table of strings (121 pointers + 121 data relocations), reduce that table to the actual character data and use a secondary table for looking up the correct row in the main table. No functional change.
1.76	25-Sep-2021	rillig	indent: rename option variable to be more expressive No functional change.
1.75	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.74	25-Sep-2021	rillig	indent: convert parser_state from ibool to bool indent.c:400:5: error: suggest parentheses around assignment used as truth value io.c:271:32: error: ‘~’ on a boolean expression No functional change intended.
1.73	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.72	25-Sep-2021	rillig	indent: merge duplicate code for initializing buffers No functional change.
1.71	25-Sep-2021	rillig	indent: clean up initialization of options The default values in 'struct pro' were redundant but all consistent, even with the commented defaults in main_parse_command_line. No functional change.
1.70	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.69	25-Sep-2021	rillig	indent: move statistical values into a separate struct No functional change.
1.68	25-Sep-2021	rillig	indent: add nonnull memory allocation functions The only functional change is a single error message.
1.67	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.66	25-Sep-2021	rillig	indent: inline macro 'token' No functional change.
1.65	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.64	25-Sep-2021	rillig	indent: rename variables of type token_type The previous variable name 'code' conflicts with the buffer of the same name. No functional change.
1.63	24-Sep-2021	rillig	indent: group global variables for label buffer into struct No functional change.
1.62	24-Sep-2021	rillig	indent: group global variables for the comment buffer No functional change.
1.61	25-Aug-2021	rillig	indent: fix lint warnings about type conversions on ilp32 No functional change.
1.60	26-Mar-2021	rillig	indent: fix Clang build everywhere but on amd64 No idea why Clang didn't complain about this on amd64, only on all other platforms.
1.59	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.58	13-Mar-2021	rillig	indent: add debug logging for switching the input buffer No functional change outside debug mode.
1.57	13-Mar-2021	rillig	indent: distinguish between 'column' and 'indentation' column == 1 + indentation. In addition, indentation is a relative distance while column is an absolute position. Therefore, don't confuse these two concepts, to prevent off-by-one errors. No functional change.
1.56	13-Mar-2021	rillig	indent: rename pr_comment to process_comment, clean up documentation No functional change.
1.55	13-Mar-2021	rillig	indent: fix handling of '/' in string literal in preprocessing line Previously, the '/' in the string literal had been interpreted as the beginning of a comment, which was wrong. Because of that, the variable declaration in the following line was still interpreted as part of the comment. The comment even continued until the end of the file. Due to indent's forgiving nature, it neither complained nor even mentioned that anything had gone wrong. The decision of rather producing wrong output than failing early is a dangerous one. At least, there should have been an error message that at the end of the file, the parser was still in a a comment, expecting the closing '*/'.
1.54	13-Mar-2021	rillig	indent: split 'main_loop' into several functions No functional change.
1.53	13-Mar-2021	rillig	indent: split 'main' into manageable parts Since several years (maybe even decades) compilers know how to inline static functions that are only used once. Therefore there is no need to have overly long functions anymore, especially not 'main', which is only called a single time and thus does not add any noticeable performance degradation. No functional change.
1.52	13-Mar-2021	rillig	indent: remove redundant parentheses No functional change.
1.51	13-Mar-2021	rillig	indent: fix confusing variable names The word 'col' should only be used for the 1-based column number. This name is completely inappropriate for a line length since that provokes off-by-one errors. The name 'cols' would be acceptable although confusing since it sounds so similar to 'col'. Therefore, rename variables that are related to the maximum line length to 'line_length' since that makes for obvious code and nicely relates to the description of the option in the manual page. No functional change.
1.50	13-Mar-2021	rillig	indent: inline calls to count_spaces and count_spaces_until These two functions operated on column numbers instead of indentation, which required adjustments of '+ 1' and '- 1'. Their names were completely wrong since these functions did not count anything, instead they computed the column. No functional change.
1.49	13-Mar-2021	rillig	indent: replace compute_code_column with compute_code_indent The goal is to only ever be concerned about the _indentation_ of a token, never the _column_ it appears in. Having only one of these avoids off-by-one errors. No functional change.
1.48	13-Mar-2021	rillig	indent: add debug logging for actually writing to the output file Together with the results of the tokenizer and the 4 buffers for token, label, code and comment, the debug log now provides a good high-level view on how the indentation happens and where to look for the many remaining bugs.
1.47	13-Mar-2021	rillig	indent: replace pad_output with output_indent Calculating the indentation is simpler than calculating the column, since that saves the constant addition and subtraction of the 1. No functional change.
1.46	12-Mar-2021	rillig	indent: replace 'target' with 'indent' in function names The word 'target' was not as specific as possible. No functional change.
1.45	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.44	12-Mar-2021	rillig	indent: manually fix indentation No functional change.
1.43	11-Mar-2021	rillig	indent: reduce indentation of check_size functions No functional change.
1.42	11-Mar-2021	rillig	indent: remove redundant cast after allocation functions No functional change.
1.41	09-Mar-2021	rillig	indent: extract search_brace from main No functional change.
1.40	09-Mar-2021	rillig	indent: extract capsicum code out of the main function No functional change.
1.39	09-Mar-2021	rillig	indent: rename a few more token types The previous names were either too short or ambiguous. No functional change.
1.38	09-Mar-2021	rillig	indent: make token names more precise The previous 'casestmt' was wrong since a case label is not a statement at all. The previous 'swstmt' was overly short, and wrong as well, since it represents only the 'switch (expr)' part, which is not a complete switch statement. Same for 'ifstmt', 'whilestmt', 'forstmt'. The previous word 'head' was not precise enough since it didn't specify exactly where the head ends and the body starts. Especially for handling the dangling else, this distinction is important. No functional change.
1.37	09-Mar-2021	rillig	indent: rename a few tokens to be more obvious For casual readers it is not obvious whether the 'sp' meant 'special' or 'space' or something entirely different.
1.36	09-Mar-2021	rillig	indent: manually indent comments It's strange that indent's own code is not formatted by indent itself, which would be a good demonstration of its capabilities. In its current state, I don't trust indent to get even the tokenization correct, therefore the only safe way is to format the code manually.
1.35	08-Mar-2021	rillig	indent: inline macro for backslash No functional change.
1.34	08-Mar-2021	rillig	indent: convert big macros to functions Each of these buffers is only modified in a single file. This makes it unnecessary to declare the macros in the global header.
1.33	08-Mar-2021	rillig	indent: fix printing of uninitialized 'token' in debug output
1.32	07-Mar-2021	rillig	indent: sprinkle a few const No functional change.
1.31	07-Mar-2021	rillig	indent: use named constants for the different types of keywords This reduces the magic numbers in the code. Most of these had their designated constant name written in a nearby comment anyway. The one instance where arithmetic was performed on this new enum type (in indent.c) was a bit tricky to understand. The combination rw_continue_or_inline_or_restrict looks strange, the 'continue' should intuitively belong to the other control flow keywords in rw_break_or_goto_or_return. No functional change.
1.30	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.29	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.28	06-Mar-2021	rillig	indent: fix space-tab alignment in indent's own code These parts are not fixed automatically by indent since they are in box comments. No functional change.
1.27	23-Apr-2020	joerg	Avoid common symbol declarations
1.26	19-Oct-2019	christos	use stdarg, annotate function as __printflike and fix broken formats.
1.25	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.24	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.23	05-Sep-2016	sevan	branches: 1.23.14; Drop main() prototype.
1.22	25-Feb-2016	ginsbach	Fix obvious contraction spelling mistakes by adding missing apostrophes.
1.21	22-Feb-2016	ginsbach	Use warnx(3).
1.20	22-Feb-2016	ginsbach	Use errx(3).
1.19	04-Sep-2014	mrg	port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.18	12-Apr-2009	lukem	branches: 1.18.24; Fix WARNS=4 issues (-Wshadow -Wcast-qual -Wsign-compare)
1.17	21-Jul-2008	lukem	branches: 1.17.6; Remove the \n and tabs from the __COPYRIGHT() strings. Tweak to use a consistent format.
1.16	30-Oct-2004	dsl	branches: 1.16.28; Add (unsigned char) cast to ctype functions
1.15	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.14	19-Jun-2003	christos	PR/21645: Mishka: Localized comments don't work with indent.
1.13	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.12	20-Aug-2001	wiz	precede, not preceed.
1.11	16-Jun-2001	kleink	Handle a labeled statement at the beginning of a function correctly; from Nagae Hidetake <nagae@tk.airnet.ne.jp> in PR bin/12781.
1.10	19-Dec-1998	christos	char -> unsigned char, braces for gcc-2.8.1
1.9	08-Oct-1998	wsanchez	Get rid of multiply defined common symbols
1.8	06-Sep-1998	mellon	Support indenting standard input. When indenting standard input, write output to standard output.
1.7	25-Aug-1998	ross	Add { and } to shut up egcs. Reformat the more questionable code.
1.6	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.5	18-Oct-1997	mrg	merge lite-2.
1.4	09-Jan-1997	tls	RCS ID police
1.3	07-May-1996	jtc	Include appropriate header files to bring prototypes into scope. Removed explicit errno declarations.
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	08-Jun-1993	mrg	4.4BSD-Lite2
1.16.28.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.17.6.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.18.24.1	21-Sep-2014	snj	Pull up following revision(s) (requested by mrg in ticket #110): usr.bin/indent/io.c: revision 1.15 usr.bin/indent/indent_globs.h: revision 1.10 usr.bin/indent/args.c: revision 1.11 usr.bin/indent/indent.1: revision 1.23 usr.bin/indent/indent.c: revision 1.19 port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.23.14.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.23.14.1	10-Jun-2019	christos	Sync with HEAD
1.390.2.1	02-Aug-2025	perseant	Sync with HEAD
1.211	07-Jan-2025	rillig	indent: condense and simplify parsing code
1.210	04-Jan-2025	rillig	indent: make debug log more uniform
1.209	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.208	03-Jan-2025	rillig	indent: fix line breaks in else-if sequences The flag ps.want_newline did not adequately model the conditions under which a line break should be inserted, thus the redesign. A welcome side effect is that in statements like 'if (cond);', the semicolon is now placed on a separate line, thus becoming more visible.
1.207	03-Dec-2023	rillig	branches: 1.207.2; indent: inline input-related macros No binary change.
1.206	03-Dec-2023	rillig	indent: group input-related variables into a struct No functional change.
1.205	03-Dec-2023	rillig	indent: use line number of the token start in diagnostics Previously, the line number of the end of the token was used, which was confusing in debug mode.
1.204	26-Jun-2023	rillig	indent: implement 'blank line above first statement in function body'
1.203	23-Jun-2023	rillig	indent: properly store parser state in debug mode The stacks in the parser state are allocated now and need to be copied individually. The test whether two paren stacks are equal was broken since 2023-06-14 14:11:28.
1.202	16-Jun-2023	rillig	indent: merge lexer symbols for type in/outside parentheses
1.201	16-Jun-2023	rillig	indent: fix indentation and linebreaks in typedef declarations
1.200	16-Jun-2023	rillig	indent: don't force a blank line between '}' and preprocessing line
1.199	16-Jun-2023	rillig	indent: rename a field of the parser state The previous name 'comment_in_first_line' was misleading, as it could mean that there was a comment in the first line of the file. No functional change.
1.198	15-Jun-2023	rillig	indent: rename state variable to be more accurate No binary change.
1.197	14-Jun-2023	rillig	indent: clean up the code, add a few tests
1.196	14-Jun-2023	rillig	indent: allow more than 128 brace levels
1.195	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.194	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.193	14-Jun-2023	rillig	indent: clean up debugging code
1.192	14-Jun-2023	rillig	indent: clean up handling of comments One less moving part in the parser state. No functional change.
1.191	14-Jun-2023	rillig	indent: remove another flag from parser state When processing a comment, the flag ps.next_col_1 was not used for the next token, but for a line within a comment. As its scope was limited to a single comment, there is no need to store it any longer than that No functional change.
1.190	14-Jun-2023	rillig	indent: remove a redundant flag from the parser state No functional change.
1.189	14-Jun-2023	rillig	indent: merge parser symbols for stmt and stmt_list They were handled in exactly the same way.
1.188	10-Jun-2023	rillig	indent: rename misleading variable The name started with 'line_start', but the value is not always the value from the beginning of the line. No functional change.
1.187	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.186	10-Jun-2023	rillig	indent: in debug mode, null-terminate buffers
1.185	10-Jun-2023	rillig	indent: clean up function and variable names
1.184	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.183	09-Jun-2023	rillig	indent: trim trailing blank lines
1.182	09-Jun-2023	rillig	indent: group lexer symbols by topic, sort processing functions No functional change.
1.181	09-Jun-2023	rillig	indent: don't treat function call expressions as cast expressions
1.180	09-Jun-2023	rillig	indent: when an indentation is ambiguous, indent one level further The '-eei' mode now applies whenever the indentation from a multi-line expression could be confused with a following statement.
1.179	08-Jun-2023	rillig	indent: remove fragile heuristic for detecting cast expressions The assumption that in an expression of the form '(a * anything)', the '*' marks a pointer type was too simple-minded. For now, fix the obvious cases and leave the others for later. If needed, they can be worked around using the '-T' option.
1.178	08-Jun-2023	rillig	indent: clean up and condense code No functional change.
1.177	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.176	06-Jun-2023	rillig	indent: compute indentation of 'case' labels on-demand One less moving part to keep track of. No functional change.
1.175	05-Jun-2023	rillig	indent: sync debug output with parser state
1.174	05-Jun-2023	rillig	indent: format own source code
1.173	05-Jun-2023	rillig	indent: do not report broken lines, report configuration on stderr
1.172	05-Jun-2023	rillig	indent: rename variables, clean up comments No binary change.
1.171	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.170	04-Jun-2023	rillig	indent: track the kind of '{' on the parser stack
1.169	04-Jun-2023	rillig	indent: fix debug output of the parser symbol stack Even though the stack always contains a stmt_list as first element, print it nevertheless to avoid confusion about starting at index 1, and to provide the full picture.
1.168	04-Jun-2023	rillig	indent: rename struct field, for better symmetry No binary change outside debug mode.
1.167	04-Jun-2023	rillig	lint: use separate lexer symbols for 'case' and 'default' It's not strictly necessary since these tokens behave in the same way, still, the code is more straight-forward when there are separate tokens.
1.166	04-Jun-2023	rillig	indent: classify 'inline' as a modifier rather than a word
1.165	04-Jun-2023	rillig	indent: use separate lexer symbols for the different kinds of ':'
1.164	04-Jun-2023	rillig	indent: handle the indentation of 'case' in a simpler way
1.163	04-Jun-2023	rillig	indent: separate code for handling parentheses and brackets Handling parentheses is more complicated than for brackets.
1.162	02-Jun-2023	rillig	indent: improve heuristics of classifying '*' as pointer or operator
1.161	02-Jun-2023	rillig	indent: clean up Only print the 'token' buffer in debug mode if it is interesting, group the blocks in handling of '(' tokens by topic, remove obsolete comment from test.
1.160	02-Jun-2023	rillig	indent: fix formatting of declarations with preprocessing lines
1.159	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.158	23-May-2023	rillig	indent: fix spacing in declarations in for loops
1.157	22-May-2023	rillig	indent: implement suppressing optional blank lines
1.156	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.155	20-May-2023	rillig	indent: implement blank line above block comment
1.154	20-May-2023	rillig	indent: implement blank line after function body
1.153	20-May-2023	rillig	indent: implement blank lines around conditional compilation
1.152	20-May-2023	rillig	indent: add debug logging for brace indentation No functional change outside debug mode, as the initialization of di_stack[0] was redundant.
1.151	20-May-2023	rillig	indent: separate detection of function definitions from lexing '*' No functional change.
1.150	18-May-2023	rillig	indent: document the funcname token
1.149	18-May-2023	rillig	indent: rename a few functions No functional change.
1.148	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.147	18-May-2023	rillig	indent: remove unnecessary variable size optimization Due to the enum that follows in the struct, the short variable was padded to 4 bytes anyway. No functional change.
1.146	16-May-2023	rillig	indent: directly access the input buffer No functional change.
1.145	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.144	16-May-2023	rillig	indent: fix handling of INDENT OFF/ON comments Previously, the 'INDENT OFF' comments were interpreted when the newline token from the line above the comment was processed, which was earlier than could be reasonably expected. The 'INDENT ON' comments were interpreted equally early, which led to the situation that the 'INDENT OFF' comments were preserved literally but the 'INDENT ON' comments weren't.
1.143	16-May-2023	rillig	indent: move parsing of 'INDENT OFF/ON' comments to the lexer No functional change.
1.142	15-May-2023	rillig	indent: clean up detection of whether parentheses form a cast No functional change.
1.141	15-May-2023	rillig	indent: fix duplicate space between comment and binary operator
1.140	15-May-2023	rillig	indent: format its own code, extend some comments With manual corrections, as there are still some bugs left. No functional change.
1.139	15-May-2023	rillig	indent: fix indentation of statements after controlling expression
1.138	15-May-2023	rillig	indent: fix indentation of expressions in -nlp -eei mode
1.137	15-May-2023	rillig	indent: fix indentation of multi-line '?:' expressions in functions
1.136	15-May-2023	rillig	indent: group parser state by topic No functional change.
1.135	15-May-2023	rillig	indent: clean up memory allocation No functional change.
1.134	15-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.133	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.132	14-May-2023	rillig	indent: only null-terminate the buffers if necessary The only case where a buffer is used as a C-style string is when looking up a keyword. No functional change.
1.131	14-May-2023	rillig	indent: reduce code for scanning tokens The input line is guaranteed to end with '\n', so there's no need to carry another pointer around. No functional change.
1.130	14-May-2023	rillig	indent: remove foreign RCS IDs
1.129	14-May-2023	rillig	indent: miscellaneous cleanups
1.128	13-May-2023	rillig	indent: implement 'blank after declarations'
1.127	13-May-2023	rillig	indent: use enum instead of magic numbers for tracking declarations No functional change.
1.126	13-May-2023	rillig	indent: improve names of option variables No functional change.
1.125	13-May-2023	rillig	indent: rename struct fields for buffers No binary change except for assertion line numbers.
1.124	13-May-2023	rillig	indent: clean up a condition, add comments No functional change.
1.123	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.122	12-May-2023	rillig	indent: rename placeholder symbol for parser stack No functional change outside debug mode.
1.121	12-May-2023	rillig	indent: remove statistics The numbers from the statistics were wrong.
1.120	12-May-2023	rillig	indent: condense code for handling spaced expressions No functional change outside debug mode.
1.119	11-May-2023	rillig	indent: don't touch comments in preprocessing lines The indentation of multi-line comments was wrong, and the code for handling them was too complicated.
1.118	11-May-2023	rillig	indent: remove unused code
1.117	11-May-2023	rillig	indent: remove broken code for handling blank lines This fixes several bugs where blank lines were erroneously added or removed, treating these old bugs for new bugs in different places. These new bugs are expected to be easier to fix, as the old bugs will not interfere anymore.
1.116	11-May-2023	rillig	indent: add debug output for tracking comments and braces
1.115	11-May-2023	rillig	indent: move parser state variables to the parser_state struct Include the variables in the debug output.
1.114	11-May-2023	rillig	indent: move force_nl into the parser state This way, it is included in the debug output. No functional change.
1.113	11-May-2023	rillig	indent: remove buggy code for swapping tokens It is not the job of an indenter to swap tokens, even if it's only about placing comments elsewhere. The code that swapped the tokens was complicated, buggy and impossible to understand. In -br (brace right) mode, indent no longer moves a '{' from the beginning of a line to the end of the previous line, as that was handled by the token swapping code as well. This change is unintended, but it will be easier to re-add that now that the code is simpler.
1.112	23-Apr-2022	rillig	indent: group global variables related to output control No functional change.
1.111	13-Feb-2022	rillig	indent: rename parser_state.p_l_follow and paren_level The previous variable names were misleading. Paren_level is not the current level of parentheses but the one from the beginning of the current output line. For better accuracy, rename it to line_start_paren_level. P_l_follow is not the level of parentheses that will be active at some point in the future, as the previous name suggested. Instead, it is the level of parentheses right now. For better accuracy, rename it to nparen. This nicely matches its main usage, which is as index to the parser_state.paren array. No binary change.
1.110	13-Feb-2022	rillig	indent: replace bitmasking code with struct The struct directly represents the properties of a pair of parentheses, without forcing the human reader to decode any bitset. This makes it easier to find the remaining bugs in the heuristic for determining the kind of parentheses. No functional change outside debug mode.
1.109	13-Feb-2022	rillig	indent: change parser_state.cast_mask to 0-based indexing Having 1-based indexing was completely unexpected, and it didn't match the 0-based indexing of parser_state.paren_indents. No functional change.
1.108	12-Feb-2022	rillig	indent: fix indentation of enum constants in typedef (since 2019-04-04) The solution is not elegant since it adds a small state machine inside the parser state, but at least these states only depend on the sequence of token types and not on any other part of the parser state. Reported in PR#55453.
1.107	28-Nov-2021	rillig	indent: treat L"string" as a single token There is never whitespace between the 'L' and the string literal or the character constant. There might be a backslash-newline between them, but that case was not handled before either. No functional change.
1.106	28-Nov-2021	rillig	indent: clean up and document input handling The transformation of moving comments from after an 'if (expr)' after the following brace has a large implementation cost (about 300 lines of code) and makes input handling quite complicated. Document the overall idea to save future readers some time. No functional change.
1.105	27-Nov-2021	rillig	indent: accept a few formatting suggestions from indent The remaining issues are still that the conditions look ambiguous even with -eei, and that __attribute__ is broken into a separate line. No functional change.
1.104	27-Nov-2021	rillig	indent: rename dump functions to output No functional change.
1.103	26-Nov-2021	rillig	indent: add buf_add_range for adding characters to a buffer No functional change.
1.102	25-Nov-2021	rillig	indent: rename ps.in_function_parameters to match reality This flag is only set while parsing the parameters of a function definition, but not for a function declaration. See buffer_add in the test fmt_decl. No functional change.
1.101	25-Nov-2021	rillig	indent: fix comment for ps.in_decl In C, there are no declaration statements. There are declarations and statements, but no combination thereof.
1.100	25-Nov-2021	rillig	indent: rename ps.in_stmt to in_stmt_or_decl The previous name didn't match reality. No functional change.
1.99	25-Nov-2021	rillig	indent: rename ps.ind_stmt to in_stmt_cont This makes a comment redundant. No functional change.
1.98	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.97	19-Nov-2021	rillig	indent: replace ps.procname with ps.is_function_definition Omly the first character of ps.procname was ever read, and it was only compared to '\0'. Using a bool for this means simpler code, less memory and fewer wasted CPU cycles due to the removed strncpy. No functional change.
1.96	19-Nov-2021	rillig	indent: unexport inbuf No functional change.
1.95	19-Nov-2021	rillig	indent: use character input API from pr_comment.c No functional change.
1.94	19-Nov-2021	rillig	indent: remove all references to inbuf from indent.c No functional change.
1.93	19-Nov-2021	rillig	indent: move character input handling from indent.c to io.c No functional change.
1.92	19-Nov-2021	rillig	indent: move character input from indent.c to io.c No functional change.
1.91	19-Nov-2021	rillig	indent: use character input API from the tokenizer No functional change.
1.90	19-Nov-2021	rillig	indent: move character input handling from lexi.c to io.c No functional change.
1.89	19-Nov-2021	rillig	indent: replace direct access to the input buffer This is a preparation for abstracting away all the low-level details of handling the input. The goal is to fix the current bugs regarding line number counting, out of bounds memory access, and generally unreadable code. No functional change.
1.88	19-Nov-2021	rillig	indent: rename input buffer variables From reading the names 'save_com' and 'sc_end', it was not obvious enough that these two variables are the limits of the same buffer, the names were just too unrelated. No functional change.
1.87	19-Nov-2021	rillig	indent: group variables for input handling No functional change.
1.86	07-Nov-2021	rillig	indent: various cleanups Make several comments more precise. Rename process_end_of_file to process_eof to match the token name. Change the order of assignments in analyze_comment to keep the com_ind computations closer together. In copy_comment_wrap, use pointer difference instead of pointer addition to stay away from undefined behavior. No functional change.
1.85	07-Nov-2021	rillig	indent: rename ps.decl_nest to decl_level This better matches the comment. No functional change.
1.84	07-Nov-2021	rillig	indent: reduce negations in process_else, clean up comments No functional change.
1.83	07-Nov-2021	rillig	indent: document the comment buffer more accurately
1.82	07-Nov-2021	rillig	indent: split copy_comment into wrapping and non-wrapping These two cases are processed in an almost entirely different way. In particular, copy_comment_nowrap should copy the comment verbatim, which is not obvious from the current code, due to the many conditions and the complex control flow. No functional change.
1.81	07-Nov-2021	rillig	indent: rename type_at_paren_level_0 to type_outside_parentheses For symmetry with type_in_parentheses. No functional change.
1.80	07-Nov-2021	rillig	indent: distinguish between typename in parentheses and other words This gets rid of two members of parser_state. No functional change for well-formed programs. The sequence of '++int' or '--size_t' may be formatted differently than before, but no program is expected to contain that sequence. Rename lsym_ident to lsym_word since 'ident' was too specific. This token type is used for constants and string literals as well. Strictly speaking, a string literal is not a word, but at least it's better than before.
1.79	07-Nov-2021	rillig	indent: rename 'inbuf' functions to 'inp' The variable 'inp' used to be named 'inbuf'. Make the function names correspond to the variable name again. No functional change.
1.78	05-Nov-2021	rillig	indent: the '+ 1' in dump_line_code is not an off-by-one error
1.77	05-Nov-2021	rillig	indent: rename ps.curr_newline to next_col_1 For symmetry with ps.curr_col_1. No functional change.
1.76	03-Nov-2021	rillig	indent: inline indentation_after, shorten function name to ind_add There were only few calls to indentation_after, so inlining it spares the need to look at yet another function definition. Another effect is that code.s and code.e appear in the code as a pair now, instead of a single code.s, making the scope of the function call obvious. In ind_add, there is no need to check for '\0' anymore since none of the buffers can ever contain a null character, these are filtered out by inbuf_read_line. No functional change.
1.75	01-Nov-2021	rillig	indent: fix missing blank after 'return' (since 2021-10-31) In indent.c 1.200 from 2021-10-31, the subtypes of identifier tokens were removed since they were redundant. An unintended side effect was that a parenthesized expression after 'return' was no longer separated by a blank. Before that change, 'return' was tokenized as an lsym_ident with subtype kw_other, and want_space_before_lparen handled this case in the last line. After the change, 'return' was treated as an ordinary identifier, and unless the option '-pcs' (blank after function call) was given, the blank was removed. The other keywords that had kw_other are not affected since they do not expect a '(' afterwards. These keywords are 'break', 'continue', 'goto', 'inline' and 'restrict'. Curiously, there was not a single test case that covered 'return(expr)'. While here, remove the trailing ',' from the enum lexer_symbol, which is not allowed in standard C, it is a GNU extension. Lint doesn't complain about this since the default LINTFLAGS include '-g' for GCC mode.
1.74	31-Oct-2021	rillig	indent: clean up Initialize buffers in reading order, make comments more expressive, rename add_typename to register_typename, remove unused macro. No functional change.
1.73	31-Oct-2021	rillig	indent: replace kw_tag with lsym_tag This leaves only one special type of token, which is lsym_ident, which in some cases represents a type name and in other cases an identifier, constant or string literal. No functional change.
1.72	31-Oct-2021	rillig	indent: replace simple cases of keyword_kind with lexer_symbol The remaining keyword kinds 'tag' and 'type' require a bit more thought, so do them in a separate step. No functional change.
1.71	31-Oct-2021	rillig	indent: rename lsym_type to better reflect reality Type names that occur in parentheses are parsed as lsym_ident having the subtype kw_type instead. No functional change.
1.70	31-Oct-2021	rillig	indent: add separate lexer symbol for offsetof No functional change.
1.69	31-Oct-2021	rillig	indent: add separate lexer symbol for sizeof The plan is to get rid of the type keyword_kind, which largely overlaps with lexer_symbol. No functional change.
1.68	31-Oct-2021	rillig	indent: clean up definition of keywords Rename kw_struct_or_union_or_enum to the shorter kw_tag. Merge kw_jump with kw_inline_or_restrict since they are handled in the same way. No functional change.
1.67	30-Oct-2021	rillig	indent: move debugging functions to a separate section
1.66	30-Oct-2021	rillig	indent: rename prev_newline and prev_col_1 to curr These two flags describe the token that is currently processed. In process_binary_op, curr_newline can never be true since newline is not a binary operator, so remove that condition. No functional change.
1.65	30-Oct-2021	rillig	indent: clean up lexical analyzer Use traditional type for small unsigned numbers instead of uint8_t; the required header was not included. Remove assertion for debug mode; lint takes care of ensuring that the enum constants match the length of the names array. Constify a name array. Move the comparison function for bsearch closer to its caller. No functional change.
1.64	30-Oct-2021	rillig	indent: inline macro label_offset No functional change.
1.63	29-Oct-2021	rillig	indent: fix missing blank before binary operator
1.62	29-Oct-2021	rillig	indent: merge isblank and is_hspace into ch_isblank No functional change.
1.61	29-Oct-2021	rillig	indent: reorder global variables to be more intuitive The buffer 'inp' comes first. From there, a single token is read into the buffer 'token'. From there, it usually ends up in 'code'. The buffer 'token' does not belong to the group of the other 3 buffers, which together make up a line of formatted output. No functional change.
1.60	29-Oct-2021	rillig	indent: use prev/curr/next to refer to the current token The word 'last' just didn't match with 'next'. No functional change.
1.59	29-Oct-2021	rillig	indent: group members of parser_state by topic No functional change.
1.58	29-Oct-2021	rillig	indent: rename ps.dumped_decl_indent and indent_declaration The word 'dump' in 'ps.dumped_decl_indent' was too close to dump_line, which led to confusion since the variable controls whether the indentation has been added to the code buffer, which happens way before actually dumping the current line to the output file. The function name 'indent_declaration' was too unspecific, it did not reveal where the indentation of the declaration actually happened. No functional change.
1.57	29-Oct-2021	rillig	indent: keep p_l_follow nonnegative, use consistent comparison No functional change.
1.56	29-Oct-2021	rillig	indent: spell 'parentheses' properly in messages and comments
1.55	28-Oct-2021	rillig	indent: clean up indentation, comments, reduce No functional change.
1.54	28-Oct-2021	rillig	indent: reduce negations in search_stmt_lookahead No functional change.
1.53	28-Oct-2021	rillig	indent: clean up comments and function names Having accurate names for the lexer symbols and the parser symbols makes most of the comments redundant. Remove these. Rename process_decl to process_type, to match the name of the corresponding lexer symbol. In this phase, it's just a single type token, not a whole declaration. No functional change.
1.52	26-Oct-2021	rillig	indent: make ps.keyword easier to understand Previously, ps.keyword did not have any documentation and was not straight-forward. In some cases it was reset to kw_0, in others it was set to an interesting value. The idea behind it was to remember the kind of word of the previous token, to decide whether to have a space between sizeof or offsetof and a following '('. No functional change.
1.51	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.50	25-Oct-2021	rillig	indent: rename search_brace to search_stmt No functional change.
1.49	25-Oct-2021	rillig	indent: split type token_type into 3 separate types Previously, token_type was used for 3 different purposes: 1. symbol types from the lexer 2. symbol types on the parser stack 3. kind of control statement for 'if (expr)' and similar statements Splitting the 41 constants into separate types makes it immediately clear that the parser stack never handles comments, preprocessing lines, newlines, form feeds, the inner structure of expressions. Previously, the constant switch_expr was especially confusing since it was used for 3 different purposes: when returned from lexi, it represented the keyword 'switch', in the parser stack it represented 'switch (expr)', and it was used for a statement head as well. The only overlap between the lexer symbols and the parser symbols are '{' and '}', and the keywords 'do' and 'else'. To increase confusion, the constants of the previous token_type were in apparently random order and before 2021, they had cryptic, highly abbreviated names. No functional change.
1.48	24-Oct-2021	rillig	indent: rename form_feed to tt_lex_form_feed No functional change.
1.47	24-Oct-2021	rillig	indent: split kw_for_or_if_or_while into separate constants No functional change.
1.46	24-Oct-2021	rillig	indent: split kw_do_or_else into separate constants It was unnecessarily confusing to have the token types keyword_do_else, keyword_do and keyword_else at the same time, without any hint in what they differed. Some of the token types seem to be used by the lexer while others are used in the parse stack. Maybe all token types can be partitioned into these groups, which would suggest to use two different types for them. And if not, it's still clearer to have this distinction in the names of the constants. No functional change.
1.45	24-Oct-2021	rillig	indent: rename nitems to array_length
1.44	24-Oct-2021	rillig	indent: replace global variable use_ff with function parameter
1.43	20-Oct-2021	rillig	indent: add reminder to make the code understandable for humans
1.42	20-Oct-2021	rillig	indent: rename ps.last_u_d to match its comment No functional change.
1.41	20-Oct-2021	rillig	indent: rename parser stack variables No functional change.
1.40	20-Oct-2021	rillig	indent: rename blankline_requested variables The words 'prefix' and 'postfix' sounded too much like horizontal concepts, like in operators. The actual purpose of these variables is to add blank lines before and after the current line, so use the same wording as in the command line options. No functional change.
1.39	20-Oct-2021	rillig	indent: rename next_blank_lines to blank_lines_to_output The previous name was already an improvement over the name before that (n_real_blanklines), but didn't express the intended purpose clearly enough, so try another name. No functional change.
1.38	09-Oct-2021	rillig	indent: extract common code for advancing a single tab No functional change.
1.37	08-Oct-2021	rillig	indent: rename in_or_st to init_or_struct This makes a few comments redundant. No functional change.
1.36	08-Oct-2021	rillig	indent: convert ps.box_com to local variable This variable is only used in a single function, and that function does not call any other function that could replace the parser state or install a temporary parser state. No functional change.
1.35	08-Oct-2021	rillig	indent: rename fill_buffer to inbuf_read_line No functional change.
1.34	08-Oct-2021	rillig	indent: run indent on indent.h The formatting looks mostly OK. Some struct members had excessively long names, leaving no space for their corresponding comments. Renamed some of them using well-known abbreviations. The formatting for debug_vis_range is messed up, no idea why. It is clearly a function declaration, not a function definition, so there is no need to place the function name in column 1. No functional change.
1.33	08-Oct-2021	rillig	indent: replace column calculations with indent, part 4/4
1.32	08-Oct-2021	rillig	indent: rename tokens lparen and rparen to be more precise No functional change.
1.31	08-Oct-2021	rillig	indent: merge headers into a single file No functional change.
1.30	07-Oct-2021	rillig	indent: remove global variable option_source It is only needed at startup, while parsing the options. The string "?" was not needed at all. No functional change.
1.29	05-Oct-2021	rillig	indent: merge duplicate code into is_hspace No functional change.
1.28	05-Oct-2021	rillig	indent: merge duplicate code for reading from input buffer No functional change.
1.27	03-Oct-2021	rillig	indent: rename functions There was no good reason for using the different verbs 'scan' and 'set' for two functions that essentially do the same. No functional change.
1.26	27-Sep-2021	rillig	indent: use binary instead of linear search when adding types No functional change.
1.25	25-Sep-2021	rillig	indent: merge duplicate code for token buffers No functional change.
1.24	25-Sep-2021	rillig	indent: clean up argument handling No functional change.
1.23	25-Sep-2021	rillig	indent: reduce code and data size for lexing of numbers Instead of having a table of strings (121 pointers + 121 data relocations), reduce that table to the actual character data and use a secondary table for looking up the correct row in the main table. No functional change.
1.22	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.21	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.20	25-Sep-2021	rillig	indent: use standard definition for bool, true, false
1.19	25-Sep-2021	rillig	indent: clean up initialization of options The default values in 'struct pro' were redundant but all consistent, even with the commented defaults in main_parse_command_line. No functional change.
1.18	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.17	25-Sep-2021	rillig	indent: add nonnull memory allocation functions The only functional change is a single error message.
1.16	14-Mar-2021	rillig	indent: give indent a try at formatting its own code Formatting indent.h required the following manual corrections afterwards: The first tab in the comment in line 1 was replaced with a space but shouldn't be. The spacing around the '...' in function prototypes was completely wrong. It looked like 'const char ,...)__printflike', without any spaces. The '' of the return type 'const char *' was tied to the function name, even though this declaration was only for a single function. In such a case, it's more appropriate to line up the function names. The function-like macros were not indented to -di. This is something that I would not expect from indent, so it's ok to do that manually.
1.15	13-Mar-2021	rillig	indent: remove disabled duplicate RCS ID from header By convention, headers don't record their RCS ID.
1.14	13-Mar-2021	rillig	indent: rename pr_comment to process_comment, clean up documentation No functional change.
1.13	13-Mar-2021	rillig	indent: inline calls to count_spaces and count_spaces_until These two functions operated on column numbers instead of indentation, which required adjustments of '+ 1' and '- 1'. Their names were completely wrong since these functions did not count anything, instead they computed the column. No functional change.
1.12	13-Mar-2021	rillig	indent: replace column computation with indentation computation No functional change.
1.11	13-Mar-2021	rillig	indent: replace compute_code_column with compute_code_indent The goal is to only ever be concerned about the _indentation_ of a token, never the _column_ it appears in. Having only one of these avoids off-by-one errors. No functional change.
1.10	13-Mar-2021	rillig	indent: replace compute_label_column with compute_label_indent Using the invariant 'column == 1 + indent'. This removes several overly complicated '+ 1' from the code that are not needed conceptually. No functional change.
1.9	13-Mar-2021	rillig	indent: add debug logging for actually writing to the output file Together with the results of the tokenizer and the 4 buffers for token, label, code and comment, the debug log now provides a good high-level view on how the indentation happens and where to look for the many remaining bugs.
1.8	13-Mar-2021	rillig	indent: replace pad_output with output_indent Calculating the indentation is simpler than calculating the column, since that saves the constant addition and subtraction of the 1. No functional change.
1.7	12-Mar-2021	rillig	indent: add 'const', rename variables, reorder formula for tab width Column counting starts at 1. This 1 should rather be at the beginning of the formula since it is thought of being added at the very beginning of the line, not at the end. When adding a tab, the newly added tab is added at the end of the string, therefore that '+ 1' should be at the end of the formula as well. No functional change.
1.6	12-Mar-2021	rillig	indent: replace 'target' with 'indent' in function names The word 'target' was not as specific as possible. No functional change.
1.5	07-Mar-2021	rillig	indent: in debug mode, output detailed token information The main ingredient for understanding how indent works is the tokenizer and the 4 buffers in which the text is collected. Inspecting this debug log for the test comment-line-end makes it obvious why indent messes up code that contains '//' comments. The cause is that indent interprets '//' as an operator, just like '&&' or '\|\|'. The sequence '/////' is interpreted as a single operator as well, by the way. Since '//' is interpreted as an ordinary operator, any words following it are plain identifiers, usually several of them in a row, which is a syntax error. Depending on the context, the operator '//' is either a unary operator (no space around) or a binary operator (space around). This explains why the word 'line-end' is expanded to 'line - end'. No functional change outside of debug mode.
1.4	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.3	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.2	19-Oct-2019	christos	use stdarg, annotate function as __printflike and fix broken formats.
1.1	04-Apr-2019	kamil	branches: 1.1.1; 1.1.2; Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself. Part II, checkin new files.
1.1.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.2.2	10-Jun-2019	christos	Sync with HEAD
1.1.2.1	04-Apr-2019	christos	file indent.h was added on branch phil-wifi on 2019-06-10 22:10:20 +0000
1.1.1.1	04-Apr-2019	kamil	FreeBSD indent r340138
1.207.2.1	02-Aug-2025	perseant	Sync with HEAD
1.13	08-Oct-2021	rillig	indent: merge headers into a single file No functional change.
1.12	27-Sep-2021	rillig	indent: rename rwcode to keyword_kind, various cleanup No idea what the 'rw' in 'rwcode' meant, it had been imported that way 28 years ago. Since rwcode specifies the kind of a keyword, the prefix 'kw_' makes sense. No functional change.
1.11	09-Mar-2021	rillig	indent: rename a few more token types The previous names were either too short or ambiguous. No functional change.
1.10	09-Mar-2021	rillig	indent: make token names more precise The previous 'casestmt' was wrong since a case label is not a statement at all. The previous 'swstmt' was overly short, and wrong as well, since it represents only the 'switch (expr)' part, which is not a complete switch statement. Same for 'ifstmt', 'whilestmt', 'forstmt'. The previous word 'head' was not precise enough since it didn't specify exactly where the head ends and the body starts. Especially for handling the dangling else, this distinction is important. No functional change.
1.9	09-Mar-2021	rillig	indent: rename a few tokens to be more obvious For casual readers it is not obvious whether the 'sp' meant 'special' or 'space' or something entirely different.
1.8	07-Mar-2021	rillig	indent: use named constants for the different types of keywords This reduces the magic numbers in the code. Most of these had their designated constant name written in a nearby comment anyway. The one instance where arithmetic was performed on this new enum type (in indent.c) was a bit tricky to understand. The combination rw_continue_or_inline_or_restrict looks strange, the 'continue' should intuitively belong to the other control flow keywords in rw_break_or_goto_or_return. No functional change.
1.7	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.6	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.5	07-Aug-2003	agc	branches: 1.5.98; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.4	18-Oct-1997	mrg	merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.5.98.1	10-Jun-2019	christos	Sync with HEAD
1.49	08-Oct-2021	rillig	indent: merge headers into a single file No functional change.
1.48	07-Oct-2021	rillig	indent: rename bp_save to saved_inp_s, be_save to saved_inp_e Using the same naming convention makes it easier to relate the variables. No functional change.
1.47	07-Oct-2021	rillig	indent: group variables for the input buffer The input buffer follows the same concept as the intermediate buffers for label, code, comment and token, so use the same type for it. No functional change.
1.46	07-Oct-2021	rillig	indent: move definition of bufsize from header to implementation No functional change.
1.45	07-Oct-2021	rillig	indent: rename opt.btype_2 to brace_same_line No functional change.
1.44	07-Oct-2021	rillig	indent: fix wrong or outdated comments No functional change.
1.43	07-Oct-2021	rillig	indent: clean up colon handling No functional change.
1.42	05-Oct-2021	rillig	indent: rename n_real_blanklines The word 'n' was not as helpful as possible, the word 'real' did not give any clue at all about the variable's purpose. No functional change.
1.41	27-Sep-2021	rillig	indent: rename rwcode to keyword_kind, various cleanup No idea what the 'rw' in 'rwcode' meant, it had been imported that way 28 years ago. Since rwcode specifies the kind of a keyword, the prefix 'kw_' makes sense. No functional change.
1.40	26-Sep-2021	rillig	indent: fix documentation of opt.case_indent See io.c, compute_label_indent.
1.39	26-Sep-2021	rillig	indent: unexport global variables The variable match_state was write-only and was thus removed. No functional change.
1.38	26-Sep-2021	rillig	indent: negate and rename option.leave_comma The old name did not mirror the description in the manual page, and it was the only option that is negated. Inverting it allows the options table to be compressed.
1.37	25-Sep-2021	rillig	indent: misc cleanup No functional change.
1.36	25-Sep-2021	rillig	indent: convert found_err to bool That variable had slipped through the migration since it consequently used int for the declaration, the definition and all assignments. No functional change.
1.35	25-Sep-2021	rillig	indent: un-abbreviate a few parser_state members, clean up comments No functional change.
1.34	25-Sep-2021	rillig	indent: remove dead code for printing comments after empty lines This code has been commented out for at least 29 years. No functional change.
1.33	25-Sep-2021	rillig	indent: rename option variable to be more expressive No functional change.
1.32	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.31	25-Sep-2021	rillig	indent: convert parser_state from ibool to bool indent.c:400:5: error: suggest parentheses around assignment used as truth value io.c:271:32: error: ‘~’ on a boolean expression No functional change intended.
1.30	25-Sep-2021	rillig	indent: convert options from ibool to bool No functional change intended.
1.29	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.28	25-Sep-2021	rillig	indent: use standard definition for bool, true, false
1.27	25-Sep-2021	rillig	indent: move statistical values into a separate struct No functional change.
1.26	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.25	25-Sep-2021	rillig	indent: inline macro 'token' No functional change.
1.24	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.23	24-Sep-2021	rillig	indent: group global variables for label buffer into struct No functional change.
1.22	24-Sep-2021	rillig	indent: group global variables for the comment buffer No functional change.
1.21	13-Mar-2021	rillig	indent: fix documentation of parser_state.paren_indents The column position is not the same as the indentation (off-by-one).
1.20	13-Mar-2021	rillig	indent: distinguish between 'column' and 'indentation' column == 1 + indentation. In addition, indentation is a relative distance while column is an absolute position. Therefore, don't confuse these two concepts, to prevent off-by-one errors. No functional change.
1.19	13-Mar-2021	rillig	indent: rename pr_comment to process_comment, clean up documentation No functional change.
1.18	13-Mar-2021	rillig	indent: fix confusing variable names The word 'col' should only be used for the 1-based column number. This name is completely inappropriate for a line length since that provokes off-by-one errors. The name 'cols' would be acceptable although confusing since it sounds so similar to 'col'. Therefore, rename variables that are related to the maximum line length to 'line_length' since that makes for obvious code and nicely relates to the description of the option in the manual page. No functional change.
1.17	08-Mar-2021	rillig	indent: inline macro for backslash No functional change.
1.16	08-Mar-2021	rillig	indent: convert big macros to functions Each of these buffers is only modified in a single file. This makes it unnecessary to declare the macros in the global header.
1.15	07-Mar-2021	rillig	lint: move keyword 'continue' over to the other control flow keywords No functional change since neither rw_jump nor rw_inline_or_restrict is mentioned in any switch statement, and lint didn't find any other suspicious enum operations.
1.14	07-Mar-2021	rillig	indent: use named constants for the different types of keywords This reduces the magic numbers in the code. Most of these had their designated constant name written in a nearby comment anyway. The one instance where arithmetic was performed on this new enum type (in indent.c) was a bit tricky to understand. The combination rw_continue_or_inline_or_restrict looks strange, the 'continue' should intuitively belong to the other control flow keywords in rw_break_or_goto_or_return. No functional change.
1.13	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.12	23-Apr-2020	joerg	Avoid common symbol declarations
1.11	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.10	04-Sep-2014	mrg	branches: 1.10.16; port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.9	12-Apr-2009	lukem	branches: 1.9.24; Fix WARNS=4 issues (-Wshadow -Wcast-qual -Wsign-compare)
1.8	07-Aug-2003	agc	branches: 1.8.42; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.7	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.6	08-Oct-1998	wsanchez	Get rid of multiply defined common symbols
1.5	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.4	18-Oct-1997	mrg	merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.8.42.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.9.24.1	21-Sep-2014	snj	Pull up following revision(s) (requested by mrg in ticket #110): usr.bin/indent/io.c: revision 1.15 usr.bin/indent/indent_globs.h: revision 1.10 usr.bin/indent/args.c: revision 1.11 usr.bin/indent/indent.1: revision 1.23 usr.bin/indent/indent.c: revision 1.19 port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.10.16.1	10-Jun-2019	christos	Sync with HEAD
1.237	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.236	12-Dec-2024	rillig	indent: add error handling for I/O errors Suggested by lint2.
1.235	03-Dec-2023	rillig	branches: 1.235.2; indent: inline input-related macros No binary change.
1.234	03-Dec-2023	rillig	indent: group input-related variables into a struct No functional change.
1.233	03-Dec-2023	rillig	indent: use line number of the token start in diagnostics Previously, the line number of the end of the token was used, which was confusing in debug mode.
1.232	27-Jun-2023	rillig	indent: fix 'blank line above first statement in function body'
1.231	26-Jun-2023	rillig	indent: implement 'blank line above first statement in function body'
1.230	26-Jun-2023	rillig	indent: in -bad mode, don't add a blank line above a comment or '}'
1.229	17-Jun-2023	rillig	indent: clean up Extract duplicate code for handling line continuations. Prevent theoretic undefined behavior in strspn, as inp.s is not null-terminated. Remove adding extra space characters when processing comments, as these are not necessary to force a line of output. No functional change.
1.228	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.227	16-Jun-2023	rillig	indent: don't force a blank line between '}' and preprocessing line
1.226	16-Jun-2023	rillig	indent: rename a field of the parser state The previous name 'comment_in_first_line' was misleading, as it could mean that there was a comment in the first line of the file. No functional change.
1.225	15-Jun-2023	rillig	indent: consolidate handling of statement continuations
1.224	15-Jun-2023	rillig	indent: rename state variable to be more accurate No binary change.
1.223	15-Jun-2023	rillig	indent: fix indentation of multi-line enum constant initializers
1.222	15-Jun-2023	rillig	indent: miscellaneous cleanups, more tests for edge cases
1.221	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.220	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.219	14-Jun-2023	rillig	indent: clean up debugging code
1.218	14-Jun-2023	rillig	indent: clean up handling of comments One less moving part in the parser state. No functional change.
1.217	10-Jun-2023	rillig	indent: rename misleading variable The name started with 'line_start', but the value is not always the value from the beginning of the line. No functional change.
1.216	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.215	10-Jun-2023	rillig	indent: in debug mode, null-terminate buffers
1.214	10-Jun-2023	rillig	indent: distinguish blank lines from newline characters
1.213	10-Jun-2023	rillig	indent: fix indentation of continuation lines in initializers
1.212	10-Jun-2023	rillig	indent: extract output of an indented line to separate function
1.211	10-Jun-2023	rillig	indent: clean up function names and order in output
1.210	10-Jun-2023	rillig	indent: clean up function and variable names
1.209	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.208	09-Jun-2023	rillig	indent: trim trailing blank lines
1.207	09-Jun-2023	rillig	indent: when an indentation is ambiguous, indent one level further The '-eei' mode now applies whenever the indentation from a multi-line expression could be confused with a following statement.
1.206	09-Jun-2023	rillig	indent: format its own code
1.205	09-Jun-2023	rillig	indent: indent multi-line expressions according to parentheses This reverts the FreeBSD change from 2004-02-12 that had been imported on 2019-04-04.
1.204	08-Jun-2023	rillig	indent: fix indentation in multi-line else-if conditions
1.203	08-Jun-2023	rillig	indent: clean up and condense code No functional change.
1.202	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.201	06-Jun-2023	rillig	indent: condense code for writing tabs No functional change.
1.200	06-Jun-2023	rillig	indent: sort functions in call order No functional change.
1.199	06-Jun-2023	rillig	indent: compute indentation of 'case' labels on-demand One less moving part to keep track of. No functional change.
1.198	05-Jun-2023	rillig	indent: clean up comments
1.197	05-Jun-2023	rillig	indent: don't remove blank line after 'if (expr) {'
1.196	05-Jun-2023	rillig	indent: fix formatting of 'do' statements
1.195	05-Jun-2023	rillig	indent: clean up handling of whitespace No functional change.
1.194	05-Jun-2023	rillig	indent: let the output routines keep track of the indentation No functional change.
1.193	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.192	04-Jun-2023	rillig	indent: force at least one space after the colon of a label
1.191	04-Jun-2023	rillig	indent: track the kind of '{' on the parser stack
1.190	04-Jun-2023	rillig	indent: ensure that the 'block init level' never goes negative No functional change.
1.189	04-Jun-2023	rillig	indent: fix indentation of initializers in compound expressions
1.188	04-Jun-2023	rillig	indent: handle the indentation of 'case' in a simpler way
1.187	23-May-2023	rillig	indent: separate code for handling enums from the lexer The lexer's responsibility is to generate tokens, it's not supposed to update the parser state. Centralize the state transitions that control indentation of enum constants to keep the lexer code clean. Skip comments, newlines and preprocessing lines when updating the parser state for enum constants and for '*' in declarations.
1.186	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.185	22-May-2023	rillig	indent: implement suppressing optional blank lines
1.184	20-May-2023	rillig	indent: don't insert blank line between two closing lines
1.183	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.182	20-May-2023	rillig	indent: implement blank line above block comment
1.181	20-May-2023	rillig	indent: implement blank line after function body
1.180	20-May-2023	rillig	indent: ensure that no blank lines are inserted in INDENT OFF mode No blank lines were inserted previously, but the code looked suspicious as if that were possible.
1.179	20-May-2023	rillig	indent: implement blank lines around conditional compilation
1.178	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.177	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.176	18-May-2023	rillig	indent: remove unnecessary variable size optimization Due to the enum that follows in the struct, the short variable was padded to 4 bytes anyway. No functional change.
1.175	16-May-2023	rillig	indent: directly access the input buffer No functional change.
1.174	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.173	16-May-2023	rillig	indent: fix handling of INDENT OFF/ON comments Previously, the 'INDENT OFF' comments were interpreted when the newline token from the line above the comment was processed, which was earlier than could be reasonably expected. The 'INDENT ON' comments were interpreted equally early, which led to the situation that the 'INDENT OFF' comments were preserved literally but the 'INDENT ON' comments weren't.
1.172	16-May-2023	rillig	indent: move parsing of 'INDENT OFF/ON' comments to the lexer No functional change.
1.171	15-May-2023	rillig	indent: fix cast detection In process_lparen_or_lbracket, ps.paren[...].maybe_cast was not initialized, which may have been the cause for seemingly random spacing around binary operators. While here, clean up the code by reducing the number of accesses to the parser state.
1.170	15-May-2023	rillig	indent: indent multi-line conditions No functional change.
1.169	15-May-2023	rillig	indent: fix indentation of statements after controlling expression
1.168	15-May-2023	rillig	indent: fix indentation of expressions in -nlp -eei mode
1.167	15-May-2023	rillig	indent: remove redundant include lines
1.166	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.165	14-May-2023	rillig	indent: only null-terminate the buffers if necessary The only case where a buffer is used as a C-style string is when looking up a keyword. No functional change.
1.164	14-May-2023	rillig	indent: reduce code for scanning tokens The input line is guaranteed to end with '\n', so there's no need to carry another pointer around. No functional change.
1.163	14-May-2023	rillig	indent: remove foreign RCS IDs
1.162	14-May-2023	rillig	indent: miscellaneous cleanups
1.161	13-May-2023	rillig	indent: do not add a blank at the beginning of a line Most calls to output_line did already reset the variable. There may be some untested edge cases in or after comments, but these should be fine as well.
1.160	13-May-2023	rillig	indent: implement 'blank after declarations'
1.159	13-May-2023	rillig	indent: use enum instead of magic numbers for tracking declarations No functional change.
1.158	13-May-2023	rillig	indent: improve names of option variables No functional change.
1.157	13-May-2023	rillig	indent: rename struct fields for buffers No binary change except for assertion line numbers.
1.156	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.155	12-May-2023	rillig	indent: remove statistics The numbers from the statistics were wrong.
1.154	11-May-2023	rillig	indent: clean up input buffer handling No functional change.
1.153	11-May-2023	rillig	indent: don't touch comments in preprocessing lines The indentation of multi-line comments was wrong, and the code for handling them was too complicated.
1.152	11-May-2023	rillig	tests/indent: add more tests for preprocessing directives
1.151	11-May-2023	rillig	indent: remove unused code
1.150	11-May-2023	rillig	indent: remove broken code for handling blank lines This fixes several bugs where blank lines were erroneously added or removed, treating these old bugs for new bugs in different places. These new bugs are expected to be easier to fix, as the old bugs will not interfere anymore.
1.149	11-May-2023	rillig	indent: add debug output for tracking comments and braces
1.148	23-Apr-2022	rillig	indent: group global variables related to output control No functional change.
1.147	13-Feb-2022	rillig	indent: consistently use nparen for indexing parser_state.paren No binary change.
1.146	13-Feb-2022	rillig	indent: rename parser_state.p_l_follow and paren_level The previous variable names were misleading. Paren_level is not the current level of parentheses but the one from the beginning of the current output line. For better accuracy, rename it to line_start_paren_level. P_l_follow is not the level of parentheses that will be active at some point in the future, as the previous name suggested. Instead, it is the level of parentheses right now. For better accuracy, rename it to nparen. This nicely matches its main usage, which is as index to the parser_state.paren array. No binary change.
1.145	13-Feb-2022	rillig	indent: replace bitmasking code with struct The struct directly represents the properties of a pair of parentheses, without forcing the human reader to decode any bitset. This makes it easier to find the remaining bugs in the heuristic for determining the kind of parentheses. No functional change outside debug mode.
1.144	12-Feb-2022	rillig	indent: fix indentation of enum constants in typedef (since 2019-04-04) The solution is not elegant since it adds a small state machine inside the parser state, but at least these states only depend on the sequence of token types and not on any other part of the parser state. Reported in PR#55453.
1.143	28-Nov-2021	rillig	indent: clean up and document input handling The transformation of moving comments from after an 'if (expr)' after the following brace has a large implementation cost (about 300 lines of code) and makes input handling quite complicated. Document the overall idea to save future readers some time. No functional change.
1.142	27-Nov-2021	rillig	indent: accept a few formatting suggestions from indent The remaining issues are still that the conditions look ambiguous even with -eei, and that __attribute__ is broken into a separate line. No functional change.
1.141	27-Nov-2021	rillig	indent: rename dump functions to output No functional change.
1.140	27-Nov-2021	rillig	indent: add assertions for input handling Just to document the invariants; the code is already OK.
1.139	26-Nov-2021	rillig	indent: enhance debug logging for input handling
1.138	26-Nov-2021	rillig	indent: replace inp_enlarge with inp_add Previously, inbuf.inp.s was only updated at the very end of reading a line from the input file, which meant that during debugging, it pointed to invalid memory. Updating all fields in inbuf.inp after every reallocation makes the code less tricky to understand. No functional change.
1.137	26-Nov-2021	rillig	indent: split inp_read_line into smaller functions No functional change.
1.136	26-Nov-2021	rillig	indent: extract inp_from_file from inp_read_line No functional change.
1.135	26-Nov-2021	rillig	indent: remove code that fixes malformed preprocessor directives
1.134	26-Nov-2021	rillig	indent: move ind_add from io.c to indent.c It's a general-purpose function that is not directly related to input or output.
1.133	25-Nov-2021	rillig	indent: prevent undefined behavior in inp_line_start No functional change.
1.132	25-Nov-2021	rillig	indent: update cross-reference comments for bug in comment handling The function was renamed in io.c 1.122 from 2021-11-19.
1.131	25-Nov-2021	rillig	indent: rename ps.in_stmt to in_stmt_or_decl The previous name didn't match reality. No functional change.
1.130	25-Nov-2021	rillig	indent: rename ps.ind_stmt to in_stmt_cont This makes a comment redundant. No functional change.
1.129	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.128	19-Nov-2021	rillig	indent: keep inbuf.save_com_s and inbuf.save_com_e in sync No functional change.
1.127	19-Nov-2021	rillig	indent: fix included headers
1.126	19-Nov-2021	rillig	indent: clean up io.c No functional change.
1.125	19-Nov-2021	rillig	indent: replace ps.procname with ps.is_function_definition Omly the first character of ps.procname was ever read, and it was only compared to '\0'. Using a bool for this means simpler code, less memory and fewer wasted CPU cycles due to the removed strncpy. No functional change.
1.124	19-Nov-2021	rillig	indent: unexport inbuf No functional change.
1.123	19-Nov-2021	rillig	indent: use character input API from pr_comment.c No functional change.
1.122	19-Nov-2021	rillig	indent: remove all references to inbuf from indent.c No functional change.
1.121	19-Nov-2021	rillig	indent: move character input handling from indent.c to io.c No functional change.
1.120	19-Nov-2021	rillig	indent: move character input from indent.c to io.c No functional change.
1.119	19-Nov-2021	rillig	indent: use character input API from the tokenizer No functional change.
1.118	19-Nov-2021	rillig	indent: move character input handling from lexi.c to io.c No functional change.
1.117	19-Nov-2021	rillig	indent: group variables for input handling No functional change.
1.116	07-Nov-2021	rillig	indent: rename 'inbuf' functions to 'inp' The variable 'inp' used to be named 'inbuf'. Make the function names correspond to the variable name again. No functional change.
1.115	05-Nov-2021	rillig	indent: the '+ 1' in dump_line_code is not an off-by-one error
1.114	04-Nov-2021	rillig	indent: do not discard former error comments anymore Since io.c 1.20 from 2019-10-19, indent has not placed error comments in the code anymore. Since these comments are supposed to be cleaned up immediately, there is no point in having code for handling them.
1.113	04-Nov-2021	rillig	indent: extract compute_code_indent_lineup into separate function Having 9 different paths in a single function made it more complicated to understand than necessary. No functional change.
1.112	04-Nov-2021	rillig	indent: fix off-by-one confusion in paren_indent The variable was called 'indent' but actually contained a 'column', which was off by one. No functional change.
1.111	04-Nov-2021	rillig	indent: replace column computation with indentation computation No functional change.
1.110	04-Nov-2021	rillig	indent: group conditions in compute_code_indent by topic No functional change.
1.109	03-Nov-2021	rillig	indent: inline indentation_after, shorten function name to ind_add There were only few calls to indentation_after, so inlining it spares the need to look at yet another function definition. Another effect is that code.s and code.e appear in the code as a pair now, instead of a single code.s, making the scope of the function call obvious. In ind_add, there is no need to check for '\0' anymore since none of the buffers can ever contain a null character, these are filtered out by inbuf_read_line. No functional change.
1.108	30-Oct-2021	rillig	indent: inline macro label_offset No functional change.
1.107	29-Oct-2021	rillig	indent: merge isblank and is_hspace into ch_isblank No functional change.
1.106	29-Oct-2021	rillig	indent: fix undefined behavior in buffer handling Adding an arbitrary integer to a pointer may result in an out of bounds pointer, so replace the addition with a pointer subtraction. In the buffer handling functions, handle 'buf' and 'l' before 's' and 'e', since they are pairs. In inbuf_read_line, use 's' instead of 'buf' to make the code easier to understand for human readers. No functional change.
1.105	29-Oct-2021	rillig	indent: reorder global variables to be more intuitive The buffer 'inp' comes first. From there, a single token is read into the buffer 'token'. From there, it usually ends up in 'code'. The buffer 'token' does not belong to the group of the other 3 buffers, which together make up a line of formatted output. No functional change.
1.104	29-Oct-2021	rillig	indent: rename ps.dumped_decl_indent and indent_declaration The word 'dump' in 'ps.dumped_decl_indent' was too close to dump_line, which led to confusion since the variable controls whether the indentation has been added to the code buffer, which happens way before actually dumping the current line to the output file. The function name 'indent_declaration' was too unspecific, it did not reveal where the indentation of the declaration actually happened. No functional change.
1.103	27-Oct-2021	rillig	indent: fix indentation of local variable declarations This had been broken since the import of FreeBSD indent in 2019.
1.102	24-Oct-2021	rillig	indent: clean up format of warnings and errors Previously, warnings and errors had the form of C block comments. Before NetBSD io.c 1.20 from 2019-10-19, this format made sense because the diagnostics could end up in the same output stream as the formatted output. Since NetBSD io.c 1.20 from 2019-10-19, all diagnostics are redirected to stderr. This change was not mentioned in the commit message back then, it makes sense nevertheless. Since stdout and stderr now are properly separated, there is no need anymore to keep the weird format for warnings and errors. Switch to the standard 'error: file:line' format. Move the function 'diag' to indent.c to have access to the name of the current input file.
1.101	24-Oct-2021	rillig	indent: replace global variable use_ff with function parameter
1.100	24-Oct-2021	rillig	indent: sort includes
1.99	20-Oct-2021	rillig	indent: rename blankline_requested variables The words 'prefix' and 'postfix' sounded too much like horizontal concepts, like in operators. The actual purpose of these variables is to add blank lines before and after the current line, so use the same wording as in the command line options. No functional change.
1.98	20-Oct-2021	rillig	indent: rename next_blank_lines to blank_lines_to_output The previous name was already an improvement over the name before that (n_real_blanklines), but didn't express the intended purpose clearly enough, so try another name. No functional change.
1.97	19-Oct-2021	rillig	indent: always keep next_blank_lines >= 0 No functional change.
1.96	19-Oct-2021	rillig	indent: use simpler code for copying the input buffer In debug mode, this reduces the amount of debug output lines. No functional change in default mode.
1.95	19-Oct-2021	rillig	indent: if a file ends with indent off, don't add space-newline
1.94	11-Oct-2021	rillig	indent: use bool for suppress_blanklines It only ever got assigned the values 0 and 1. No functional change.
1.93	11-Oct-2021	rillig	indent: remove dead code
1.92	09-Oct-2021	rillig	indent: condense code for calculating indentations No functional change.
1.91	09-Oct-2021	rillig	indent: extract common code for advancing a single tab No functional change.
1.90	08-Oct-2021	rillig	indent: improve local variable names No functional change.
1.89	08-Oct-2021	rillig	indent: rename fill_buffer to inbuf_read_line No functional change.
1.88	08-Oct-2021	rillig	indent: run indent on indent.h The formatting looks mostly OK. Some struct members had excessively long names, leaving no space for their corresponding comments. Renamed some of them using well-known abbreviations. The formatting for debug_vis_range is messed up, no idea why. It is clearly a function declaration, not a function definition, so there is no need to place the function name in column 1. No functional change.
1.87	08-Oct-2021	rillig	indent: fix formatting of C99 comments The first attempt at formatting C99 comments was conceptually wrong. It accessed the next token in dump_line, even though that function should only ever look at the buffers for the label, the code and the current comment. (Understanding that part of the code was difficult at that time due to the sheer number of global variables.) The complicated and ever-growing condition for whether to output the token was a hack and in retrospect doesn't make sense at all, that's why it only came close to the intended effect. Some unintended side effects were that the C99 comments had an additional space in front of them, and that in some cases an empty line followed the comment, and that the comments were not aligned. Previously, the newline that terminates the C99 comment was included in the comment. Separating the newline from the comment fixed all these unintended side effects. The only downside is that the multi-line statement is not indented, but that should be easy to fix.
1.86	08-Oct-2021	rillig	indent: split dump_line into smaller functions No functional change.
1.85	08-Oct-2021	rillig	indent: replace column calculations with indent, part 4/4
1.84	08-Oct-2021	rillig	indent: replace column calculations with indent, part 3 No functional change.
1.83	08-Oct-2021	rillig	indent: replace column calculations with indent, part 2 No functional change.
1.82	08-Oct-2021	rillig	indent: calculate indentation instead of column This avoids constantly adding and subtracting 1. No functional change.
1.81	08-Oct-2021	rillig	indent: reduce indentation in dump_line
1.80	07-Oct-2021	rillig	indent: rename bp_save to saved_inp_s, be_save to saved_inp_e Using the same naming convention makes it easier to relate the variables. No functional change.
1.79	07-Oct-2021	rillig	indent: group variables for the input buffer The input buffer follows the same concept as the intermediate buffers for label, code, comment and token, so use the same type for it. No functional change.
1.78	07-Oct-2021	rillig	indent: use braces around multi-line statements No functional change.
1.77	07-Oct-2021	rillig	indent: let the code breathe a bit by inserting empty lines No functional change.
1.76	07-Oct-2021	rillig	indent: remove redundant comments No functional change.
1.75	07-Oct-2021	rillig	indent: clean up colon handling No functional change.
1.74	07-Oct-2021	rillig	indent: raise WARNS from the default 5 up to 6
1.73	05-Oct-2021	rillig	indent: rewrite parse_indent_comment for human readers The generated code is still very similar, GCC does a good job at inlining strncmp. No functional change.
1.72	05-Oct-2021	rillig	indent: rename n_real_blanklines The word 'n' was not as helpful as possible, the word 'real' did not give any clue at all about the variable's purpose. No functional change.
1.71	05-Oct-2021	rillig	indent: rename local char variable, reduce scope of counters No functional change.
1.70	05-Oct-2021	rillig	indent: use proper escape sequence for form feed This escape sequence has been available since at least 1978.
1.69	05-Oct-2021	rillig	indent: merge duplicate code into is_hspace No functional change.
1.68	26-Sep-2021	rillig	indent: unexport global variables The variable match_state was write-only and was thus removed. No functional change.
1.67	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.66	25-Sep-2021	rillig	indent: misc cleanup No functional change.
1.65	25-Sep-2021	rillig	indent: convert found_err to bool That variable had slipped through the migration since it consequently used int for the declaration, the definition and all assignments. No functional change.
1.64	25-Sep-2021	rillig	indent: un-abbreviate a few parser_state members, clean up comments No functional change.
1.63	25-Sep-2021	rillig	indent: remove dead code for printing comments after empty lines This code has been commented out for at least 29 years. No functional change.
1.62	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.61	25-Sep-2021	rillig	indent: convert parser_state from ibool to bool indent.c:400:5: error: suggest parentheses around assignment used as truth value io.c:271:32: error: ‘~’ on a boolean expression No functional change intended.
1.60	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.59	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.58	25-Sep-2021	rillig	indent: move statistical values into a separate struct No functional change.
1.57	25-Sep-2021	rillig	indent: add nonnull memory allocation functions The only functional change is a single error message.
1.56	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.55	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.54	24-Sep-2021	rillig	indent: group global variables for label buffer into struct No functional change.
1.53	24-Sep-2021	rillig	indent: extract parse_indent_comment from fill_buffer No functional change.
1.52	24-Sep-2021	rillig	indent: group global variables for the comment buffer No functional change.
1.51	24-Sep-2021	rillig	indent: rename local variable in fill_buffer The local variable name 'com' prevented grouping the global variables combuf, s_com, e_com and l_com into a struct named 'com'. No functional change.
1.50	24-Sep-2021	rillig	indent: fix token duplication after C99 comment The code that keeps blank lines after C99 comments still looks wrong, but at least it's better than before.
1.49	14-Mar-2021	rillig	indent: make compute_code_indent more readable The '?:' operator computing the factor was too hard to read. When quickly scanning the code, the 1 in the expression looked too much like it would be added to the indentation, which would turn the indentation length into a column number, and that again would smell like an off-by-one error. No functional change.
1.48	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.47	13-Mar-2021	rillig	indent: add debug logging for switching the input buffer No functional change outside debug mode.
1.46	13-Mar-2021	rillig	indent: align comments in indent's own code No functional change.
1.45	13-Mar-2021	rillig	indent: rename local variable in dump_line This clarifies that the variable names a column, not an indentation.
1.44	13-Mar-2021	rillig	indent: in dump_line, reduce scope of local variable This allows the variable 'target' in the lower half of the function to get a more specific name. No functional change.
1.43	13-Mar-2021	rillig	indent: distinguish between 'column' and 'indentation' column == 1 + indentation. In addition, indentation is a relative distance while column is an absolute position. Therefore, don't confuse these two concepts, to prevent off-by-one errors. No functional change.
1.42	13-Mar-2021	rillig	indent: fix confusing variable names The word 'col' should only be used for the 1-based column number. This name is completely inappropriate for a line length since that provokes off-by-one errors. The name 'cols' would be acceptable although confusing since it sounds so similar to 'col'. Therefore, rename variables that are related to the maximum line length to 'line_length' since that makes for obvious code and nicely relates to the description of the option in the manual page. No functional change.
1.41	13-Mar-2021	rillig	indent: inline calls to count_spaces and count_spaces_until These two functions operated on column numbers instead of indentation, which required adjustments of '+ 1' and '- 1'. Their names were completely wrong since these functions did not count anything, instead they computed the column. No functional change.
1.40	13-Mar-2021	rillig	indent: replace column computation with indentation computation No functional change.
1.39	13-Mar-2021	rillig	indent: replace compute_code_column with compute_code_indent The goal is to only ever be concerned about the _indentation_ of a token, never the _column_ it appears in. Having only one of these avoids off-by-one errors. No functional change.
1.38	13-Mar-2021	rillig	indent: replace compute_label_column with compute_label_indent Using the invariant 'column == 1 + indent'. This removes several overly complicated '+ 1' from the code that are not needed conceptually. No functional change.
1.37	13-Mar-2021	rillig	indent: manually fix indentation in indent's own source code
1.36	13-Mar-2021	rillig	indent: add debug logging for actually writing to the output file Together with the results of the tokenizer and the 4 buffers for token, label, code and comment, the debug log now provides a good high-level view on how the indentation happens and where to look for the many remaining bugs.
1.35	13-Mar-2021	rillig	indent: remove strange debugging code that went in the output file Whenever the code to be output contained the magic byte 0x80, instead of writing this byte, indent wrote the column number at the beginning of the code snippet, times 7. Especially the 'times 7' does not make any sense at all. In ISO-8859-1, this character position is not assigned. In Microsoft Codepage 1252 it is the Euro sign. In UTF-8 (which was probably not on the author's list when the code was originally written) it occurs as the middle byte for code points like U+2026 (horizontal ellipsis) from the block General Punctuation. Remove this strange code, thereby fixing indent for UTF-8 code. The code had been there since at least 1993-04-09, when it was first imported to NetBSD.
1.34	13-Mar-2021	rillig	indent: replace pad_output with output_indent Calculating the indentation is simpler than calculating the column, since that saves the constant addition and subtraction of the 1. No functional change.
1.33	13-Mar-2021	rillig	indent: clean up verbose documentation comments from the 1970s Since C90, there is no need to repeat the type of the function parameters. In the whole code of indent, there is a lot of confusion between the concepts of a 'column' (which is a position on the screen, counting starts at 1) and 'indentation' (which is a length, not a position). To avoid this confusion, the code will be rewritten anyway very soon. Repeatedly adding and subtracting 1 from the 'current column' is not elegant, this should rather be done by consistently measuring only the indentation from the left border (at offset 0), as a distance, not as an absolute position.
1.32	12-Mar-2021	rillig	indent: add 'const', rename variables, reorder formula for tab width Column counting starts at 1. This 1 should rather be at the beginning of the formula since it is thought of being added at the very beginning of the line, not at the end. When adding a tab, the newly added tab is added at the end of the string, therefore that '+ 1' should be at the end of the formula as well. No functional change.
1.31	12-Mar-2021	rillig	indent: replace 'target' with 'indent' in function names The word 'target' was not as specific as possible. No functional change.
1.30	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.29	12-Mar-2021	rillig	indent: make output_string inline GCC 9.3.0 didn't notice that the argument to this function is always a string literal, which makes it worthwhile to inline the call.
1.28	12-Mar-2021	rillig	indent: add helper functions for doing the actual output This allows to add debug logging to these few functions instead of all other places that might output something. Reducing the possible output formats to a few primitives makes dump_line simpler, especially the fprintf calls. It also removes the non-constant printf string. The call to output_int may be meant for debugging, as the character 0x80 is unlikely to appear in any real-world code. No functional change.
1.27	08-Mar-2021	rillig	indent: remove redundant initializer in dump_line No functional change.
1.26	08-Mar-2021	rillig	indent: move comment about dump_line to column 1 It looked misplaced on the right side since that area is usually reserved for small remarks, not long explanations. No functional change.
1.25	08-Mar-2021	rillig	indent: always use braces in do-while loops Having a 'while' at the beginning of a line looks as if it would start a loop. It's confusing when it _ends_ a loop instead.
1.24	07-Mar-2021	rillig	indent: fix handling of '//' end-of-line comments
1.23	07-Mar-2021	rillig	indent: remove redundant parentheses around return value No functional change.
1.22	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.21	06-Mar-2021	rillig	indent: fix space-tab alignment in indent's own code These parts are not fixed automatically by indent since they are in box comments. No functional change.
1.20	19-Oct-2019	christos	use stdarg, annotate function as __printflike and fix broken formats.
1.19	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.18	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.17	25-Feb-2016	ginsbach	branches: 1.17.16; Fix obvious contraction spelling mistakes by adding missing apostrophes.
1.16	22-Feb-2016	ginsbach	Use errx(3).
1.15	04-Sep-2014	mrg	port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.14	12-Apr-2009	lukem	branches: 1.14.24; Fix WARNS=4 issues (-Wshadow -Wcast-qual -Wsign-compare)
1.13	16-Oct-2003	itojun	branches: 1.13.42; safer use of realloc
1.12	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.11	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.10	14-Oct-2000	is	Due to infinite wisdom by the language designers, the difference of pointers has a type of (int) on i386 and (long) on sparc, and I don't even want to know what else on other cpu types.
1.9	19-Dec-1998	christos	branches: 1.9.2; 1.9.10; char -> unsigned char, braces for gcc-2.8.1
1.8	25-Aug-1998	ross	Add { and } to shut up egcs. Reformat the more questionable code.
1.7	30-Mar-1998	mrg	use static int instead of static
1.6	19-Oct-1997	mrg	fix compile warnings on the alpha.
1.5	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.4	18-Oct-1997	mrg	merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.9.10.1	17-Oct-2000	tv	Pullup 1.10 [is]: Due to infinite wisdom by the language designers, the difference of pointers has a type of (int) on i386 and (long) on sparc, and I don't even want to know what else on other cpu types.
1.9.2.1	19-Oct-2000	he	Pull up revision 1.10 (requested by is): The type of the difference between pointers is implementation- defined (long on sparc and int on most others). Compensate when printing.
1.13.42.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.14.24.1	21-Sep-2014	snj	Pull up following revision(s) (requested by mrg in ticket #110): usr.bin/indent/io.c: revision 1.15 usr.bin/indent/indent_globs.h: revision 1.10 usr.bin/indent/args.c: revision 1.11 usr.bin/indent/indent.1: revision 1.23 usr.bin/indent/indent.c: revision 1.19 port the -ut / -nut options from freebsd. -ut (default) enables tabs in output, the -nut uses spaces.
1.17.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.17.16.1	10-Jun-2019	christos	Sync with HEAD
1.235.2.1	02-Aug-2025	perseant	Sync with HEAD
1.242	03-Dec-2023	rillig	indent: inline input-related macros No binary change.
1.241	03-Dec-2023	rillig	indent: use line number of the token start in diagnostics Previously, the line number of the end of the token was used, which was confusing in debug mode.
1.240	03-Dec-2023	rillig	indent: fix line number counting in function definition In a function definition that is split on two lines, if the first line ends with a '*', the following line break didn't include the line number.
1.239	26-Jun-2023	rillig	indent: improve heuristics for '*' as pointer in for loops
1.238	26-Jun-2023	rillig	indent: improve heuristics for '*' as a pointer type
1.237	26-Jun-2023	rillig	indent: clean up indentation
1.236	25-Jun-2023	rillig	indent: move cast detection from the lexer to the main processor It is not the job of the lexer to modify the parser state.
1.235	25-Jun-2023	rillig	indent: treat 'complex' and 'imaginary' as type modifiers, not as types
1.234	25-Jun-2023	rillig	indent: fix formatting of parenthesized name in function definition
1.233	25-Jun-2023	rillig	indent: don't use strspn on inp_p, as it is not null-terminated No functional change.
1.232	17-Jun-2023	rillig	indent: clean up Extract duplicate code for handling line continuations. Prevent theoretic undefined behavior in strspn, as inp.s is not null-terminated. Remove adding extra space characters when processing comments, as these are not necessary to force a line of output. No functional change.
1.231	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.230	16-Jun-2023	rillig	indent: merge lexer symbols for type in/outside parentheses
1.229	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.228	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.227	14-Jun-2023	rillig	indent: remove another flag from parser state When processing a comment, the flag ps.next_col_1 was not used for the next token, but for a line within a comment. As its scope was limited to a single comment, there is no need to store it any longer than that No functional change.
1.226	14-Jun-2023	rillig	indent: remove a redundant flag from the parser state No functional change.
1.225	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.224	10-Jun-2023	rillig	indent: clean up function names, fix blank lines in debug output
1.223	10-Jun-2023	rillig	indent: in debug mode, null-terminate buffers
1.222	10-Jun-2023	rillig	indent: clean up function and variable names
1.221	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.220	09-Jun-2023	rillig	indent: clean up lexer No functional change.
1.219	09-Jun-2023	rillig	indent: improve heuristics for function declaration vs. definition
1.218	09-Jun-2023	rillig	indent: format its own code
1.217	08-Jun-2023	rillig	indent: remove fragile heuristic for detecting cast expressions The assumption that in an expression of the form '(a * anything)', the '*' marks a pointer type was too simple-minded. For now, fix the obvious cases and leave the others for later. If needed, they can be worked around using the '-T' option.
1.216	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.215	06-Jun-2023	rillig	indent: sort functions in call order No functional change.
1.214	04-Jun-2023	rillig	indent: do not parse '&&&&&&&' as a single binary operator
1.213	04-Jun-2023	rillig	indent: fix '*=' to be a binary operator, not a unary one
1.212	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.211	04-Jun-2023	rillig	indent: rename struct field, for better symmetry No binary change outside debug mode.
1.210	04-Jun-2023	rillig	lint: use separate lexer symbols for 'case' and 'default' It's not strictly necessary since these tokens behave in the same way, still, the code is more straight-forward when there are separate tokens.
1.209	04-Jun-2023	rillig	indent: classify 'inline' as a modifier rather than a word
1.208	04-Jun-2023	rillig	indent: use separate lexer symbols for the different kinds of ':'
1.207	04-Jun-2023	rillig	indent: separate code for handling parentheses and brackets Handling parentheses is more complicated than for brackets.
1.206	23-May-2023	rillig	indent: separate code for handling enums from the lexer The lexer's responsibility is to generate tokens, it's not supposed to update the parser state. Centralize the state transitions that control indentation of enum constants to keep the lexer code clean. Skip comments, newlines and preprocessing lines when updating the parser state for enum constants and for '*' in declarations.
1.205	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.204	23-May-2023	rillig	indent: fix spacing in declarations in for loops
1.203	22-May-2023	rillig	indent: adjust indentation in lexer No binary change.
1.202	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.201	20-May-2023	rillig	indent: clean up lexing of word tokens No functional change.
1.200	20-May-2023	rillig	indent: separate detection of function definitions from lexing '*' No functional change.
1.199	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.198	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.197	16-May-2023	rillig	indent: directly access the input buffer No functional change.
1.196	16-May-2023	rillig	indent: allow comments in column 1 to be formatted
1.195	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.194	16-May-2023	rillig	indent: fix handling of INDENT OFF/ON comments Previously, the 'INDENT OFF' comments were interpreted when the newline token from the line above the comment was processed, which was earlier than could be reasonably expected. The 'INDENT ON' comments were interpreted equally early, which led to the situation that the 'INDENT OFF' comments were preserved literally but the 'INDENT ON' comments weren't.
1.193	16-May-2023	rillig	indent: move parsing of 'INDENT OFF/ON' comments to the lexer No functional change.
1.192	15-May-2023	rillig	indent: clean up detection of whether parentheses form a cast No functional change.
1.191	15-May-2023	rillig	indent: improve type guessing, fix formatting of declarations
1.190	15-May-2023	rillig	indent: remove backslash line continuation outside preprocessing The indenter did not handle these backslashes well, interpreting them as unary operators, and they are an edge case anyway. Line continuations in string literals and character constants are kept.
1.189	15-May-2023	rillig	indent: indent multi-line conditions No functional change.
1.188	15-May-2023	rillig	indent: let indent format its own code With manual corrections, as indent does not properly indent multi-line '?:' expressions nor multi-line controlling expressions.
1.187	15-May-2023	rillig	indent: clean up memory allocation No functional change.
1.186	15-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.185	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.184	14-May-2023	rillig	indent: only null-terminate the buffers if necessary The only case where a buffer is used as a C-style string is when looking up a keyword. No functional change.
1.183	14-May-2023	rillig	indent: reduce code for scanning tokens The input line is guaranteed to end with '\n', so there's no need to carry another pointer around. No functional change.
1.182	14-May-2023	rillig	indent: remove foreign RCS IDs
1.181	14-May-2023	rillig	indent: miscellaneous cleanups
1.180	14-May-2023	rillig	indent: reduce binary size No functional change.
1.179	13-May-2023	rillig	indent: fix lexing of numbers that are spread over multiple lines
1.178	13-May-2023	rillig	indent: rename struct fields for buffers No binary change except for assertion line numbers.
1.177	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.176	12-May-2023	rillig	indent: condense code for handling spaced expressions No functional change outside debug mode.
1.175	11-May-2023	rillig	indent: move parser state variables to the parser_state struct Include the variables in the debug output.
1.174	11-May-2023	rillig	indent: move force_nl into the parser state This way, it is included in the debug output. No functional change.
1.173	11-May-2023	rillig	indent: remove buggy code for swapping tokens It is not the job of an indenter to swap tokens, even if it's only about placing comments elsewhere. The code that swapped the tokens was complicated, buggy and impossible to understand. In -br (brace right) mode, indent no longer moves a '{' from the beginning of a line to the end of the previous line, as that was handled by the token swapping code as well. This change is unintended, but it will be easier to re-add that now that the code is simpler.
1.172	13-Feb-2022	rillig	indent: rename parser_state.p_l_follow and paren_level The previous variable names were misleading. Paren_level is not the current level of parentheses but the one from the beginning of the current output line. For better accuracy, rename it to line_start_paren_level. P_l_follow is not the level of parentheses that will be active at some point in the future, as the previous name suggested. Instead, it is the level of parentheses right now. For better accuracy, rename it to nparen. This nicely matches its main usage, which is as index to the parser_state.paren array. No binary change.
1.171	13-Feb-2022	rillig	indent: replace bitmasking code with struct The struct directly represents the properties of a pair of parentheses, without forcing the human reader to decode any bitset. This makes it easier to find the remaining bugs in the heuristic for determining the kind of parentheses. No functional change outside debug mode.
1.170	13-Feb-2022	rillig	indent: change parser_state.cast_mask to 0-based indexing Having 1-based indexing was completely unexpected, and it didn't match the 0-based indexing of parser_state.paren_indents. No functional change.
1.169	12-Feb-2022	rillig	indent: fix indentation of enum constants in typedef (since 2019-04-04) The solution is not elegant since it adds a small state machine inside the parser state, but at least these states only depend on the sequence of token types and not on any other part of the parser state. Reported in PR#55453.
1.168	12-Feb-2022	rillig	indent: extend debug logging for the parser state The member names in struct parser_state are not trustworthy, for example in_decl does not correspond to the intuitive definition of "inside a declaration". To cope with this uncertainty, output the full state of the parser state to the debug log, not only the changes. This helps to track the inner state for small differences in the input, such as between 'typedef enum { TA, TB } TT' and 'enum { EA, EB } ET'. This hopefully helps in fixing PR#55453. No functional change outside debug mode.
1.167	28-Nov-2021	rillig	indent: treat L"string" as a single token There is never whitespace between the 'L' and the string literal or the character constant. There might be a backslash-newline between them, but that case was not handled before either. No functional change.
1.166	27-Nov-2021	rillig	indent: illustrate probably_looking_at_definition with examples No functional change.
1.165	27-Nov-2021	rillig	indent: fix out of bounds memory access (since 2021-11-25)
1.164	25-Nov-2021	rillig	indent: rename ps.in_function_parameters to match reality This flag is only set while parsing the parameters of a function definition, but not for a function declaration. See buffer_add in the test fmt_decl. No functional change.
1.163	25-Nov-2021	rillig	indent: improve heuristic for spaces around '*' in declarations
1.162	25-Nov-2021	rillig	indent: eliminate 3 negations in tokenizer No functional change.
1.161	25-Nov-2021	rillig	indent: extract lex_asterisk_unary into separate function No functional change.
1.160	25-Nov-2021	rillig	indent: condense code for building tokens from characters No functional change.
1.159	25-Nov-2021	rillig	indent: in lexi, assign lsym and next_unary in consistent order No functional change.
1.158	25-Nov-2021	rillig	indent: fix heuristic for declaration/definition to post-1990 reality
1.157	25-Nov-2021	rillig	indent: fix space after function name for option '-pcs'
1.156	25-Nov-2021	rillig	indent: fix spacing for unknown type names in declarations
1.155	25-Nov-2021	rillig	indent: extract probably_looking_at_definition to separate function This heuristic guesses wrong in many cases and will some cleanups. No functional change.
1.154	25-Nov-2021	rillig	indent: merge duplicate code for parsing 'struct s *' No functional change.
1.153	25-Nov-2021	rillig	indent: fix formatting of a few declarations involving unknown types
1.152	25-Nov-2021	rillig	indent: rename ps.in_stmt to in_stmt_or_decl The previous name didn't match reality. No functional change.
1.151	25-Nov-2021	rillig	indent: rename ps.ind_stmt to in_stmt_cont This makes a comment redundant. No functional change.
1.150	20-Nov-2021	rillig	indent: clean up lint annotation and tests
1.149	20-Nov-2021	rillig	indent: fix tokenizing of word-like tokens (since 2019-04-04) After a backslash-newline, the first character of the next line is only part of the identifier if it is an identifier character. indent-2000.10.11.14.46.04 \| int var \ \| +name = 4; indent-2012.11.20.03.02.57 indent-2014.09.04.04.06.07 \| int var \ \| +name = 4; indent-2019.02.03.03.19.29 indent-2019.04.04.15.27.35 \| int var+name = 4; indent-2021.11.19.20.23.17 indent \| int var + name = 4;
1.148	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.147	19-Nov-2021	rillig	indent: replace ps.procname with ps.is_function_definition Omly the first character of ps.procname was ever read, and it was only compared to '\0'. Using a bool for this means simpler code, less memory and fewer wasted CPU cycles due to the removed strncpy. No functional change.
1.146	19-Nov-2021	rillig	indent: fix formatting of function definitions (since 2019-04-04) In the definition of a function with a pointer return type, the formatting depended on the name of the function. Function names matching [A-Za-z+] were formatted correctly, those containing [$0-9_] weren't.
1.145	19-Nov-2021	rillig	indent: merge duplicate code into is_identifier_part No functional change.
1.144	19-Nov-2021	rillig	indent: fix lost function name (since 2019-04-04) When indent searched for an identifier followed by a '(', to see whether the identifier is a function name, it didn't care that the input buffer could be resized due to a long line, which had made the pointer 'tp' invalid. Fix this by stopping the search at the end of the line. A better approach would be to have an unlimited lookahead buffer for situations like these. The code that deals with character input has already been extracted to io.c, so it's possible to implement that now. While here, fix another access to undefined memory, after the loop. There is still the issue of overwriting procname[0] with a blank, which results in inconsistent formatting depending on the function name, probably another case of accessing undefined memory, although the results have been reproducible, but that may have been pure luck. The formatted code looks clearly broken, but that's still better than losing a token and destroying the whole file.
1.143	19-Nov-2021	rillig	indent: use character input API from the tokenizer No functional change.
1.142	19-Nov-2021	rillig	indent: move character input handling from lexi.c to io.c No functional change.
1.141	19-Nov-2021	rillig	indent: replace direct access to the input buffer This is a preparation for abstracting away all the low-level details of handling the input. The goal is to fix the current bugs regarding line number counting, out of bounds memory access, and generally unreadable code. No functional change.
1.140	19-Nov-2021	rillig	indent: group variables for input handling No functional change.
1.139	18-Nov-2021	rillig	indent: prevent use-after-free bug Triggered by the following artificial program: ---- snip ---- int * f ( void) { } ---- snap ----
1.138	07-Nov-2021	rillig	indent: various cleanups Make several comments more precise. Rename process_end_of_file to process_eof to match the token name. Change the order of assignments in analyze_comment to keep the com_ind computations closer together. In copy_comment_wrap, use pointer difference instead of pointer addition to stay away from undefined behavior. No functional change.
1.137	07-Nov-2021	rillig	indent: rename ps.decl_nest to decl_level This better matches the comment. No functional change.
1.136	07-Nov-2021	rillig	indent: move ps.p_l_follow closer to lsym_type_outside_parentheses This makes it easier to see the relation between these two. No functional change.
1.135	07-Nov-2021	rillig	indent: rename type_at_paren_level_0 to type_outside_parentheses For symmetry with type_in_parentheses. No functional change.
1.134	07-Nov-2021	rillig	indent: distinguish between typename in parentheses and other words This gets rid of two members of parser_state. No functional change for well-formed programs. The sequence of '++int' or '--size_t' may be formatted differently than before, but no program is expected to contain that sequence. Rename lsym_ident to lsym_word since 'ident' was too specific. This token type is used for constants and string literals as well. Strictly speaking, a string literal is not a word, but at least it's better than before.
1.133	07-Nov-2021	rillig	indent: rename 'inbuf' functions to 'inp' The variable 'inp' used to be named 'inbuf'. Make the function names correspond to the variable name again. No functional change.
1.132	05-Nov-2021	rillig	indent: consistently use token.e[-1] for the last added character No functional change.
1.131	05-Nov-2021	rillig	indent: add debug output for remaining members of parser_status
1.130	05-Nov-2021	rillig	indent: rename ps.curr_newline to next_col_1 For symmetry with ps.curr_col_1. No functional change.
1.129	01-Nov-2021	rillig	indent: fix missing blank after 'return' (since 2021-10-31) In indent.c 1.200 from 2021-10-31, the subtypes of identifier tokens were removed since they were redundant. An unintended side effect was that a parenthesized expression after 'return' was no longer separated by a blank. Before that change, 'return' was tokenized as an lsym_ident with subtype kw_other, and want_space_before_lparen handled this case in the last line. After the change, 'return' was treated as an ordinary identifier, and unless the option '-pcs' (blank after function call) was given, the blank was removed. The other keywords that had kw_other are not affected since they do not expect a '(' afterwards. These keywords are 'break', 'continue', 'goto', 'inline' and 'restrict'. Curiously, there was not a single test case that covered 'return(expr)'. While here, remove the trailing ',' from the enum lexer_symbol, which is not allowed in standard C, it is a GNU extension. Lint doesn't complain about this since the default LINTFLAGS include '-g' for GCC mode.
1.128	31-Oct-2021	rillig	indent: clean up Initialize buffers in reading order, make comments more expressive, rename add_typename to register_typename, remove unused macro. No functional change.
1.127	31-Oct-2021	rillig	indent: remove redundant keyword.is_type It is still confusing that not all type keywords end up as lsym_type. Those that occur inside parentheses end up as identifiers instead. To see whether an identifier is a typename, query ps.curr_is_type and ps.prev_is_type. No functional change.
1.126	31-Oct-2021	rillig	indent: replace kw_tag with lsym_tag This leaves only one special type of token, which is lsym_ident, which in some cases represents a type name and in other cases an identifier, constant or string literal. No functional change.
1.125	31-Oct-2021	rillig	indent: replace simple cases of keyword_kind with lexer_symbol The remaining keyword kinds 'tag' and 'type' require a bit more thought, so do them in a separate step. No functional change.
1.124	31-Oct-2021	rillig	indent: rename lsym_type to better reflect reality Type names that occur in parentheses are parsed as lsym_ident having the subtype kw_type instead. No functional change.
1.123	31-Oct-2021	rillig	indent: remove support for pre-1978 variable initialization
1.122	31-Oct-2021	rillig	indent: in debug log, print token subtype in same line The keyword 'void' is parsed as lsym_type in some cases and lsym_ident in others. Its corresponding keyword is always kw_type though. Put the subtype into the same line as the other token information.
1.121	31-Oct-2021	rillig	indent: add separate lexer symbol for offsetof No functional change.
1.120	31-Oct-2021	rillig	indent: add separate lexer symbol for sizeof The plan is to get rid of the type keyword_kind, which largely overlaps with lexer_symbol. No functional change.
1.119	31-Oct-2021	rillig	indent: clean up definition of keywords Rename kw_struct_or_union_or_enum to the shorter kw_tag. Merge kw_jump with kw_inline_or_restrict since they are handled in the same way. No functional change.
1.118	31-Oct-2021	rillig	indent: condense lexi_alnum No functional change.
1.117	30-Oct-2021	rillig	indent: rename prev_newline and prev_col_1 to curr These two flags describe the token that is currently processed. In process_binary_op, curr_newline can never be true since newline is not a binary operator, so remove that condition. No functional change.
1.116	30-Oct-2021	rillig	indent: in debug output, list the new token first
1.115	30-Oct-2021	rillig	indent: clean up lexical analyzer Use traditional type for small unsigned numbers instead of uint8_t; the required header was not included. Remove assertion for debug mode; lint takes care of ensuring that the enum constants match the length of the names array. Constify a name array. Move the comparison function for bsearch closer to its caller. No functional change.
1.114	29-Oct-2021	rillig	indent: remove redundant comments, remove punctuation from debug log The comment about 'null stmt' between braces probably meant 'no statements between braces'. The comments at psym_switch_expr only repeated what the code says or had been outdated 29 years ago already since opt.case_indent does not have to be 'one level down'. In the debug log, the quotes around the symbol names are not necessary after a ':'. The parse stack also does not need this much punctuation. Reducing a do-while loop to nothing instead of a statement saves a few CPU cycles. It works because after each lbrace, a stmt is pushed to the parser stack. This stmt can only ever be reduced to a stmt_list but never be removed.
1.113	29-Oct-2021	rillig	indent: in debug mode, log only differences for most ps members
1.112	29-Oct-2021	rillig	indent: add detailed debug logging for the parser state
1.111	29-Oct-2021	rillig	indent: merge isblank and is_hspace into ch_isblank No functional change.
1.110	29-Oct-2021	rillig	indent: use prev/curr/next to refer to the current token The word 'last' just didn't match with 'next'. No functional change.
1.109	29-Oct-2021	rillig	indent: keep p_l_follow nonnegative, use consistent comparison No functional change.
1.108	29-Oct-2021	rillig	indent: spell 'parentheses' properly in messages and comments
1.107	28-Oct-2021	rillig	indent: remove unused local variable in lexi Since the previous commit, lexi is always called with the same argument, so remove that parameter. The previous commit broke the debug logging by not printing "transient state" anymore. Replace this with "rolled back parser state" at the caller's site. No functional change.
1.106	28-Oct-2021	rillig	indent: reduce negations in search_stmt_lookahead No functional change.
1.105	26-Oct-2021	rillig	indent: make ps.keyword easier to understand Previously, ps.keyword did not have any documentation and was not straight-forward. In some cases it was reset to kw_0, in others it was set to an interesting value. The idea behind it was to remember the kind of word of the previous token, to decide whether to have a space between sizeof or offsetof and a following '('. No functional change.
1.104	26-Oct-2021	rillig	indent: fix debug logging The parser state is not always 'ps', so the debug logging must use the correct state as well.
1.103	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.102	26-Oct-2021	rillig	indent: merge duplicate code in lexi_alnum
1.101	25-Oct-2021	rillig	indent: improve debug logging Output the various details in chronological order.
1.100	25-Oct-2021	rillig	indent: split type token_type into 3 separate types Previously, token_type was used for 3 different purposes: 1. symbol types from the lexer 2. symbol types on the parser stack 3. kind of control statement for 'if (expr)' and similar statements Splitting the 41 constants into separate types makes it immediately clear that the parser stack never handles comments, preprocessing lines, newlines, form feeds, the inner structure of expressions. Previously, the constant switch_expr was especially confusing since it was used for 3 different purposes: when returned from lexi, it represented the keyword 'switch', in the parser stack it represented 'switch (expr)', and it was used for a statement head as well. The only overlap between the lexer symbols and the parser symbols are '{' and '}', and the keywords 'do' and 'else'. To increase confusion, the constants of the previous token_type were in apparently random order and before 2021, they had cryptic, highly abbreviated names. No functional change.
1.99	24-Oct-2021	rillig	indent: rename form_feed to tt_lex_form_feed No functional change.
1.98	24-Oct-2021	rillig	indent: split kw_for_or_if_or_while into separate constants No functional change.
1.97	24-Oct-2021	rillig	indent: split kw_do_or_else into separate constants It was unnecessarily confusing to have the token types keyword_do_else, keyword_do and keyword_else at the same time, without any hint in what they differed. Some of the token types seem to be used by the lexer while others are used in the parse stack. Maybe all token types can be partitioned into these groups, which would suggest to use two different types for them. And if not, it's still clearer to have this distinction in the names of the constants. No functional change.
1.96	24-Oct-2021	rillig	indent: define lexi_end as function instead of macro
1.95	24-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards. Indent still does not get extra_expr_indent correctly, it also indents global variables after tagged declarations too deep. No functional change.
1.94	24-Oct-2021	rillig	indent: rename nitems to array_length
1.93	24-Oct-2021	rillig	indent: sort includes
1.92	20-Oct-2021	rillig	indent: rename ps.last_u_d to match its comment No functional change.
1.91	11-Oct-2021	rillig	indent: use separate variables for lexi_alnum and lexi These two uses of the variable are independent of each other. No functional change.
1.90	11-Oct-2021	rillig	indent: clean up comments in lexi and lexi_alnum No functional change.
1.89	11-Oct-2021	rillig	indent: extract lexi_alnum from lexi No functional change.
1.88	09-Oct-2021	rillig	indent: fix lint warning about bsearch discarding 'const' lexi.c(433): warning: call to 'bsearch' effectively discards 'const' from argument [346]
1.87	08-Oct-2021	rillig	indent: merge duplicate code in lexer No functional change.
1.86	08-Oct-2021	rillig	indent: rename in_or_st to init_or_struct This makes a few comments redundant. No functional change.
1.85	08-Oct-2021	rillig	indent: remove 'global' from the list of keywords Since 1978, 'global' has not been a keyword in C. Moreover, it was declared as a type while its name would rather suggest a storage class. Removing the keyword fixes the formatting of variables named 'global'.
1.84	08-Oct-2021	rillig	indent: clean up typename handling Unexport typenames list. Replace standard binary search with custom binary search that returns the inserting position. In is_typename, take advantage of the buffer type instead of using the standard C recipe for str_ends_with. No functional change.
1.83	08-Oct-2021	rillig	indent: enhance comments for lex_number state machine No functional change.
1.82	08-Oct-2021	rillig	indent: improve local variable names No functional change.
1.81	08-Oct-2021	rillig	indent: rename fill_buffer to inbuf_read_line No functional change.
1.80	08-Oct-2021	rillig	indent: constify detection of function names No functional change.
1.79	08-Oct-2021	rillig	indent: rename tokens lparen and rparen to be more precise No functional change.
1.78	07-Oct-2021	rillig	indent: group variables for the input buffer The input buffer follows the same concept as the intermediate buffers for label, code, comment and token, so use the same type for it. No functional change.
1.77	07-Oct-2021	rillig	indent: clean up code, remove outdated wrong comments No functional change.
1.76	07-Oct-2021	rillig	indent: use braces around multi-line statements No functional change.
1.75	07-Oct-2021	rillig	indent: let the code breathe a bit by inserting empty lines No functional change.
1.74	07-Oct-2021	rillig	indent: fix wrong or outdated comments No functional change.
1.73	07-Oct-2021	rillig	indent: raise WARNS from the default 5 up to 6
1.72	05-Oct-2021	rillig	indent: use buffer type in debug_print_buf That function had been created before 'struct buffer' was invented, therefore it used two pointers as parameters. Remove this redundancy. No functional change.
1.71	05-Oct-2021	rillig	indent: run indent on lexi.c, with manual corrections The variables 'keywords' and 'typenames' were indented using 8 spaces, even though -di0 was in effect, which should result in a single space, and -ut was in effect, which should result in a single tab instead of 8 spaces. The option -eei does not work as advertised, the controlling expressions are only indented by the normal amount, which easily leads to confusion as to whether the code belongs to the condition or the following statement.
1.70	05-Oct-2021	rillig	indent: untangle complicated condition in probably_typedef No functional change.
1.69	05-Oct-2021	rillig	indent: use proper escape sequence for form feed This escape sequence has been available since at least 1978.
1.68	05-Oct-2021	rillig	indent: merge duplicate code into is_hspace No functional change.
1.67	05-Oct-2021	rillig	indent: clean up code for appending to buffers Use *e++ for appending and e[-1] for testing the previously appended character, like in other places in the code. No functional change.
1.66	05-Oct-2021	rillig	indent: merge duplicate code for reading from input buffer No functional change.
1.65	03-Oct-2021	rillig	indent: fix lint warning about signed '>>' Lint couldn't infer that indent's list of type names will practically never contain more that 2 billion entries and that the result of '>>' would be the same in all cases.
1.64	27-Sep-2021	rillig	indent: use binary instead of linear search when adding types No functional change.
1.63	27-Sep-2021	rillig	indent: extract is_typename from lexi No functional change.
1.62	27-Sep-2021	rillig	indent: rename rwcode to keyword_kind, various cleanup No idea what the 'rw' in 'rwcode' meant, it had been imported that way 28 years ago. Since rwcode specifies the kind of a keyword, the prefix 'kw_' makes sense. No functional change.
1.61	26-Sep-2021	rillig	indent: unexport global variables The variable match_state was write-only and was thus removed. No functional change.
1.60	26-Sep-2021	rillig	indent: unexport keyword table, clean up No functional change.
1.59	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.58	25-Sep-2021	rillig	indent: merge duplicate code for token buffers No functional change.
1.57	25-Sep-2021	rillig	indent: extract probably_typedef into separate function This condition is complicated enough that it warrants being split into several clauses, maybe even with an explanation. No functional change.
1.56	25-Sep-2021	rillig	indent: reduce code and data size for lexing of numbers Instead of having a table of strings (121 pointers + 121 data relocations), reduce that table to the actual character data and use a secondary table for looking up the correct row in the main table. No functional change.
1.55	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.54	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.53	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.52	25-Sep-2021	rillig	indent: make lex_char_or_string simpler The previous code was so tricky that every second line needed a comment that explains what's going on. Replace the complicated code with the usual straight-forward string-copying patterns. No functional change.
1.51	25-Sep-2021	rillig	indent: add nonnull memory allocation functions The only functional change is a single error message.
1.50	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.49	25-Sep-2021	rillig	indent: inline macro 'token' No functional change.
1.48	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.47	25-Sep-2021	rillig	indent: rename variables of type token_type The previous variable name 'code' conflicts with the buffer of the same name. No functional change.
1.46	24-Sep-2021	rillig	indent: group global variables for label buffer into struct No functional change.
1.45	24-Sep-2021	rillig	indent: group global variables for the comment buffer No functional change.
1.44	24-Sep-2021	rillig	indent: fix space-tab in indentation
1.43	26-Aug-2021	rillig	indent: extract lex_number, lex_word, lex_char_or_string No functional change.
1.42	25-Aug-2021	rillig	indent: fix lint warnings about type conversions on ilp32 No functional change.
1.41	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.40	13-Mar-2021	rillig	indent: remove redundant parentheses No functional change.
1.39	13-Mar-2021	rillig	indent: add debug logging for actually writing to the output file Together with the results of the tokenizer and the 4 buffers for token, label, code and comment, the debug log now provides a good high-level view on how the indentation happens and where to look for the many remaining bugs.
1.38	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.37	12-Mar-2021	rillig	indent: fix misleading indentation in indent's own code No functional change.
1.36	12-Mar-2021	rillig	indent: move code for tokenizing numbers further up Having it directly below the table makes it easier understandable. I also tried to omit this function entirely by moving the code into the initializer itself, but that made the code redundant and furthermore increased the size of the resulting binary, probably because of the new relocation records. No functional change.
1.35	11-Mar-2021	rillig	indent: reduce indentation of check_size functions No functional change.
1.34	11-Mar-2021	rillig	indent: remove redundant cast after allocation functions No functional change.
1.33	11-Mar-2021	rillig	indent: use consistent array indexing No functional change.
1.32	11-Mar-2021	rillig	indent: merge duplicate code for reading from the input buffer No functional change.
1.31	09-Mar-2021	rillig	indent: rename a few more token types The previous names were either too short or ambiguous. No functional change.
1.30	09-Mar-2021	rillig	indent: make token names more precise The previous 'casestmt' was wrong since a case label is not a statement at all. The previous 'swstmt' was overly short, and wrong as well, since it represents only the 'switch (expr)' part, which is not a complete switch statement. Same for 'ifstmt', 'whilestmt', 'forstmt'. The previous word 'head' was not precise enough since it didn't specify exactly where the head ends and the body starts. Especially for handling the dangling else, this distinction is important. No functional change.
1.29	09-Mar-2021	rillig	indent: rename a few tokens to be more obvious For casual readers it is not obvious whether the 'sp' meant 'special' or 'space' or something entirely different.
1.28	09-Mar-2021	rillig	indent: manually indent comments It's strange that indent's own code is not formatted by indent itself, which would be a good demonstration of its capabilities. In its current state, I don't trust indent to get even the tokenization correct, therefore the only safe way is to format the code manually.
1.27	08-Mar-2021	rillig	indent: split bsearch comparison function It may have been a clever trick to use the same memory layout for struct templ and a string pointer, but it's not worth the extra comment and difficulty in understanding the code. No functional change.
1.26	08-Mar-2021	rillig	indent: inline macro for backslash No functional change.
1.25	08-Mar-2021	rillig	indent: convert big macros to functions Each of these buffers is only modified in a single file. This makes it unnecessary to declare the macros in the global header.
1.24	07-Mar-2021	rillig	indent: fix handling of '//' end-of-line comments
1.23	07-Mar-2021	rillig	indent: remove redundant parentheses around return value No functional change.
1.22	07-Mar-2021	rillig	lint: move keyword 'continue' over to the other control flow keywords No functional change since neither rw_jump nor rw_inline_or_restrict is mentioned in any switch statement, and lint didn't find any other suspicious enum operations.
1.21	07-Mar-2021	rillig	indent: use named constants for the different types of keywords This reduces the magic numbers in the code. Most of these had their designated constant name written in a nearby comment anyway. The one instance where arithmetic was performed on this new enum type (in indent.c) was a bit tricky to understand. The combination rw_continue_or_inline_or_restrict looks strange, the 'continue' should intuitively belong to the other control flow keywords in rw_break_or_goto_or_return. No functional change.
1.20	07-Mar-2021	rillig	indent: in debug mode, output detailed token information The main ingredient for understanding how indent works is the tokenizer and the 4 buffers in which the text is collected. Inspecting this debug log for the test comment-line-end makes it obvious why indent messes up code that contains '//' comments. The cause is that indent interprets '//' as an operator, just like '&&' or '\|\|'. The sequence '/////' is interpreted as a single operator as well, by the way. Since '//' is interpreted as an ordinary operator, any words following it are plain identifiers, usually several of them in a row, which is a syntax error. Depending on the context, the operator '//' is either a unary operator (no space around) or a binary operator (space around). This explains why the word 'line-end' is expanded to 'line - end'. No functional change outside of debug mode.
1.19	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.18	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.17	19-Oct-2019	christos	use stdarg, annotate function as __printflike and fix broken formats.
1.16	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.15	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.14	05-Jun-2016	dholland	branches: 1.14.16; Fix CSRG-era typo: typedef, not typdef. Spotted by Piotr Stefaniak.
1.13	12-Apr-2009	lukem	Fix WARNS=4 issues (-Wshadow -Wcast-qual -Wsign-compare)
1.12	07-Aug-2003	agc	branches: 1.12.42; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.11	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.10	22-Mar-2002	kristerw	Recognize all C9x integer constants (ISO/IEC 9899:1999 section 6.4.4.1) Patch taken from FreeBSD. Fixes PR bin/9219.
1.9	15-Mar-1999	kristerw	Made indent recognize the [fF], [uU], [lL], [uU][lL], [lL][lL], and [uU][lL][lL] constant suffixes. (PR bin/6516 by Brian Ginsbach)
1.8	19-Dec-1998	christos	char -> unsigned char, braces for gcc-2.8.1
1.7	25-Aug-1998	ross	Add { and } to shut up egcs. Reformat the more questionable code.
1.6	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.5	18-Oct-1997	mrg	merge lite-2.
1.4	09-Sep-1997	agc	Bump number of elements in specials array from 100 to 1000. Typedefs are added to this array, and it silently ignores any attempts to enter more elements when the array is full.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.12.42.1	13-May-2009	jym	Sync with HEAD. Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html
1.14.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.14.16.1	10-Jun-2019	christos	Sync with HEAD
1.85	07-Jan-2025	rillig	indent: condense and simplify parsing code
1.84	07-Jan-2025	rillig	indent: fix indentation of statement after deeply nested 'if'
1.83	07-Jan-2025	rillig	indent: fix indentation of comment above 'else' Previously, indent assumed that no 'else' would follow.
1.82	04-Jan-2025	rillig	indent: fix indentation of adjacent multi-line initializers The main topic of this change is parse.c:66, which makes the indentation of statements uniform with the indentation of other parser symbols. That change had the side effect of messing up the indentation of files whose first line does not start in column 1, such as in ps_ind_level.c. To fix this side effect, the initial indentation must be determined before pushing the placeholder token psym_stmt during initialization.
1.81	04-Jan-2025	rillig	indent: make debug log more uniform
1.80	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.79	18-Jun-2023	rillig	branches: 1.79.2; indent: untangle code for handling the statement indentation The expression 'psyms.level-- - 2' did too much in a single line, so extract the '--' to a separate statement, to highlight the symmetry between the 'sym' and 'ind_level' code. No functional change.
1.78	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.77	14-Jun-2023	rillig	indent: clean up the code, add a few tests
1.76	14-Jun-2023	rillig	indent: allow more than 128 brace levels
1.75	14-Jun-2023	rillig	indent: fix out-of-bounds read when reducing a statement Since parse.c 1.73 from today. The parser symbol psym_stmt_list that was removed in that commit acted as a stop symbol, so that psyms_reduce_stmt would save a memory access.
1.74	14-Jun-2023	rillig	indent: clean up array indexing for parser symbols With 'top' pointing to the actual top element, the array was indexed in the closed range from 0 to top. All other arrays are indexed by the usual half-open interval from 0 to len. No functional change.
1.73	14-Jun-2023	rillig	indent: merge parser symbols for stmt and stmt_list They were handled in exactly the same way.
1.72	10-Jun-2023	rillig	indent: fix stack overflow, add more tests For several parser symbols, 2 symbols are pushed in a row, which led to an out-of-bounds write.
1.71	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.70	09-Jun-2023	rillig	indent: format its own code
1.69	07-Jun-2023	rillig	indent: extract the stack of parser symbols to a separate struct No functional change.
1.68	06-Jun-2023	rillig	indent: sort functions in call order No functional change.
1.67	06-Jun-2023	rillig	indent: compute indentation of 'case' labels on-demand One less moving part to keep track of. No functional change.
1.66	05-Jun-2023	rillig	indent: rename variables, clean up comments No binary change.
1.65	04-Jun-2023	rillig	indent: track the kind of '{' on the parser stack
1.64	03-Jun-2023	rillig	indent: clean up handling of brace indentation No functional change.
1.63	02-Jun-2023	rillig	indent: fix formatting of declarations with preprocessing lines
1.62	23-May-2023	rillig	indent: split debug output into paragraphs The paragraphs separate the different processing steps: getting a token from the lexer, processing the token, updating the parser state, sending a finished line to the output.
1.61	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.60	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.59	16-May-2023	rillig	indent: allow comments in column 1 to be formatted
1.58	15-May-2023	rillig	indent: format its own code, extend some comments With manual corrections, as there are still some bugs left. No functional change.
1.57	15-May-2023	rillig	indent: remove redundant include lines
1.56	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.55	14-May-2023	rillig	indent: remove foreign RCS IDs
1.54	13-May-2023	rillig	indent: move debugging code to separate file No functional change.
1.53	12-May-2023	rillig	indent: rename placeholder symbol for parser stack No functional change outside debug mode.
1.52	12-May-2023	rillig	tests/indent: test pushing the placeholder symbol to the parser stack
1.51	12-May-2023	rillig	indent: condense code for handling spaced expressions No functional change outside debug mode.
1.50	11-May-2023	rillig	indent: remove buggy code for swapping tokens It is not the job of an indenter to swap tokens, even if it's only about placing comments elsewhere. The code that swapped the tokens was complicated, buggy and impossible to understand. In -br (brace right) mode, indent no longer moves a '{' from the beginning of a line to the end of the previous line, as that was handled by the token swapping code as well. This change is unintended, but it will be easier to re-add that now that the code is simpler.
1.49	22-Apr-2022	rillig	indent: remove FreeBSD IDs Most of the IDs were empty anyway.
1.48	07-Nov-2021	rillig	indent: various cleanups Make several comments more precise. Rename process_end_of_file to process_eof to match the token name. Change the order of assignments in analyze_comment to keep the com_ind computations closer together. In copy_comment_wrap, use pointer difference instead of pointer addition to stay away from undefined behavior. No functional change.
1.47	29-Oct-2021	rillig	indent: remove redundant comments, remove punctuation from debug log The comment about 'null stmt' between braces probably meant 'no statements between braces'. The comments at psym_switch_expr only repeated what the code says or had been outdated 29 years ago already since opt.case_indent does not have to be 'one level down'. In the debug log, the quotes around the symbol names are not necessary after a ':'. The parse stack also does not need this much punctuation. Reducing a do-while loop to nothing instead of a statement saves a few CPU cycles. It works because after each lbrace, a stmt is pushed to the parser stack. This stmt can only ever be reduced to a stmt_list but never be removed.
1.46	29-Oct-2021	rillig	indent: remove redundant comments The comments only repeated what the constants for the parser symbols already express in their names. In the past, the names of these constants were inconsistent and misleading; back then, it made sense to make the comments express the actual meaning of the constants.
1.45	29-Oct-2021	rillig	indent: reduce indentation in parse, extract decl_level No functional change.
1.44	28-Oct-2021	rillig	indent: clean up indentation, comments, reduce No functional change.
1.43	28-Oct-2021	rillig	indent: clean up comments and function names Having accurate names for the lexer symbols and the parser symbols makes most of the comments redundant. Remove these. Rename process_decl to process_type, to match the name of the corresponding lexer symbol. In this phase, it's just a single type token, not a whole declaration. No functional change.
1.42	26-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards, to compensate for the remaining bugs in indent. Without the type definitions in .indent.pro, the opening braces of the functions kw_name and lexi_alnum would not be at the beginning of the line.
1.41	25-Oct-2021	rillig	indent: do not output token in debug mode When the parse stack is manipulated, the text of the token is not relevant anymore and may even be confusing, for example when parsing if_expr, the token may contain "}".
1.40	25-Oct-2021	rillig	indent: rename search_brace to search_stmt No functional change.
1.39	25-Oct-2021	rillig	indent: split type token_type into 3 separate types Previously, token_type was used for 3 different purposes: 1. symbol types from the lexer 2. symbol types on the parser stack 3. kind of control statement for 'if (expr)' and similar statements Splitting the 41 constants into separate types makes it immediately clear that the parser stack never handles comments, preprocessing lines, newlines, form feeds, the inner structure of expressions. Previously, the constant switch_expr was especially confusing since it was used for 3 different purposes: when returned from lexi, it represented the keyword 'switch', in the parser stack it represented 'switch (expr)', and it was used for a statement head as well. The only overlap between the lexer symbols and the parser symbols are '{' and '}', and the keywords 'do' and 'else'. To increase confusion, the constants of the previous token_type were in apparently random order and before 2021, they had cryptic, highly abbreviated names. No functional change.
1.38	24-Oct-2021	rillig	indent: split kw_do_or_else into separate constants It was unnecessarily confusing to have the token types keyword_do_else, keyword_do and keyword_else at the same time, without any hint in what they differed. Some of the token types seem to be used by the lexer while others are used in the parse stack. Maybe all token types can be partitioned into these groups, which would suggest to use two different types for them. And if not, it's still clearer to have this distinction in the names of the constants. No functional change.
1.37	24-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards. Indent still does not get extra_expr_indent correctly, it also indents global variables after tagged declarations too deep. No functional change.
1.36	20-Oct-2021	rillig	indent: rename parser stack variables No functional change.
1.35	08-Oct-2021	rillig	indent: clean up comments, parentheses, debug messages, boolean operator No functional change.
1.34	08-Oct-2021	rillig	indent: clean up 'parse', add test for dangling else No functional change.
1.33	07-Oct-2021	rillig	indent: rename opt.btype_2 to brace_same_line No functional change.
1.32	07-Oct-2021	rillig	indent: let the code breathe a bit by inserting empty lines No functional change.
1.31	07-Oct-2021	rillig	indent: clean up comments No functional change.
1.30	07-Oct-2021	rillig	indent: remove redundant comments No functional change.
1.29	05-Oct-2021	rillig	indent: fix Clang-Tidy warnings, clean up bakcopy The comment above and inside bakcopy had been outdated for at least the last 28 years, the backup file is named "%s.BAK", not ".B%s". Prevent buffer overflow for very long filenames (sprintf -> snprintf).
1.28	05-Oct-2021	rillig	indent: rename local char variable, reduce scope of counters No functional change.
1.27	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.26	25-Sep-2021	rillig	indent: un-abbreviate a few parser_state members, clean up comments No functional change.
1.25	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.24	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.23	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.22	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.21	25-Sep-2021	rillig	indent: inline macro 'token' No functional change.
1.20	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.19	25-Sep-2021	rillig	indent: rename variables of type token_type The previous variable name 'code' conflicts with the buffer of the same name. No functional change.
1.18	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.17	09-Mar-2021	rillig	indent: make token names more precise The previous 'casestmt' was wrong since a case label is not a statement at all. The previous 'swstmt' was overly short, and wrong as well, since it represents only the 'switch (expr)' part, which is not a complete switch statement. Same for 'ifstmt', 'whilestmt', 'forstmt'. The previous word 'head' was not precise enough since it didn't specify exactly where the head ends and the body starts. Especially for handling the dangling else, this distinction is important. No functional change.
1.16	09-Mar-2021	rillig	indent: extract reduce_stmt from reduce This refactoring reduces the indentation of the code, as well as removing any ambiguity as to which 'switch' statement a 'break' belongs, as there are no more nested 'switch' statements. No functional change.
1.15	09-Mar-2021	rillig	indent: manually indent comments It's strange that indent's own code is not formatted by indent itself, which would be a good demonstration of its capabilities. In its current state, I don't trust indent to get even the tokenization correct, therefore the only safe way is to format the code manually.
1.14	07-Mar-2021	rillig	indent: in debug mode, output detailed token information The main ingredient for understanding how indent works is the tokenizer and the 4 buffers in which the text is collected. Inspecting this debug log for the test comment-line-end makes it obvious why indent messes up code that contains '//' comments. The cause is that indent interprets '//' as an operator, just like '&&' or '\|\|'. The sequence '/////' is interpreted as a single operator as well, by the way. Since '//' is interpreted as an ordinary operator, any words following it are plain identifiers, usually several of them in a row, which is a syntax error. Depending on the context, the operator '//' is either a unary operator (no space around) or a binary operator (space around). This explains why the word 'line-end' is expanded to 'line - end'. No functional change outside of debug mode.
1.13	07-Mar-2021	rillig	indent: for the token types, use enum instead of #define This makes it easier to step through the code in a debugger. No functional change.
1.12	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.11	06-Mar-2021	rillig	indent: fix space-tab alignment in indent's own code These parts are not fixed automatically by indent since they are in box comments. No functional change.
1.10	19-Oct-2019	christos	use stdarg, annotate function as __printflike and fix broken formats.
1.9	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.8	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.7	07-Aug-2003	agc	branches: 1.7.98; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.6	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.5	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.4	18-Oct-1997	mrg	merge lite-2.
1.3	09-Jan-1997	tls	RCS ID police
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.7.98.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.7.98.1	10-Jun-2019	christos	Sync with HEAD
1.79.2.1	02-Aug-2025	perseant	Sync with HEAD
1.174	04-Jan-2025	rillig	indent: make debug output easier readable The previous format had the values of the parser state on the left side and the corresponding names on the right side. While it looked nicely aligned, it was not suitable for focusing on the actual data. Replace this format with the more common "key: value" format. Use the names of the enum constants in the debug log, instead of the previous "nice" names that needed one more level of mental translation and in some cases contained unbalanced punctuation such as '{'.
1.173	03-Dec-2023	rillig	branches: 1.173.2; indent: inline input-related macros No binary change.
1.172	03-Dec-2023	rillig	indent: use line number of the token start in diagnostics Previously, the line number of the end of the token was used, which was confusing in debug mode.
1.171	23-Jun-2023	rillig	indent: fix scanning of no-wrap comments (since 2021.11.07.10.34.03) The "refactoring" back then tried to be too clever.
1.170	18-Jun-2023	rillig	indent: only add blank lines before actual block comments
1.169	18-Jun-2023	rillig	indent: remove support for backspace in code and comments The C code in the whole tree does not contain a single literal backspace.
1.168	17-Jun-2023	rillig	indent: clean up Extract duplicate code for handling line continuations. Prevent theoretic undefined behavior in strspn, as inp.s is not null-terminated. Remove adding extra space characters when processing comments, as these are not necessary to force a line of output. No functional change.
1.167	17-Jun-2023	rillig	indent: miscellaneous cleanups No binary change.
1.166	16-Jun-2023	rillig	indent: rename a field of the parser state The previous name 'comment_in_first_line' was misleading, as it could mean that there was a comment in the first line of the file. No functional change.
1.165	14-Jun-2023	rillig	indent: allow more than 20 nested parentheses or brackets
1.164	14-Jun-2023	rillig	indent: clean up handling of comments One less moving part in the parser state. No functional change.
1.163	14-Jun-2023	rillig	indent: remove another flag from parser state When processing a comment, the flag ps.next_col_1 was not used for the next token, but for a line within a comment. As its scope was limited to a single comment, there is no need to store it any longer than that No functional change.
1.162	14-Jun-2023	rillig	indent: remove a redundant flag from the parser state No functional change.
1.161	10-Jun-2023	rillig	indent: miscellaneous cleanups
1.160	10-Jun-2023	rillig	indent: in debug mode, null-terminate buffers
1.159	10-Jun-2023	rillig	indent: rename and sort variables in parser state No functional change.
1.158	09-Jun-2023	rillig	indent: format its own code
1.157	09-Jun-2023	rillig	indent: preserve block comments with delimiters
1.156	09-Jun-2023	rillig	indent: express more clearly how delimited and no-wrap comments relate No functional change.
1.155	06-Jun-2023	rillig	indent: right-trim single-line comments
1.154	06-Jun-2023	rillig	indent: clean up formatting of comments No functional change.
1.153	06-Jun-2023	rillig	indent: simplify handling of comments No functional change.
1.152	06-Jun-2023	rillig	indent: split printing of comments into smaller functions No functional change.
1.151	04-Jun-2023	rillig	indent: remove read pointer from buffers that don't need it The only buffer that needs a read pointer is the current input line in 'inp'. No functional change.
1.150	04-Jun-2023	rillig	indent: fix out-of-bounds read when reading a comment
1.149	21-May-2023	rillig	tests/indent: fix outdated or wrong comments
1.148	20-May-2023	rillig	indent: extract the output state from the parser state The parser state depends on the preprocessing lines, the output state shouldn't.
1.147	20-May-2023	rillig	indent: implement blank line above block comment
1.146	18-May-2023	rillig	indent: manually wrap overly long lines No functional change.
1.145	18-May-2023	rillig	indent: switch to standard code style Taken from share/misc/indent.pro. Indent does not wrap code to fit into the line width, it only does so for comments. The 'INDENT OFF' sections and too long lines will be addressed in a follow-up commit. No functional change.
1.144	16-May-2023	rillig	indent: directly access the input buffer No functional change.
1.143	16-May-2023	rillig	indent: remove support for form feed characters inside a line Form feeds are occasionally used to split code into pages, and this use is still supported. Having a form feed in the middle of a line is exotic.
1.142	15-May-2023	rillig	indent: fix line wrapping of comments to the right of code
1.141	15-May-2023	rillig	indent: let indent format its own code With manual corrections, as indent does not properly indent multi-line '?:' expressions nor multi-line controlling expressions.
1.140	15-May-2023	rillig	indent: clean up memory and buffer management Remove the need to explicitly initialize the buffers. To avoid subtracting null pointers or comparing them using '<', migrate the buffers from the (start, end) form to the (start, len) form. This form also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s == buf.e' is used. Make buffer.st const, to avoid accidental modification of the buffer's content. Replace 'buf.e++ = ch' with buf_add_char, to avoid having to keep track how much unwritten space is left in the buffer. Remove all safety margins, that is, no more unchecked access to buf.st[-1] or appending using 'buf.e++'. Fix line number counting in lex_word for words that contain line breaks. No functional change.
1.139	14-May-2023	rillig	indent: only null-terminate the buffers if necessary The only case where a buffer is used as a C-style string is when looking up a keyword. No functional change.
1.138	14-May-2023	rillig	indent: shorten variable names for indenting comments No functional change.
1.137	14-May-2023	rillig	indent: fix handling of multiple block comments in a line
1.136	14-May-2023	rillig	indent: in comments, keep a leading tab This kind of comments is used for the CVS IDs at the top of files.
1.135	14-May-2023	rillig	indent: fix vertical spacing after declarations A comment is not supposed to change the state of the 'blank line after declaration', but it did. The initialization of saved_just_saw_decl was wrong though since it tried to capture the previous value after it had already been overwritten.
1.134	14-May-2023	rillig	indent: reduce code for scanning tokens The input line is guaranteed to end with '\n', so there's no need to carry another pointer around. No functional change.
1.133	14-May-2023	rillig	indent: remove foreign RCS IDs
1.132	13-May-2023	rillig	indent: use enum instead of magic numbers for tracking declarations No functional change.
1.131	13-May-2023	rillig	indent: rename struct fields for buffers No binary change except for assertion line numbers.
1.130	12-May-2023	rillig	indent: remove statistics The numbers from the statistics were wrong.
1.129	11-May-2023	rillig	indent: remove broken code for handling blank lines This fixes several bugs where blank lines were erroneously added or removed, treating these old bugs for new bugs in different places. These new bugs are expected to be easier to fix, as the old bugs will not interfere anymore.
1.128	09-May-2022	rillig	indent: clean up control flow, remove Capsicum No functional change.
1.127	23-Apr-2022	rillig	indent: group global variables related to output control No functional change.
1.126	27-Nov-2021	rillig	indent: rename dump functions to output No functional change.
1.125	26-Nov-2021	rillig	indent: add buf_add_range for adding characters to a buffer No functional change.
1.124	25-Nov-2021	rillig	indent: make fits_in_one_line independent of ps This makes it easier to understand the function. Having the dependency to the input line handling is already complicated enough. No functional change.
1.123	25-Nov-2021	rillig	indent: fix accidentally joined and broken comments (since 2019-04-04) The fixed version is not perfect as it gets the indentation of the last line of the first comment wrong, but at least indent doesn't generate malformed output anymore.
1.122	25-Nov-2021	rillig	tests/indent: demonstrate bugs in comment processing
1.121	19-Nov-2021	rillig	indent: reduce casts to unsigned char for character classification No functional change.
1.120	19-Nov-2021	rillig	indent: use character input API from pr_comment.c No functional change.
1.119	19-Nov-2021	rillig	indent: replace direct access to the input buffer This is a preparation for abstracting away all the low-level details of handling the input. The goal is to fix the current bugs regarding line number counting, out of bounds memory access, and generally unreadable code. No functional change.
1.118	19-Nov-2021	rillig	indent: rename input buffer variables From reading the names 'save_com' and 'sc_end', it was not obvious enough that these two variables are the limits of the same buffer, the names were just too unrelated. No functional change.
1.117	19-Nov-2021	rillig	indent: group variables for input handling No functional change.
1.116	07-Nov-2021	rillig	indent: various cleanups Make several comments more precise. Rename process_end_of_file to process_eof to match the token name. Change the order of assignments in analyze_comment to keep the com_ind computations closer together. In copy_comment_wrap, use pointer difference instead of pointer addition to stay away from undefined behavior. No functional change.
1.115	07-Nov-2021	rillig	indent: remove code that accessed out-of-bounds data from buffer Saving and restoring the exact buffer position had been necessary before NetBSD pr_comment.c 1.114. The code accessed the buffer data out of the bounds of [com.s, com.e), which was rather surprising. More specifically, it accessed com.e[-1] in a case where com.e == com.s, relying on com.e[-1] being ' ' in most cases and '*' in the case where the comment delimiter was written to a separate output line. Make the code easier understandable by only ever accessing the buffer data in the bounds [buf.s, buf.e). No functional change.
1.114	07-Nov-2021	rillig	indent: only access buffer data in the range [buf.s, buf.e) No functional change.
1.113	07-Nov-2021	rillig	indent: do not expand comment buffer unnecessarily This may have been a simple typo or a really tricky optimization that isn't obvious even after studying the code for several months. Either of these is bad, so use the standard form of resetting the buffer. No functional change.
1.112	07-Nov-2021	rillig	indent: remove dead code in analyze_comment The case of an otherwise empty line is already handled further above.
1.111	07-Nov-2021	rillig	indent: remove dead code from copy_comment_wrap C99 comments are not wrapped, therefore there is no need to check for them in copy_comment_wrap. No functional change.
1.110	07-Nov-2021	rillig	indent: move documentation from process_comment to copy_comment_wrap
1.109	07-Nov-2021	rillig	indent: clean up process_comment The assignments to the variables were redundant, they are already done by analyze_comment. No functional change.
1.108	07-Nov-2021	rillig	indent: remove redundant assignment At that point, ps.next_col_1 is already false. No functional change.
1.107	07-Nov-2021	rillig	indent: make end of comment detection of nowrap comments simpler No functional change.
1.106	07-Nov-2021	rillig	indent: clean up overcomplicated conditions in copy_comment_nowrap No functional change.
1.105	07-Nov-2021	rillig	indent: skip redundant conditions in copy_comment_nowrap No functional change.
1.104	07-Nov-2021	rillig	indent: clean up copy_comment_nowrap The action for '\f' was the same as the default action. Replacing 'switch' with 'if' makes the code shorter. No functional change.
1.103	07-Nov-2021	rillig	indent: make copy_comment_nowrap simpler Since a nowrap comment is copied unmodified, it need not depend on any maximum line length. No functional change.
1.102	07-Nov-2021	rillig	indent: remove dead code from process_comment_nowrap In comments that are preserved, no additional leading ' * ' is inserted. No functional change.
1.101	07-Nov-2021	rillig	indent: remove dead code from copy_comment_wrap No functional change.
1.100	07-Nov-2021	rillig	indent: remove dead code from copy_comment_nowrap No functional change.
1.99	07-Nov-2021	rillig	indent: split copy_comment into wrapping and non-wrapping These two cases are processed in an almost entirely different way. In particular, copy_comment_nowrap should copy the comment verbatim, which is not obvious from the current code, due to the many conditions and the complex control flow. No functional change.
1.98	07-Nov-2021	rillig	indent: rename 'inbuf' functions to 'inp' The variable 'inp' used to be named 'inbuf'. Make the function names correspond to the variable name again. No functional change.
1.97	05-Nov-2021	rillig	indent: rename ps.curr_newline to next_col_1 For symmetry with ps.curr_col_1. No functional change.
1.96	04-Nov-2021	rillig	indent: fix parsing of C99 comments containing '*/'
1.95	04-Nov-2021	rillig	indent: split process_comments into separate functions No functional change.
1.94	03-Nov-2021	rillig	indent: inline indentation_after, shorten function name to ind_add There were only few calls to indentation_after, so inlining it spares the need to look at yet another function definition. Another effect is that code.s and code.e appear in the code as a pair now, instead of a single code.s, making the scope of the function call obvious. In ind_add, there is no need to check for '\0' anymore since none of the buffers can ever contain a null character, these are filtered out by inbuf_read_line. No functional change.
1.93	30-Oct-2021	rillig	indent: rename prev_newline and prev_col_1 to curr These two flags describe the token that is currently processed. In process_binary_op, curr_newline can never be true since newline is not a binary operator, so remove that condition. No functional change.
1.92	30-Oct-2021	rillig	indent: fix bounds check for sc_buf Some years ago, save_com was an array of characters, used as temporary storage. When sc_buf was added, this code was forgotten. The bounds check must be on the array itself, not on an iterator that points somewhere in that array.
1.91	30-Oct-2021	rillig	indent: fix assertion in fits_in_one_line
1.90	29-Oct-2021	rillig	indent: merge isblank and is_hspace into ch_isblank No functional change.
1.89	29-Oct-2021	rillig	indent: fix undefined behavior in buffer handling Adding an arbitrary integer to a pointer may result in an out of bounds pointer, so replace the addition with a pointer subtraction. In the buffer handling functions, handle 'buf' and 'l' before 's' and 'e', since they are pairs. In inbuf_read_line, use 's' instead of 'buf' to make the code easier to understand for human readers. No functional change.
1.88	29-Oct-2021	rillig	indent: use prev/curr/next to refer to the current token The word 'last' just didn't match with 'next'. No functional change.
1.87	26-Oct-2021	rillig	indent: clean up process_comment There is no undefined behavior since the compared characters are always from the basic execution character set. All other cases are covered by the condition above for now_len. Fix debug logging for non-ASCII characters, previously a character was output as \xffffffc3.
1.86	26-Oct-2021	rillig	indent: reduce indentation in process_comment No functional change.
1.85	26-Oct-2021	rillig	indent: make reformatting of comments simpler No functional change.
1.84	25-Oct-2021	rillig	indent: split type token_type into 3 separate types Previously, token_type was used for 3 different purposes: 1. symbol types from the lexer 2. symbol types on the parser stack 3. kind of control statement for 'if (expr)' and similar statements Splitting the 41 constants into separate types makes it immediately clear that the parser stack never handles comments, preprocessing lines, newlines, form feeds, the inner structure of expressions. Previously, the constant switch_expr was especially confusing since it was used for 3 different purposes: when returned from lexi, it represented the keyword 'switch', in the parser stack it represented 'switch (expr)', and it was used for a statement head as well. The only overlap between the lexer symbols and the parser symbols are '{' and '}', and the keywords 'do' and 'else'. To increase confusion, the constants of the previous token_type were in apparently random order and before 2021, they had cryptic, highly abbreviated names. No functional change.
1.83	24-Oct-2021	rillig	indent: run indent on its own source code With manual corrections afterwards. Indent still does not get extra_expr_indent correctly, it also indents global variables after tagged declarations too deep. No functional change.
1.82	24-Oct-2021	rillig	indent: replace global variable use_ff with function parameter
1.81	24-Oct-2021	rillig	indent: clean up comments
1.80	20-Oct-2021	rillig	indent: rename blankline_requested variables The words 'prefix' and 'postfix' sounded too much like horizontal concepts, like in operators. The actual purpose of these variables is to add blank lines before and after the current line, so use the same wording as in the command line options. No functional change.
1.79	14-Oct-2021	rillig	indent: clarify that 25 is a magic number The extra line width for comments to the very right is just that, an arbitrarily chosen number. It neither has to be a multiple of 8, nor of the tabsize nor of the indentation. Since 25 is neither of these, this makes it a perfect choice, allowing these extreme comments to have 22 characters per line with -sc (leading asterisks in comment continuations, the default) or 25 without. No functional change.
1.78	14-Oct-2021	rillig	indent: turn ps.com_ind into local variable This makes it immediately clear that ps.com_ind is computed in the first part of process_comment and then only used from there. No functional change.
1.77	13-Oct-2021	rillig	indent: extract fits_in_one_line from process_comment No functional change.
1.76	12-Oct-2021	rillig	indent: in process_comment, migrate int variable to bool No functional change.
1.75	12-Oct-2021	rillig	indent: in process_comment, negate box_com to may_wrap In the new line 213, may_wrap could only be true and was therefore a redundant condition. No functional change.
1.74	12-Oct-2021	rillig	indent: negate a few condition in process_comment No functional change.
1.73	12-Oct-2021	rillig	indent: fix formatting of single-line comments (since today) The change in pr_comment.c 1.70 from 3 hours ago did not cover all edge cases correctly. Now it works for comments that are aligned with tabs.
1.72	12-Oct-2021	rillig	indent: replace unreachable code with assertion
1.71	12-Oct-2021	rillig	indent: use high-level buffer API for processing comments Document the trickery of copying the last word from the previous line since that it not obvious at all. No functional change.
1.70	12-Oct-2021	rillig	indent: fix wrapping for comments in otherwise empty lines The comment above the code was wrong. The leading 3 characters were indeed ignored, but the first of them was '/', not ' '. Of the trailing 3 characters, 2 were not ignored. The start and end of the comment would not cancel out, they would rather sum up.
1.69	09-Oct-2021	rillig	indent: extract common code for advancing a single tab No functional change.
1.68	08-Oct-2021	rillig	indent: replace unreachable code with assertions The input buffer is always supposed to be terminated with a newline. The function inbuf_read_line silently skips null characters. Since the input buffer is redirected to a temporary buffer in some cases, do not simply remove this supposed dead code, but replace it with assertions. In any case, if the code for calling inbuf_read_line had been reachable and actually allocated a different buffer, continuing to use the pointer p would have invoked undefined behavior anyway.
1.67	08-Oct-2021	rillig	indent: use standard error handling for unterminated comment Just writing it to stdout is bad, especially when indent is used in filter mode. Silently continuing after such an error is bad as well. echo '/*' \| indent
1.66	08-Oct-2021	rillig	indent: remove redundant conditions No functional change.
1.65	08-Oct-2021	rillig	indent: convert ps.box_com to local variable This variable is only used in a single function, and that function does not call any other function that could replace the parser state or install a temporary parser state. No functional change.
1.64	08-Oct-2021	rillig	indent: rename fill_buffer to inbuf_read_line No functional change.
1.63	08-Oct-2021	rillig	indent: run indent on indent.h The formatting looks mostly OK. Some struct members had excessively long names, leaving no space for their corresponding comments. Renamed some of them using well-known abbreviations. The formatting for debug_vis_range is messed up, no idea why. It is clearly a function declaration, not a function definition, so there is no need to place the function name in column 1. No functional change.
1.62	08-Oct-2021	rillig	indent: fix formatting of C99 comments The first attempt at formatting C99 comments was conceptually wrong. It accessed the next token in dump_line, even though that function should only ever look at the buffers for the label, the code and the current comment. (Understanding that part of the code was difficult at that time due to the sheer number of global variables.) The complicated and ever-growing condition for whether to output the token was a hack and in retrospect doesn't make sense at all, that's why it only came close to the intended effect. Some unintended side effects were that the C99 comments had an additional space in front of them, and that in some cases an empty line followed the comment, and that the comments were not aligned. Previously, the newline that terminates the C99 comment was included in the comment. Separating the newline from the comment fixed all these unintended side effects. The only downside is that the multi-line statement is not indented, but that should be easy to fix.
1.61	08-Oct-2021	rillig	indent: reduce scope of t_ptr in process_comment No functional change.
1.60	08-Oct-2021	rillig	indent: replace column calculations with indent, part 4/4
1.59	08-Oct-2021	rillig	indent: replace column calculations with indent, part 3 No functional change.
1.58	07-Oct-2021	rillig	indent: group variables for the input buffer The input buffer follows the same concept as the intermediate buffers for label, code, comment and token, so use the same type for it. No functional change.
1.57	07-Oct-2021	rillig	indent: use braces around multi-line statements No functional change.
1.56	07-Oct-2021	rillig	indent: let the code breathe a bit by inserting empty lines No functional change.
1.55	07-Oct-2021	rillig	indent: clean up comments No functional change.
1.54	07-Oct-2021	rillig	indent: remove redundant comments No functional change.
1.53	05-Oct-2021	rillig	indent: fix outdated comment, add 'const'
1.52	05-Oct-2021	rillig	indent: fix spelling in comments
1.51	05-Oct-2021	rillig	indent: use proper escape sequence for form feed This escape sequence has been available since at least 1978.
1.50	05-Oct-2021	rillig	indent: merge duplicate code into is_hspace No functional change.
1.49	05-Oct-2021	rillig	indent: clean up code for appending to buffers Use *e++ for appending and e[-1] for testing the previously appended character, like in other places in the code. No functional change.
1.48	05-Oct-2021	rillig	indent: merge duplicate code for reading from input buffer No functional change.
1.47	26-Sep-2021	rillig	indent: let indent format its own code -- in supervised mode After running indent on the code, I manually selected each change that now looks better than before. The remaining changes are left for later. All in all, indent did a pretty good job, except for syntactic additions from after 1990, but that was to be expected. Examples for such additions are GCC's __attribute__ and C99 designated initializers. Indent has only few knobs to tune the indentation. The knob for the continuation indentation applies to function declarations as well as to expressions. The knob for indentation of local variable declarations applies to struct members as well, even if these are members of a top-level struct. Several code comments crossed the right margin in column 78. Several other code comments were correctly broken though. The cause for this difference was not obvious. No functional change.
1.46	25-Sep-2021	rillig	indent: merge duplicate code for token buffers No functional change.
1.45	25-Sep-2021	rillig	indent: remove dead code for printing comments after empty lines This code has been commented out for at least 29 years. No functional change.
1.44	25-Sep-2021	rillig	indent: convert remaining ibool to bool No functional change intended.
1.43	25-Sep-2021	rillig	indent: prepare for lint's strict bool mode Before C99, C had no boolean type. Instead, indent used int for that, just like many other programs. Even with C99, bool and int can be used interchangeably in many situations, such as querying '!i' or '!ptr' or 'cond == 0'. Since January 2021, lint provides the strict bool mode, which makes bool a non-arithmetic type that is incompatible with any other type. Having clearly separate types helps in understanding the code. To migrate indent to strict bool mode, the first step is to apply all changes that keep the resulting binary the same. Since sizeof(bool) is 1 and sizeof(int) is 4, the type ibool serves as an intermediate type. For now it is defined to int, later it will become bool. The current code compiles cleanly in C99 and C11 mode, as well as in lint's strict bool mode. There are a few tricky places: In args.c in 'struct pro', there are two types of options: boolean and integer. Boolean options point to a bool variable, integer options point to an int variable. To keep the current structure of the code, the pointer has been changed to 'void *'. To ensure type safety, the definition of the options is done via preprocessor magic, which in C11 mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the very bottom of the Makefile.) In indent.c in process_preprocessing, a boolean variable is post-incremented. That variable is only assigned to another variable, and that variable is only used in a boolean context. To provoke a different behavior between the '++' and the '= true', the source code to be indented would need 1 << 32 preprocessing directives, which is unlikely to happen in practice. In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever get the values 0 and 1. For these values, the expressions 'a & ~b' and 'a && !b' are equivalent, in all versions of C. The compiler may generate different code for them, though. In io.c in parse_indent_comment, the assignment to inhibit_formatting takes place in integer context. If the compiler is smart enough to detect the possible values of on_off, it may generate the same code before and after the change, but that is rather unlikely. The second step of the migration will be to replace ibool with bool, step by step, just in case there are any hidden gotchas in the code, such as sizeof or pointer casts. No change to the resulting binary.
1.42	25-Sep-2021	rillig	indent: remove ifdef for lint NetBSD lint does not need them anymore, FreeBSD does not have lint.
1.41	25-Sep-2021	rillig	indent: move statistical values into a separate struct No functional change.
1.40	25-Sep-2021	rillig	indent: add nonnull memory allocation functions The only functional change is a single error message.
1.39	25-Sep-2021	rillig	indent: group global variables for token buffer No functional change.
1.38	25-Sep-2021	rillig	indent: group global variables for code buffer No functional change.
1.37	24-Sep-2021	rillig	indent: group global variables for label buffer into struct No functional change.
1.36	24-Sep-2021	rillig	indent: group global variables for the comment buffer No functional change.
1.35	14-Mar-2021	rillig	indent: clean up check_size_comment The additional parameter last_bl_ptr was only necessary because the last blank was stored as a pointer into the buffer. By storing the index in the buffer instead, it doesn't need to be updated all the time. No functional change.
1.34	14-Mar-2021	rillig	indent: remove trailing whitespace
1.33	14-Mar-2021	rillig	indent: clean up target column computation in process_comment No functional change.
1.32	14-Mar-2021	rillig	indent: fix off-by-one error in comment wrapping The manual page says that the default maximum length of a comment line is 78. The test 'comments.0' wrongly assumed that this 78 would refer to the maximum _column_ allowed, which is off by one. Fix the wording in the test 'comments.0' and remove the (now satisfied) expectation comments in the test 'token-comment.0'. Several other tests just happened to hit that limit, fix these as well.
1.31	14-Mar-2021	rillig	indent: fix lint warnings No functional change.
1.30	13-Mar-2021	rillig	indent: remove the '+ 1' from right margin calculation in comment No functional change.
1.29	13-Mar-2021	rillig	indent: distinguish between 'column' and 'indentation' column == 1 + indentation. In addition, indentation is a relative distance while column is an absolute position. Therefore, don't confuse these two concepts, to prevent off-by-one errors. No functional change.
1.28	13-Mar-2021	rillig	indent: rename pr_comment to process_comment, clean up documentation No functional change.
1.27	13-Mar-2021	rillig	indent: remove redundant parentheses No functional change.
1.26	13-Mar-2021	rillig	indent: fix confusing variable names The word 'col' should only be used for the 1-based column number. This name is completely inappropriate for a line length since that provokes off-by-one errors. The name 'cols' would be acceptable although confusing since it sounds so similar to 'col'. Therefore, rename variables that are related to the maximum line length to 'line_length' since that makes for obvious code and nicely relates to the description of the option in the manual page. No functional change.
1.25	13-Mar-2021	rillig	indent: document undefined behavior in processing of comments No functional change.
1.24	13-Mar-2021	rillig	indent: inline calls to count_spaces and count_spaces_until These two functions operated on column numbers instead of indentation, which required adjustments of '+ 1' and '- 1'. Their names were completely wrong since these functions did not count anything, instead they computed the column. No functional change.
1.23	13-Mar-2021	rillig	indent: replace compute_code_column with compute_code_indent The goal is to only ever be concerned about the _indentation_ of a token, never the _column_ it appears in. Having only one of these avoids off-by-one errors. No functional change.
1.22	13-Mar-2021	rillig	indent: replace compute_label_column with compute_label_indent Using the invariant 'column == 1 + indent'. This removes several overly complicated '+ 1' from the code that are not needed conceptually. No functional change.
1.21	13-Mar-2021	rillig	indent: replace pad_output with output_indent Calculating the indentation is simpler than calculating the column, since that saves the constant addition and subtraction of the 1. No functional change.
1.20	12-Mar-2021	rillig	indent: replace 'target' with 'indent' in function names The word 'target' was not as specific as possible. No functional change.
1.19	12-Mar-2021	rillig	indent: use consistent indentation for 'else' Half of the code used -ce, the other half the opposite -nce. No functional change.
1.18	11-Mar-2021	rillig	indent: reduce indentation of check_size functions No functional change.
1.17	11-Mar-2021	rillig	indent: remove redundant cast after allocation functions No functional change.
1.16	11-Mar-2021	rillig	indent: use consistent array indexing No functional change.
1.15	09-Mar-2021	rillig	indent: manually indent comments It's strange that indent's own code is not formatted by indent itself, which would be a good demonstration of its capabilities. In its current state, I don't trust indent to get even the tokenization correct, therefore the only safe way is to format the code manually.
1.14	08-Mar-2021	rillig	indent: convert big macros to functions Each of these buffers is only modified in a single file. This makes it unnecessary to declare the macros in the global header.
1.13	07-Mar-2021	rillig	indent: fix handling of '//' end-of-line comments
1.12	07-Mar-2021	rillig	indent: use all headers in all files This is a prerequisite for converting the token types to an enum instead of a preprocessor define, since the return type of lexi will become token_type. Having the enum will make debugging easier. There was a single naming collision, which forced the variable in scan_profile to be renamed. All other token names are used nowhere else. No change to the resulting binary.
1.11	04-Apr-2019	kamil	Upgrade indent(1) Merge all the changes from the recent FreeBSD HEAD snapshot into our local copy. FreeBSD actively maintains this program in their sources and their repository contains over 100 commits with changes. Keep the delta between the FreeBSD and NetBSD versions to absolute minimum, mostly RCS Id and compatiblity fixes. Major chages in this import: - Added an option -ldi<N> to control indentation of local variable names. - Added option -P for loading user-provided files as profiles - Added -tsn for setting tabsize - Rename -nsac/-sac ("space after cast") to -ncs/-cs - Added option -fbs Enables (disables) splitting the function declaration and opening brace across two lines. - Respect SIMPLE_BACKUP_SUFFIX environment variable in indent(1) - Group global option variables into an options structure - Use bsearch() for looking up type keywords. - Don't produce unneeded space character in function declarators - Don't unnecessarily add a blank before a comment ends. - Don't ignore newlines after comments that follow braces. Merge the FreeBSD intend(1) tests with our ATF framework. All tests pass. Upgrade prepared by Manikishan Ghantasala. Final polishing by myself.
1.10	25-Feb-2016	ginsbach	branches: 1.10.16; Fix obvious contraction spelling mistakes by adding missing apostrophes.
1.9	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22365, verified by myself.
1.8	19-Jun-2003	christos	PR/21645: Mishka: Localized comments don't work with indent.
1.7	26-May-2002	wiz	Remove #ifndef'd __STDC__ code. ANSIfy.
1.6	19-Oct-1997	lukem	WARNSify, fix .Nm usage, deprecate register, use <err.h>, KNFify (with indent!;)
1.5	18-Oct-1997	mrg	merge lite-2.
1.4	09-Jan-1997	tls	RCS ID police
1.3	07-Aug-1993	cgd	do block commenting, if comment begins with slash-star-newline.
1.2	01-Aug-1993	mycroft	Add RCS identifiers.
1.1	09-Apr-1993	cgd	branches: 1.1.1; added, from net/2 (patch 124).
1.1.1.2	04-Apr-2019	kamil	FreeBSD indent r340138
1.1.1.1	06-Jun-1993	mrg	4.4BSD-Lite2
1.10.16.1	10-Jun-2019	christos	Sync with HEAD
1.173.2.1	02-Aug-2025	perseant	Sync with HEAD

OpenGrok