document non-literal format strings

WFORMAT is no more...

Apply modified fix for PR bin/23763
jot(1)'s output now matches the examples given in the manpage.

More format string cleanup by sommerfeld.

branches: 1.2.8; 1.2.16;
Changed to conform to NetBSD's new RCS Id convention.

branches: 1.1.1;
Initial revision

Document the default values for options in a way that hopefully makes
much more sense.

Don't require any arguments as the usage indicates; simplifies the code
and the documentation: now "jot" works like "jot -" used to (and continues
to). From Ingo Schwarze

branches: 1.15.4;
Remove workaround for ancient HTML generation code.

Bump date for previous

Add HISTORY and AUTHORS to jot

John A. Kunze requested to add himself as the author of jot
on the FreeBSD bugzilla with the following text:

Please re-instate my name (as "John A. Kunze") as AUTHOR of the
jot, rs, and lam utilities. I wrote these in 1982, around or
before the time that I worked with Jordan at UCB (he left to work
at FreeBSD and later at Apple). Thank you!

-- https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=196786

While there add a note that this tool first appeared in BSD 4.2.
This information is based on the FreeBSD manual pages.

Patch submitted by Michal Mazurek <akfaew AT jasminek.net>.

Reviewed by <dholland>

Remove unnecessary Bk/Ek pairs from SYNOPSIS.
No effective change except where I used the opportunity to sort options
and/or option descriptions.

branches: 1.11.6;
Add an example of generating a list of integers; make it the first
example. Also add a reference to seq(1).

Use semantic markup

branches: 1.9.10;
Remove a 9 from the stuttered example in order to match the only
possible reality.

branches: 1.8.18; 1.8.24;
Sort options, standardize usage message.

ok wiz@

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22365, verified by myself.

.Nm does not need a dummy argument ("") before punctuation or
for correct formatting of the SYNOPSIS any longer.

Generate <>& symbolically. I'm avoiding .../dist/... directories for now.

Clean up SYNOPSIS formatting.

Reformat in mandoc.

branches: 1.2.4;
Changed to conform to NetBSD's new RCS Id convention.

branches: 1.1.1;
Initial revision

Generate format after setting the prec value

Otherwise we end up with setting invalid printf format: "%.-1f" (from
"if (snprintf(p, sz, "%%.%df", prec) >= (int)sz)") for prec equal to -1,
which is not canonical (rejected by gcc and clang when used explicitly).

Detected by ASan (MKSANITIZER) for "jot 8".

- 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

branches: 1.26.2;
Don't require any arguments as the usage indicates; simplifies the code
and the documentation: now "jot" works like "jot -" used to (and continues
to). From Ingo Schwarze

branches: 1.25.44;
fix sign-compare issue

branches: 1.24.6;
Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

branches: 1.23.2;
Fix integer random numbers when bounds given in the wrong order.

branches: 1.22.2;
When printing non-random integers, add 0.5 to get 'schoolboy' rounding.
This makes 'jot - 9 - -.5' generate 9 9 8 8 etc again.
When printing random numbers generate a sane range when ender < begin.
For random integers we must use random()/2^31 since we don't want the
value above generated should random() return 2^31-1.

Remove the loop used to set default and calculated parameters, instead
set the defaults at compile time and calculate any derived values in
a simple switch statement.
Use names for the bit-patterns that indicate that fields are present.
Seems to behave the same way as the earlier version.

Sprinkle 'static' and 'EXIT_FAILURE'. Use getopt().
Simplify getprec() using libc functions.
Code from me partially inspired by a diff sent by xtos.

When converting fp values to integer, use floor(x) so that we get
consistent convertions not the 'round to even'.
This means that it is possibly to generate the 'stuttered' sequence
4 4 3 3 2 2 1 1 0 0 -1 -1.

Rename 's' to 'step', allow a precision of 0 to be specified (overriding
the value determined from 'begin' and 'end').

If we are doing a random integer (or character) format, then increase the
range by 1 so that the quoted upper and lower values both appear, and
with the same frequency as the intermediate values.
This reverts the change from rev 1.13, and fixes it properly!
Also fixes 'jot -r -c - 97 98' as reported to netbsd-bugs (but not as a PR!).

Remove some pointless indirections and add {} around some long single
statements.

branches: 1.15.10; 1.15.16;
PR/33709: Kevin Massey: Avoid stack smashing in args copy

branches: 1.14.2;
Apply modified fix for PR bin/23763
jot(1)'s output now matches the examples given in the manpage.

de-__P, ANSIfy prototypes

Sort options, standardize usage message.

ok wiz@

Homogenize usage messages: make the 'usage' word all lowercase, as this seems
to be the most common practice in our tree.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22365, verified by myself.

Make the randomness a tad more random. It's still not terribly
random, but it's considerably less lame now.

before% echo `jot -r 5` ; echo `jot -r 5`
66 5 78 82 66
66 5 78 82 66

after% echo `jot -r 5` ; echo `jot -r 5`
15 94 47 32 25
78 52 73 52 47

Reverse the if and else clauses for the chardata test if there is
no format string specified.

Fixes PR bin/12424 from Jim Bernard.

Use patch provided Paul Janzen <pjanzen@OpenBSD.org> to fix
bin/12316 more completely.
Highlights:
* we only allow explicitly specified set of (numberic) formats
* %% is correctly skipped, extra %foo formats are detected
* '*' in format specification is not allowed

The code in getformat() is slighly different to the OpenBSD version
of fix - the list of allowed printf(3) flags is explicit and the
condition checking 'chardata' is written slighly differently.

Also pull-in some cosmetic changes from OpenBSD - s/isdefault()/is_default()/,
error() -> errx(), usage().

Don't coredump if format given via -w doesn't contain alphanumeric
character; in such case, warn user the format is invalid.

This addresses bin/12316 by Perry E. Metzger.

branches: 1.5.10;
char -> unsigned char

WARNSify, fix .Nm usage, deprecate register, use <err.h>

Allocate proper storage.
Avoid interpretation of boring words as printf(3) style format strings.

Changed to conform to NetBSD's new RCS Id convention.

branches: 1.1.1;
Initial revision