pi.c revision 1.13 1 /* $NetBSD: pi.c,v 1.13 2009/08/13 02:10:50 dholland Exp $ */
2
3 /*
4 * Copyright (c) 1980, 1993
5 * The Regents of the University of California. All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the distribution.
15 * 3. Neither the name of the University nor the names of its contributors
16 * may be used to endorse or promote products derived from this software
17 * without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32 #include <sys/cdefs.h>
33 #ifndef lint
34 #if 0
35 static char sccsid[] = "@(#)pi.c 8.1 (Berkeley) 6/6/93";
36 #endif
37 __RCSID("$NetBSD: pi.c,v 1.13 2009/08/13 02:10:50 dholland Exp $");
38 #endif /* not lint */
39
40 #include <stdio.h>
41 #include <ctype.h>
42 #include <string.h>
43 #include <stdlib.h>
44 #include "error.h"
45
46 static char *c_linenumber;
47 static char *unk_hdr[] = {"In", "program", "???"};
48 static char **c_header = &unk_hdr[0];
49
50 static boolean alldigits(char *);
51 static boolean isdateformat(int, char **);
52 static boolean instringset(char *, char **);
53 static boolean piptr(char *);
54
55
56 /*
57 * Attempt to handle error messages produced by pi (and by pc)
58 *
59 * problem #1: There is no file name available when a file does not
60 * use a #include; this will have to be given to error
61 * in the command line.
62 * problem #2: pi doesn't always tell you what line number
63 * a error refers to; for example during the tree
64 * walk phase of code generation and error detection,
65 * an error can refer to "variable foo in procedure bletch"
66 * without giving a line number
67 * problem #3: line numbers, when available, are attached to
68 * the source line, along with the source line itself
69 * These line numbers must be extracted, and
70 * the source line thrown away.
71 * problem #4: Some error messages produce more than one line number
72 * on the same message.
73 * There are only two (I think):
74 * %s undefined on line%s
75 * %s improperly used on line%s
76 * here, the %s makes line plural or singular.
77 *
78 * Here are the error strings used in pi version 1.2 that can refer
79 * to a file name or line number:
80 *
81 * Multiply defined label in case, lines %d and %d
82 * Goto %s from line %d is into a structured statement
83 * End matched %s on line %d
84 * Inserted keyword end matching %s on line %d
85 *
86 * Here are the general pi patterns recognized:
87 * define piptr == -.*^-.*
88 * define msg = .*
89 * define digit = [0-9]
90 * definename = .*
91 * define date_format letter*3 letter*3 (digit | (digit digit))
92 * (digit | (digit digit)):digit*2 digit*4
93 *
94 * {e,E} (piptr) (msg) Encounter an error during textual scan
95 * E {digit}* - (msg) Have an error message that refers to a new line
96 * E - msg Have an error message that refers to current
97 * function, program or procedure
98 * (date_format) (name): When switch compilation files
99 * ... (msg) When refer to the previous line
100 * 'In' ('procedure'|'function'|'program') (name):
101 * pi is now complaining about 2nd pass errors.
102 *
103 * Here is the output from a compilation
104 *
105 *
106 * 2 var i:integer;
107 * e --------------^--- Inserted ';'
108 * E 2 - All variables must be declared in one var part
109 * E 5 - Include filename must end in .i
110 * Mon Apr 21 15:56 1980 test.h:
111 * 2 begin
112 * e ------^--- Inserted ';'
113 * Mon Apr 21 16:06 1980 test.p:
114 * E 2 - Function type must be specified
115 * 6 procedure foo(var x:real);
116 * e ------^--- Inserted ';'
117 * In function bletch:
118 * E - No assignment to the function variable
119 * w - variable x is never used
120 * E 6 - foo is already defined in this block
121 * In procedure foo:
122 * w - variable x is neither used nor set
123 * 9 z : = 23;
124 * E --------------^--- Undefined variable
125 * 10 y = [1];
126 * e ----------------^--- Inserted ':'
127 * 13 z := 345.;
128 * e -----------------------^--- Digits required after decimal point
129 * E 10 - Constant set involved in non set context
130 * E 11 - Type clash: real is incompatible with integer
131 * ... Type of expression clashed with type of variable in assignment
132 * E 12 - Parameter type not identical to type of var parameter x of foo
133 * In program mung:
134 * w - variable y is never used
135 * w - type foo is never used
136 * w - function bletch is never used
137 * E - z undefined on lines 9 13
138 */
139 static char *Months[] = {
140 "Jan", "Feb", "Mar", "Apr", "May", "Jun",
141 "Jul", "Aug", "Sep", "Oct","Nov", "Dec",
142 0
143 };
144 static char *Days[] = {
145 "Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat", 0
146 };
147 static char *Piroutines[] = {
148 "program", "function", "procedure", 0
149 };
150
151
152 static boolean structured, multiple;
153
154 static char *pi_Endmatched[] = {"End", "matched"};
155 static char *pi_Inserted[] = {"Inserted", "keyword", "end", "matching"};
156
157 static char *pi_multiple[] = {"Mutiply", "defined", "label", "in", "case,", "line"};
158 static char *pi_structured[] = {"is", "into", "a", "structured", "statement"};
159
160 static char *pi_und1[] = {"undefined", "on", "line"};
161 static char *pi_und2[] = {"undefined", "on", "lines"};
162 static char *pi_imp1[] = {"improperly", "used", "on", "line"};
163 static char *pi_imp2[] = {"improperly", "used", "on", "lines"};
164
165 static boolean
166 alldigits(char *string)
167 {
168 for (; *string && isdigit((unsigned char)*string); string++)
169 continue;
170 return(*string == '\0');
171 }
172
173 static boolean
174 instringset(char *member, char **set)
175 {
176 for(; *set; set++){
177 if (strcmp(*set, member) == 0)
178 return(TRUE);
179 }
180 return(FALSE);
181 }
182
183 static boolean
184 isdateformat(int wordc, char **wordv)
185 {
186 return(
187 (wordc == 5)
188 && (instringset(wordv[0], Days))
189 && (instringset(wordv[1], Months))
190 && (alldigits(wordv[2]))
191 && (alldigits(wordv[4])) );
192 }
193
194 static boolean
195 piptr(char *string)
196 {
197 if (*string != '-')
198 return(FALSE);
199 while (*string && *string == '-')
200 string++;
201 if (*string != '^')
202 return(FALSE);
203 string++;
204 while (*string && *string == '-')
205 string++;
206 return(*string == '\0');
207 }
208
209 extern int wordc;
210 extern char **wordv;
211
212 Errorclass
213 pi(void)
214 {
215 char **nwordv;
216
217 nwordv = NULL;
218 if (wordc < 2)
219 return (C_UNKNOWN);
220 if ( ( strlen(wordv[1]) == 1)
221 && ( (wordv[1][0] == 'e') || (wordv[1][0] == 'E') )
222 && ( piptr(wordv[2]) )
223 ) {
224 boolean longpiptr = 0;
225 /*
226 * We have recognized a first pass error of the form:
227 * letter ------^---- message
228 *
229 * turn into an error message of the form:
230 *
231 * file line 'pascal errortype' letter \n |---- message
232 * or of the form:
233 * file line letter |---- message
234 * when there are strlen("(*[pi]") or more
235 * preceding '-' on the error pointer.
236 *
237 * Where the | is intended to be a down arrow, so that
238 * the pi error messages can be inserted above the
239 * line in error, instead of below. (All of the other
240 * languages put their messages before the source line,
241 * instead of after it as does pi.)
242 *
243 * where the pointer to the error has been truncated
244 * by 6 characters to account for the fact that
245 * the pointer points into a tab preceded input line.
246 */
247 language = INPI;
248 (void)substitute(wordv[2], '^', '|');
249 longpiptr = position(wordv[2],'|') > (6+8);
250 nwordv = wordvsplice(longpiptr ? 2 : 4, wordc, wordv+1);
251 nwordv[0] = strdup(currentfilename);
252 nwordv[1] = strdup(c_linenumber);
253 if (!longpiptr){
254 nwordv[2] = "pascal errortype";
255 nwordv[3] = wordv[1];
256 nwordv[4] = strdup("%%%\n");
257 if (strlen(nwordv[5]) > (8-2)) /* this is the pointer */
258 nwordv[5] += (8-2); /* bump over 6 characters */
259 }
260 wordv = nwordv - 1; /* convert to 1 based */
261 wordc += longpiptr ? 2 : 4;
262 return(C_TRUE);
263 }
264 if ( (wordc >= 4)
265 && (strlen(wordv[1]) == 1)
266 && ( (*wordv[1] == 'E') || (*wordv[1] == 'w') || (*wordv[1] == 'e') )
267 && (alldigits(wordv[2]))
268 && (strlen(wordv[3]) == 1)
269 && (wordv[3][0] == '-')
270 ){
271 /*
272 * Message of the form: letter linenumber - message
273 * Turn into form: filename linenumber letter - message
274 */
275 language = INPI;
276 nwordv = wordvsplice(1, wordc, wordv + 1);
277 nwordv[0] = strdup(currentfilename);
278 nwordv[1] = wordv[2];
279 nwordv[2] = wordv[1];
280 c_linenumber = wordv[2];
281 wordc += 1;
282 wordv = nwordv - 1;
283 return(C_TRUE);
284 }
285 if ( (wordc >= 3)
286 && (strlen(wordv[1]) == 1)
287 && ( (*(wordv[1]) == 'E') || (*(wordv[1]) == 'w') || (*(wordv[1]) == 'e') )
288 && (strlen(wordv[2]) == 1)
289 && (wordv[2][0] == '-')
290 ) {
291 /*
292 * Message of the form: letter - message
293 * This happens only when we are traversing the tree
294 * during the second pass of pi, and discover semantic
295 * errors.
296 *
297 * We have already (presumably) saved the header message
298 * and can now construct a nulled error message for the
299 * current file.
300 *
301 * Turns into a message of the form:
302 * filename (header) letter - message
303 *
304 * First, see if it is a message referring to more than
305 * one line number. Only of the form:
306 * %s undefined on line%s
307 * %s improperly used on line%s
308 */
309 boolean undefined = 0;
310 int wordindex;
311
312 language = INPI;
313 if ( (undefined = (wordvcmp(wordv+2, 3, pi_und1) == 0) )
314 || (undefined = (wordvcmp(wordv+2, 3, pi_und2) == 0) )
315 || (wordvcmp(wordv+2, 4, pi_imp1) == 0)
316 || (wordvcmp(wordv+2, 4, pi_imp2) == 0)
317 ){
318 for (wordindex = undefined ? 5 : 6; wordindex <= wordc;
319 wordindex++){
320 if (nwordv) {
321 free(nwordv[0]);
322 free(nwordv);
323 }
324 nwordv = wordvsplice(2, undefined ? 2 : 3, wordv+1);
325 nwordv[0] = strdup(currentfilename);
326 nwordv[1] = wordv[wordindex];
327 if (wordindex != wordc)
328 erroradd(undefined ? 4 : 5, nwordv,
329 C_TRUE, C_UNKNOWN);
330 }
331 wordc = undefined ? 4 : 5;
332 wordv = nwordv - 1;
333 return(C_TRUE);
334 }
335
336 nwordv = wordvsplice(1+3, wordc, wordv+1);
337 nwordv[0] = strdup(currentfilename);
338 nwordv[1] = strdup(c_header[0]);
339 nwordv[2] = strdup(c_header[1]);
340 nwordv[3] = strdup(c_header[2]);
341 wordv = nwordv - 1;
342 wordc += 1 + 3;
343 return(C_THISFILE);
344 }
345 if (strcmp(wordv[1], "...") == 0){
346 /*
347 * have a continuation error message
348 * of the form: ... message
349 * Turn into form : filename linenumber message
350 */
351 language = INPI;
352 nwordv = wordvsplice(1, wordc, wordv+1);
353 nwordv[0] = strdup(currentfilename);
354 nwordv[1] = strdup(c_linenumber);
355 wordv = nwordv - 1;
356 wordc += 1;
357 return(C_TRUE);
358 }
359 if( (wordc == 6)
360 && (lastchar(wordv[6]) == ':')
361 && (isdateformat(5, wordv + 1))
362 ){
363 /*
364 * Have message that tells us we have changed files
365 */
366 language = INPI;
367 currentfilename = strdup(wordv[6]);
368 clob_last(currentfilename, '\0');
369 return(C_SYNC);
370 }
371 if( (wordc == 3)
372 && (strcmp(wordv[1], "In") == 0)
373 && (lastchar(wordv[3]) == ':')
374 && (instringset(wordv[2], Piroutines))
375 ) {
376 language = INPI;
377 c_header = wordvsplice(0, wordc, wordv+1);
378 return(C_SYNC);
379 }
380 /*
381 * now, check for just the line number followed by the text
382 */
383 if (alldigits(wordv[1])){
384 language = INPI;
385 c_linenumber = wordv[1];
386 return(C_IGNORE);
387 }
388 /*
389 * Attempt to match messages refering to a line number
390 *
391 * Multiply defined label in case, lines %d and %d
392 * Goto %s from line %d is into a structured statement
393 * End matched %s on line %d
394 * Inserted keyword end matching %s on line %d
395 */
396 multiple = structured = 0;
397 if (
398 ( (wordc == 6) && (wordvcmp(wordv+1, 2, pi_Endmatched) == 0))
399 || ( (wordc == 8) && (wordvcmp(wordv+1, 4, pi_Inserted) == 0))
400 || ( multiple = ((wordc == 9) && (wordvcmp(wordv+1,6, pi_multiple) == 0) ) )
401 || ( structured = ((wordc == 10) && (wordvcmp(wordv+6,5, pi_structured) == 0 ) ))
402 ){
403 language = INPI;
404 nwordv = wordvsplice(2, wordc, wordv+1);
405 nwordv[0] = strdup(currentfilename);
406 nwordv[1] = structured ? wordv [5] : wordv[wordc];
407 wordc += 2;
408 wordv = nwordv - 1;
409 if (!multiple)
410 return(C_TRUE);
411 erroradd(wordc, nwordv, C_TRUE, C_UNKNOWN);
412 nwordv = wordvsplice(0, wordc, nwordv);
413 nwordv[1] = wordv[wordc - 2];
414 return(C_TRUE);
415 }
416 return(C_UNKNOWN);
417 }
418