Annotation of src/usr.bin/awk/FIXES, Revision 1.21
1.21 ! millert 1: /* $OpenBSD: FIXES,v 1.20 2020/06/10 21:00:31 millert Exp $ */
1.1 tholo 2: /****************************************************************
1.6 kstailey 3: Copyright (C) Lucent Technologies 1997
1.1 tholo 4: All Rights Reserved
5:
6: Permission to use, copy, modify, and distribute this software and
7: its documentation for any purpose and without fee is hereby
8: granted, provided that the above copyright notice appear in all
9: copies and that both that the copyright notice and this
10: permission notice and warranty disclaimer appear in supporting
1.6 kstailey 11: documentation, and that the name Lucent Technologies or any of
12: its entities not be used in advertising or publicity pertaining
13: to distribution of the software without specific, written prior
14: permission.
15:
16: LUCENT DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
17: INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS.
18: IN NO EVENT SHALL LUCENT OR ANY OF ITS ENTITIES BE LIABLE FOR ANY
19: SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
20: WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER
21: IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
22: ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
23: THIS SOFTWARE.
1.1 tholo 24: ****************************************************************/
25:
26: This file lists all bug fixes, changes, etc., made since the AWK book
27: was sent to the printers in August, 1987.
1.21 ! millert 28:
! 29: Oct 25, 2018:
! 30: Added test in maketab.c to prevent generating a proctab entry
! 31: for YYSTYPE_IS_DEFINED. It was harmless but some gcc settings
! 32: generated a warning message. Thanks to Nan Xiao for report.
1.20 millert 33:
34: Aug 27, 2018:
35: Disallow '$' in printf formats; arguments evaluated in order
36: and printed in order.
37:
38: Added some casts to silence warnings on debugging printfs.
39: (Thanks, Arnold.)
1.18 millert 40:
1.19 millert 41: Aug 23, 2018:
42: A long list of fixes courtesy of Arnold Robbins,
43: to whom profound thanks.
44:
45: 1. ofs-rebuild: OFS value used to rebuild the record was incorrect.
46: Fixed August 19, 2014. Revised fix August 2018.
47:
48: 2. system-status: Instead of a floating-point division by 256, use
49: the wait(2) macros to create a reasonable exit status.
50: Fixed March 12, 2016.
51:
52: 3. space: Use provided xisblank() function instead of ispace() for
53: matching [[:blank:]].
54:
55: 4. a-format: Add POSIX standard %a and %A to supported formats. Check
56: at runtime that this format is available.
57:
58: 5. decr-NF: Decrementing NF did not change $0. This is a decades-old
59: bug. There are interactions with the old and new value of OFS as well.
60: Most of the fix came from the NetBSD awk.
61:
62: 6. string-conv: String conversions of scalars were sticky. Once a
63: conversion to string happened, even with OFMT, that value was used until
64: a new numeric value was assigned, even if OFMT differed from CONVFMT,
65: and also if CONVFMT changed.
66:
67: 7. unary-plus: Unary plus on a string constant returned the string.
68: Instead, it should convert the value to numeric and give that value.
69:
70: Also added Arnold's tests for these to awktest.tar as T.arnold.
71:
72: Aug 15, 2018:
73: fixed mangled awktest.tar (thanks, Arnold), posted all
74: current (very minor) fixes to github / onetrueawk
75:
1.18 millert 76: Jun 7, 2018:
77: (yes, a long layoff)
78: Updated some broken tests (beebe.tar, T.lilly)
79: [thanks to Arnold Robbins]
80:
81: Mar 26, 2015:
82: buffer overflow in error reporting; thanks to tobias ulmer
83: and john-mark gurney for spotting it and the fix.
84:
85: Feb 4, 2013:
86: cleaned up a handful of tests that didn't seem to actually
87: test for correct behavior: T.latin1, T.gawk.
88:
89: Jan 5, 2013:
90: added ,NULL initializer to static Cells in run.c; not really
91: needed but cleaner. Thanks to Michael Bombardieri.
1.17 millert 92:
93: Dec 20, 2012:
94: fiddled makefile to get correct yacc and bison flags. pick yacc
95: (linux) or bison (mac) as necessary.
96:
97: added __attribute__((__noreturn__)) to a couple of lines in
98: proto.h, to silence someone's enthusiastic checker.
99:
100: fixed obscure call by value bug in split(a[1],a) reported on
101: 9fans. the management of temporary values is just a mess; i
102: took a shortcut by making an extra string copy. thanks
103: to paul patience and arnold robbins for passing it on and for
104: proposed patches.
105:
106: tiny fiddle in setfval to eliminate -0 results in T.expr, which
107: has irritated me for 20+ years.
1.16 millert 108:
109: Aug 10, 2011:
110: another fix to avoid core dump with delete(ARGV); again, many thanks
111: to ruslan ermilov.
112:
113: Aug 7, 2011:
114: split(s, a, //) now behaves the same as split(s, a, "")
115:
116: Jun 12, 2011:
117: /pat/, \n /pat/ {...} is now legal, though bad style to use.
118:
119: added checks to new -v code that permits -vnospace; thanks to
120: ruslan ermilov for spotting this and providing the patch.
121:
122: removed fixed limit on number of open files; thanks to aleksey
123: cheusov and christos zoulos.
124:
125: fixed day 1 bug that resurrected deleted elements of ARGV when
126: used as filenames (in lib.c).
127:
128: minor type fiddles to make gcc -Wall -pedantic happier (but not
129: totally so); turned on -fno-strict-aliasing in makefile.
130:
131: May 6, 2011:
132: added #ifdef for isblank.
133: now allows -ffoo as well as -f foo arguments.
134: (thanks, ruslan)
135:
136: May 1, 2011:
137: after advice from todd miller, kevin lo, ruslan ermilov,
138: and arnold robbins, changed srand() to return the previous
139: seed (which is 1 on the first call of srand). the seed is
140: an Awkfloat internally though converted to unsigned int to
141: pass to the library srand(). thanks, everyone.
142:
143: fixed a subtle (and i hope low-probability) overflow error
144: in fldbld, by adding space for one extra \0. thanks to
145: robert bassett for spotting this one and providing a fix.
146:
147: removed the files related to compilation on windows. i no
148: longer have anything like a current windows environment, so
149: i can't test any of it.
1.15 millert 150:
151: May 23, 2010:
152: fixed long-standing overflow bug in run.c; many thanks to
153: nelson beebe for spotting it and providing the fix.
154:
155: fixed bug that didn't parse -vd=1 properly; thanks to santiago
156: vila for spotting it.
157:
158: Feb 8, 2010:
159: i give up. replaced isblank with isspace in b.c; there are
160: no consistent header files.
161:
162: Nov 26, 2009:
163: fixed a long-standing issue with when FS takes effect. a
164: change to FS is now noticed immediately for subsequent splits.
165:
166: changed the name getline() to awkgetline() to avoid yet another
167: name conflict somewhere.
168:
169: Feb 11, 2009:
170: temporarily for now defined HAS_ISBLANK, since that seems to
171: be the best way through the thicket. isblank arrived in C99,
172: but seems to be arriving at different systems at different
173: times.
174:
175: Oct 8, 2008:
176: fixed typo in b.c that set tmpvec wrongly. no one had ever
177: run into the problem, apparently. thanks to alistair crooks.
1.14 millert 178:
179: Oct 23, 2007:
180: minor fix in lib.c: increase inputFS to 100, change malloc
181: for fields to n+1.
182:
183: fixed memory fault caused by out of order test in setsval.
184:
185: thanks to david o'brien, freebsd, for both fixes.
186:
187: May 1, 2007:
188: fiddle in makefile to fix for BSD make; thanks to igor sobrado.
189:
190: Mar 31, 2007:
191: fixed some null pointer refs calling adjbuf.
192:
193: Feb 21, 2007:
194: fixed a bug in matching the null RE in sub and gsub. thanks to al aho
195: who actually did the fix (in b.c), and to wolfgang seeberg for finding
196: it and providing a very compact test case.
197:
198: fixed quotation in b.c; thanks to Hal Pratt and the Princeton Dante
199: Project.
200:
201: removed some no-effect asserts in run.c.
202:
203: fiddled maketab.c to not complain about bison-generated values.
204:
205: removed the obsolete -V argument; fixed --version to print the
206: version and exit.
207:
208: fixed wording and an outright error in the usage message; thanks to igor
209: sobrado and jason mcintyre.
210:
211: fixed a bug in -d that caused core dump if no program followed.
212:
213: Jan 1, 2007:
214: dropped mac.code from makefile; there are few non-MacOSX
215: mac's these days.
216:
217: Jan 17, 2006:
218: system() not flagged as unsafe in the unadvertised -safe option.
219: found it while enhancing tests before shipping the ;login: article.
220: practice what you preach.
221:
222: removed the 9-years-obsolete -mr and -mf flags.
223:
224: added -version and --version options.
225:
226: core dump on linux with BEGIN {nextfile}, now fixed.
227:
228: removed some #ifdef's in run.c and lex.c that appear to no
229: longer be necessary.
230:
231: Apr 24, 2005:
232: modified lib.c so that values of $0 et al are preserved in the END
233: block, apparently as required by posix. thanks to havard eidnes
234: for the report and code.
235:
236: Jan 14, 2005:
237: fixed infinite loop in parsing, originally found by brian tsang.
238: thanks to arnold robbins for a suggestion that started me
239: rethinking it.
240:
241: Dec 31, 2004:
242: prevent overflow of -f array in main, head off potential error in
243: call of SYNTAX(), test malloc return in lib.c, all with thanks to
244: todd miller.
1.13 millert 245:
246: Dec 22, 2004:
247: cranked up size of NCHARS; coverity thinks it can be overrun with
248: smaller size, and i think that's right. added some assertions to b.c
249: to catch places where it might overrun. the RE code is still fragile.
250:
251: Dec 5, 2004:
252: fixed a couple of overflow problems with ridiculous field numbers:
253: e.g., print $(2^32-1). thanks to ruslan ermilov, giorgos keramidas
254: and david o'brien at freebsd.org for patches. this really should
255: be re-done from scratch.
256:
257: Nov 21, 2004:
258: fixed another 25-year-old RE bug, in split. it's another failure
259: to (re-)initialize. thanks to steve fisher for spotting this and
260: providing a good test case.
261:
262: Nov 22, 2003:
263: fixed a bug in regular expressions that dates (so help me) from 1977;
264: it's been there from the beginning. an anchored longest match that
265: was longer than the number of states triggered a failure to initialize
266: the machine properly. many thanks to moinak ghosh for not only finding
267: this one but for providing a fix, in some of the most mysterious
268: code known to man.
269:
270: fixed a storage leak in call() that appears to have been there since
271: 1983 or so -- a function without an explicit return that assigns a
272: string to a parameter leaked a Cell. thanks to moinak ghosh for
273: spotting this very subtle one.
274:
275: Jul 31, 2003:
276: fixed, thanks to andrey chernov and ruslan ermilov, a bug in lex.c
277: that mis-handled the character 255 in input. (it was being compared
278: to EOF with a signed comparison.)
279:
280: Jul 29, 2003:
281: fixed (i think) the long-standing botch that included the beginning of
282: line state ^ for RE's in the set of valid characters; this led to a
283: variety of odd problems, including failure to properly match certain
284: regular expressions in non-US locales. thanks to ruslan for keeping
285: at this one.
286:
287: Jul 28, 2003:
288: n-th try at getting internationalization right, with thanks to volker
289: kiefel, arnold robbins and ruslan ermilov for advice, though they
290: should not be blamed for the outcome. according to posix, "." is the
291: radix character in programs and command line arguments regardless of
292: the locale; otherwise, the locale should prevail for input and output
293: of numbers. so it's intended to work that way.
294:
295: i have rescinded the attempt to use strcoll in expanding shorthands in
296: regular expressions (cclenter). its properties are much too
297: surprising; for example [a-c] matches aAbBc in locale en_US but abBcC
298: in locale fr_CA. i can see how this might arise by implementation
299: but i cannot explain it to a human user. (this behavior can be seen
300: in gawk as well; we're leaning on the same library.)
301:
302: the issue appears to be that strcoll is meant for sorting, where
303: merging upper and lower case may make sense (though note that unix
304: sort does not do this by default either). it is not appropriate
305: for regular expressions, where the goal is to match specific
306: patterns of characters. in any case, the notations [:lower:], etc.,
307: are available in awk, and they are more likely to work correctly in
308: most locales.
309:
310: a moratorium is hereby declared on internationalization changes.
311: i apologize to friends and colleagues in other parts of the world.
312: i would truly like to get this "right", but i don't know what
313: that is, and i do not want to keep making changes until it's clear.
314:
315: Jul 4, 2003:
316: fixed bug that permitted non-terminated RE, as in "awk /x".
317:
318: Jun 1, 2003:
319: subtle change to split: if source is empty, number of elems
320: is always 0 and the array is not set.
321:
322: Mar 21, 2003:
323: added some parens to isblank, in another attempt to make things
324: internationally portable.
325:
326: Mar 14, 2003:
327: the internationalization changes, somewhat modified, are now
328: reinstated. in theory awk will now do character comparisons
329: and case conversions in national language, but "." will always
330: be the decimal point separator on input and output regardless
331: of national language. isblank(){} has an #ifndef.
332:
333: this no longer compiles on windows: LC_MESSAGES isn't defined
334: in vc6++.
335:
336: fixed subtle behavior in field and record splitting: if FS is
337: a single character and RS is not empty, \n is NOT a separator.
338: this tortuous reading is found in the awk book; behavior now
339: matches gawk and mawk.
1.11 millert 340:
341: Dec 13, 2002:
342: for the moment, the internationalization changes of nov 29 are
343: rolled back -- programs like x = 1.2 don't work in some locales,
344: because the parser is expecting x = 1,2. until i understand this
345: better, this will have to wait.
346:
347: Nov 29, 2002:
348: modified b.c (with tiny changes in main and run) to support
349: locales, using strcoll and iswhatever tests for posix character
350: classes. thanks to ruslan ermilov (ru@freebsd.org) for code.
351: the function isblank doesn't seem to have propagated to any
352: header file near me, so it's there explicitly. not properly
353: tested on non-ascii character sets by me.
354:
355: Jun 28, 2002:
356: modified run/format() and tran/getsval() to do a slightly better
357: job on using OFMT for output from print and CONVFMT for other
358: number->string conversions, as promised by posix and done by
359: gawk and mawk. there are still places where it doesn't work
360: right if CONVFMT is changed; by then the STR attribute of the
361: variable has been irrevocably set. thanks to arnold robbins for
362: code and examples.
363:
364: fixed subtle bug in format that could get core dump. thanks to
365: Jaromir Dolecek <jdolecek@NetBSD.org> for finding and fixing.
366: minor cleanup in run.c / format() at the same time.
367:
368: added some tests for null pointers to debugging printf's, which
369: were never intended for external consumption. thanks to dave
370: kerns (dkerns@lucent.com) for pointing this out.
371:
372: GNU compatibility: an empty regexp matches anything (thanks to
373: dag-erling smorgrav, des@ofug.org). subject to reversion if
374: this does more harm than good.
375:
376: pervasive small changes to make things more const-correct, as
377: reported by gcc's -Wwrite-strings. as it says in the gcc manual,
378: this may be more nuisance than useful. provoked by a suggestion
379: and code from arnaud desitter, arnaud@nimbus.geog.ox.ac.uk
380:
381: minor documentation changes to note that this now compiles out
382: of the box on Mac OS X.
383:
384: Feb 10, 2002:
385: changed types in posix chars structure to quiet solaris cc.
386:
387: Jan 1, 2002:
388: fflush() or fflush("") flushes all files and pipes.
389:
390: length(arrayname) returns number of elements; thanks to
391: arnold robbins for suggestion.
392:
393: added a makefile.win to make it easier to build on windows.
394: based on dan allen's buildwin.bat.
395:
396: Nov 16, 2001:
397: added support for posix character class names like [:digit:],
398: which are not exactly shorter than [0-9] and perhaps no more
399: portable. thanks to dag-erling smorgrav for code.
400:
401: Feb 16, 2001:
402: removed -m option; no longer needed, and it was actually
403: broken (noted thanks to volker kiefel).
404:
405: Feb 10, 2001:
406: fixed an appalling bug in gettok: any sequence of digits, +,-, E, e,
407: and period was accepted as a valid number if it started with a period.
408: this would never have happened with the lex version.
409:
410: other 1-character botches, now fixed, include a bare $ and a
411: bare " at the end of the input.
412:
413: Feb 7, 2001:
414: more (const char *) casts in b.c and tran.c to silence warnings.
1.8 millert 415:
1.10 millert 416: Nov 15, 2000:
417: fixed a bug introduced in august 1997 that caused expressions
418: like $f[1] to be syntax errors. thanks to arnold robbins for
419: noticing this and providing a fix.
420:
421: Oct 30, 2000:
422: fixed some nextfile bugs: not handling all cases. thanks to
423: arnold robbins for pointing this out. new regressions added.
424:
425: close() is now a function. it returns whatever the library
426: fclose returns, and -1 for closing a file or pipe that wasn't
427: opened.
428:
429: Sep 24, 2000:
430: permit \n explicitly in character classes; won't work right
431: if comes in as "[\n]" but ok as /[\n]/, because of multiple
432: processing of \'s. thanks to arnold robbins.
433:
434: July 5, 2000:
435: minor fiddles in tran.c to keep compilers happy about uschar.
436: thanks to norman wilson.
437:
438: May 25, 2000:
439: yet another attempt at making 8-bit input work, with another
440: band-aid in b.c (member()), and some (uschar) casts to head
441: off potential errors in subscripts (like isdigit). also
442: changed HAT to NCHARS-2. thanks again to santiago vila.
443:
444: changed maketab.c to ignore apparently out of range definitions
445: instead of halting; new freeBSD generates one. thanks to
446: jon snader <jsnader@ix.netcom.com> for pointing out the problem.
447:
448: May 2, 2000:
449: fixed an 8-bit problem in b.c by making several char*'s into
450: unsigned char*'s. not clear i have them all yet. thanks to
451: Santiago Vila <sanvila@unex.es> for the bug report.
452:
453: Apr 21, 2000:
454: finally found and fixed a memory leak in function call; it's
455: been there since functions were added ~1983. thanks to
456: jon bentley for the test case that found it.
457:
458: added test in envinit to catch environment "variables" with
1.12 otto 459: names beginning with '='; thanks to Berend Hasselman.
1.10 millert 460:
461: Jul 28, 1999:
462: added test in defn() to catch function foo(foo), which
463: otherwise recurses until core dump. thanks to arnold
464: robbins for noticing this.
465:
1.9 millert 466: Jun 20, 1999:
467: added *bp in gettok in lex.c; appears possible to exit function
468: without terminating the string. thanks to russ cox.
469:
470: Jun 2, 1999:
471: added function stdinit() to run to initialize files[] array,
472: in case stdin, etc., are not constants; some compilers care.
473:
474: May 10, 1999:
475: replaced the ERROR ... FATAL, etc., macros with functions
476: based on vprintf, to avoid problems caused by overrunning
477: fixed-size errbuf array. thanks to ralph corderoy for the
478: impetus, and for pointing out a string termination bug in
479: qstring as well.
480:
481: Apr 21, 1999:
482: fixed bug that caused occasional core dumps with commandline
483: variable with value ending in \. (thanks to nelson beebe for
484: the test case.)
485:
1.8 millert 486: Apr 16, 1999:
487: with code kindly provided by Bruce Lilly, awk now parses
488: /=/ and similar constructs more sensibly in more places.
1.10 millert 489: Bruce also provided some helpful test cases.
1.8 millert 490:
491: Apr 5, 1999:
492: changed true/false to True/False in run.c to make it
493: easier to compile with C++. Added some casts on malloc
494: and realloc to be honest about casts; ditto. changed
495: ltype int to long in struct rrow to reduce some 64-bit
496: complaints; other changes scattered throughout for the
497: same purpose. thanks to Nelson Beebe for these portability
498: improvements.
499:
500: removed some horrible pointer-int casting in b.c and elsewhere
501: by adding ptoi and itonp to localize the casts, which are
502: all benign. fixed one incipient bug that showed up on sgi
503: in 64-bit mode.
504:
505: reset lineno for new source file; include filename in error
506: message. also fixed line number error in continuation lines.
507: (thanks to Nelson Beebe for both of these.)
508:
509: Mar 24, 1999:
510: Nelson Beebe notes that irix 5.3 yacc dies with a bogus
511: error; use a newer version or switch to bison, since sgi
512: is unlikely to fix it.
1.1 tholo 513:
1.7 millert 514: Mar 5, 1999:
1.9 millert 515: changed isnumber to is_number to avoid the problem caused by
516: versions of ctype.h that include the name isnumber.
1.7 millert 517:
518: distribution now includes a script for building on a Mac,
519: thanks to Dan Allen.
520:
521: Feb 20, 1999:
522: fixed memory leaks in run.c (call) and tran.c (setfval).
523: thanks to Stephen Nutt for finding these and providing the fixes.
524:
525: Jan 13, 1999:
526: replaced srand argument by (unsigned int) in run.c;
527: avoids problem on Mac and potentially on Unix & Windows.
528: thanks to Dan Allen.
529:
530: added a few (int) casts to silence useless compiler warnings.
531: e.g., errorflag= in run.c jump().
532:
533: added proctab.c to the bundle outout; one less thing
534: to have to compile out of the box.
535:
536: added calls to _popen and _pclose to the win95 stub for
537: pipes (thanks to Steve Adams for this helpful suggestion).
538: seems to work, though properties are not well understood
539: by me, and it appears that under some circumstances the
540: pipe output is truncated. Be careful.
541:
542: Oct 19, 1998:
543: fixed a couple of bugs in getrec: could fail to update $0
544: after a getline var; because inputFS wasn't initialized,
545: could split $0 on every character, a misleading diversion.
546:
547: fixed caching bug in makedfa: LRU was actually removing
548: least often used.
549:
550: thanks to ross ridge for finding these, and for providing
551: great bug reports.
552:
553: May 12, 1998:
554: fixed potential bug in readrec: might fail to update record
555: pointer after growing. thanks to dan levy for spotting this
556: and suggesting the fix.
557:
558: Mar 12, 1998:
559: added -V to print version number and die.
1.19 millert 560:
561: [notify dave kerns, dkerns@dacsoup.ih.lucent.com]
1.7 millert 562:
563: Feb 11, 1998:
564: subtle silent bug in lex.c: if the program ended with a number
565: longer than 1 digit, part of the input would be pushed back and
566: parsed again because token buffer wasn't terminated right.
567: example: awk 'length($0) > 10'. blush. at least i found it
568: myself.
569:
570: Aug 31, 1997:
571: s/adelete/awkdelete/: SGI uses this in malloc.h.
572: thanks to nelson beebe for pointing this one out.
573:
574: Aug 21, 1997:
575: fixed some bugs in sub and gsub when replacement includes \\.
576: this is a dark, horrible corner, but at least now i believe that
577: the behavior is the same as gawk and the intended posix standard.
578: thanks to arnold robbins for advice here.
579:
580: Aug 9, 1997:
581: somewhat regretfully, replaced the ancient lex-based lexical
582: analyzer with one written in C. it's longer, generates less code,
583: and more portable; the old one depended too much on mysterious
584: properties of lex that were not preserved in other environments.
585: in theory these recognize the same language.
586:
587: now using strtod to test whether a string is a number, instead of
588: the convoluted original function. should be more portable and
589: reliable if strtod is implemented right.
590:
591: removed now-pointless optimization in makefile that tries to avoid
592: recompilation when awkgram.y is changed but symbols are not.
593:
594: removed most fixed-size arrays, though a handful remain, some
595: of which are unchecked. you have been warned.
596:
597: Aug 4, 1997:
598: with some trepidation, replaced the ancient code that managed
599: fields and $0 in fixed-size arrays with arrays that grow on
600: demand. there is still some tension between trying to make this
601: run fast and making it clean; not sure it's right yet.
602:
603: the ill-conceived -mr and -mf arguments are now useful only
604: for debugging. previous dynamic string code removed.
605:
606: numerous other minor cleanups along the way.
1.1 tholo 607:
1.7 millert 608: Jul 30, 1997:
609: using code provided by dan levy (to whom profuse thanks), replaced
610: fixed-size arrays and awkward kludges by a fairly uniform mechanism
611: to grow arrays as needed for printf, sub, gsub, etc.
1.1 tholo 612:
1.7 millert 613: Jul 23, 1997:
614: falling off the end of a function returns "" and 0, not 0.
615: thanks to arnold robbins.
1.1 tholo 616:
1.7 millert 617: Jun 17, 1997:
618: replaced several fixed-size arrays by dynamically-created ones
619: in run.c; added overflow tests to some previously unchecked cases.
620: getline, toupper, tolower.
1.1 tholo 621:
1.7 millert 622: getline code is still broken in that recursive calls may wind
623: up using the same space. [fixed later]
1.1 tholo 624:
1.7 millert 625: increased RECSIZE to 8192 to push problems further over the horizon.
1.1 tholo 626:
1.7 millert 627: added \r to \n as input line separator for programs, not data.
628: damn CRLFs.
1.1 tholo 629:
1.7 millert 630: modified format() to permit explicit printf("%c", 0) to include
631: a null byte in output. thanks to ken stailey for the fix.
1.1 tholo 632:
1.7 millert 633: added a "-safe" argument that disables file output (print >,
634: print >>), process creation (cmd|getline, print |, system), and
635: access to the environment (ENVIRON). this is a first approximation
636: to a "safe" version of awk, but don't rely on it too much. thanks
637: to joan feigenbaum and matt blaze for the inspiration long ago.
1.1 tholo 638:
1.7 millert 639: Jul 8, 1996:
640: fixed long-standing bug in sub, gsub(/a/, "\\\\&"); thanks to
641: ralph corderoy.
1.1 tholo 642:
1.7 millert 643: Jun 29, 1996:
644: fixed awful bug in new field splitting; didn't get all the places
645: where input was done.
1.1 tholo 646:
1.7 millert 647: Jun 28, 1996:
648: changed field-splitting to conform to posix definition: fields are
649: split using the value of FS at the time of input; it used to be
650: the value when the field or NF was first referred to, a much less
651: predictable definition. thanks to arnold robbins for encouragement
652: to do the right thing.
1.1 tholo 653:
1.7 millert 654: May 28, 1996:
655: fixed appalling but apparently unimportant bug in parsing octal
656: numbers in reg exprs.
1.1 tholo 657:
1.7 millert 658: explicit hex in reg exprs now limited to 2 chars: \xa, \xaa.
1.1 tholo 659:
1.7 millert 660: May 27, 1996:
661: cleaned up some declarations so gcc -Wall is now almost silent.
1.1 tholo 662:
1.7 millert 663: makefile now includes backup copies of ytab.c and lexyy.c in case
664: one makes before looking; it also avoids recreating lexyy.c unless
665: really needed.
1.1 tholo 666:
1.7 millert 667: s/aprintf/awkprint, s/asprintf/awksprintf/ to avoid some name clashes
668: with unwisely-written header files.
1.1 tholo 669:
1.7 millert 670: thanks to jeffrey friedl for several of these.
1.1 tholo 671:
1.7 millert 672: May 26, 1996:
673: an attempt to rationalize the (unsigned) char issue. almost all
674: instances of unsigned char have been removed; the handful of places
675: in b.c where chars are used as table indices have been hand-crafted.
676: added some latin-1 tests to the regression, but i'm not confident;
677: none of my compilers seem to care much. thanks to nelson beebe for
678: pointing out some others that do care.
1.1 tholo 679:
1.7 millert 680: May 2, 1996:
681: removed all register declarations.
1.1 tholo 682:
1.7 millert 683: enhanced split(), as in gawk, etc: split(s, a, "") splits s into
684: a[1]...a[length(s)] with each character a single element.
1.1 tholo 685:
1.7 millert 686: made the same changes for field-splitting if FS is "".
1.1 tholo 687:
1.7 millert 688: added nextfile, as in gawk: causes immediate advance to next
689: input file. (thanks to arnold robbins for inspiration and code).
1.1 tholo 690:
1.7 millert 691: small fixes to regexpr code: can now handle []], [[], and
692: variants; [] is now a syntax error, rather than matching
693: everything; [z-a] is now empty, not z. far from complete
694: or correct, however. (thanks to jeffrey friedl for pointing out
695: some awful behaviors.)
1.1 tholo 696:
1.7 millert 697: Apr 29, 1996:
1.12 otto 698: replaced uchar by uschar everywhere; apparently some compilers
1.7 millert 699: usurp this name and this causes conflicts.
1.1 tholo 700:
1.7 millert 701: fixed call to time in run.c (bltin); arg is time_t *.
1.1 tholo 702:
1.7 millert 703: replaced horrible pointer/long punning in b.c by a legitimate
704: union. should be safer on 64-bit machines and cleaner everywhere.
705: (thanks to nelson beebe for pointing out some of these problems.)
1.1 tholo 706:
1.7 millert 707: replaced nested comments by #if 0...#endif in run.c, lib.c.
1.1 tholo 708:
1.7 millert 709: removed getsval, setsval, execute macros from run.c and lib.c.
710: machines are 100x faster than they were when these macros were
711: first used.
1.1 tholo 712:
1.7 millert 713: revised filenames: awk.g.y => awkgram.y, awk.lx.l => awklex.l,
714: y.tab.[ch] => ytab.[ch], lex.yy.c => lexyy.c, all in the aid of
715: portability to nameless systems.
1.1 tholo 716:
1.7 millert 717: "make bundle" now includes yacc and lex output files for recipients
718: who don't have yacc or lex.
1.1 tholo 719:
1.7 millert 720: Aug 15, 1995:
721: initialized Cells in setsymtab more carefully; some fields
722: were not set. (thanks to purify, all of whose complaints i
723: think i now understand.)
1.1 tholo 724:
1.7 millert 725: fixed at least one error in gsub that looked at -1-th element
726: of an array when substituting for a null match (e.g., $).
1.1 tholo 727:
1.7 millert 728: delete arrayname is now legal; it clears the elements but leaves
729: the array, which may not be the right behavior.
1.1 tholo 730:
1.7 millert 731: modified makefile: my current make can't cope with the test used
732: to avoid unnecessary yacc invocations.
1.1 tholo 733:
1.7 millert 734: Jul 17, 1995:
735: added dynamically growing strings to awk.lx.l and b.c
736: to permit regular expressions to be much bigger.
737: the state arrays can still overflow.
1.1 tholo 738:
1.7 millert 739: Aug 24, 1994:
740: detect duplicate arguments in function definitions (mdm).
1.1 tholo 741:
1.7 millert 742: May 11, 1994:
743: trivial fix to printf to limit string size in sub().
1.1 tholo 744:
1.7 millert 745: Apr 22, 1994:
746: fixed yet another subtle self-assignment problem:
747: $1 = $2; $1 = $1 clobbered $1.
1.1 tholo 748:
1.7 millert 749: Regression tests now use private echo, to avoid quoting problems.
1.1 tholo 750:
1.7 millert 751: Feb 2, 1994:
752: changed error() to print line number as %d, not %g.
1.1 tholo 753:
1.7 millert 754: Jul 23, 1993:
755: cosmetic changes: increased sizes of some arrays,
756: reworded some error messages.
1.1 tholo 757:
1.7 millert 758: added CONVFMT as in posix (just replaced OFMT in getsval)
1.1 tholo 759:
1.7 millert 760: FILENAME is now "" until the first thing that causes a file
761: to be opened.
1.1 tholo 762:
1.7 millert 763: Nov 28, 1992:
764: deleted yyunput and yyoutput from proto.h;
765: different versions of lex give these different declarations.
1.1 tholo 766:
1.7 millert 767: May 31, 1992:
768: added -mr N and -mf N options: more record and fields.
769: these really ought to adjust automatically.
1.1 tholo 770:
1.7 millert 771: cleaned up some error messages; "out of space" now means
772: malloc returned NULL in all cases.
1.1 tholo 773:
1.7 millert 774: changed rehash so that if it runs out, it just returns;
775: things will continue to run slow, but maybe a bit longer.
1.1 tholo 776:
1.7 millert 777: Apr 24, 1992:
778: remove redundant close of stdin when using -f -.
1.1 tholo 779:
1.7 millert 780: got rid of core dump with -d; awk -d just prints date.
1.1 tholo 781:
1.7 millert 782: Apr 12, 1992:
783: added explicit check for /dev/std(in,out,err) in redirection.
784: unlike gawk, no /dev/fd/n yet.
1.1 tholo 785:
1.7 millert 786: added (file/pipe) builtin. hard to test satisfactorily.
787: not posix.
1.1 tholo 788:
1.7 millert 789: Feb 20, 1992:
790: recompile after abortive changes; should be unchanged.
1.1 tholo 791:
1.7 millert 792: Dec 2, 1991:
793: die-casting time: converted to ansi C, installed that.
1.1 tholo 794:
1.7 millert 795: Nov 30, 1991:
796: fixed storage leak in freefa, failing to recover [N]CCL.
797: thanks to Bill Jones (jones@cs.usask.ca)
1.1 tholo 798:
1.7 millert 799: Nov 19, 1991:
800: use RAND_MAX instead of literal in builtin().
1.1 tholo 801:
1.7 millert 802: Nov 12, 1991:
803: cranked up some fixed-size arrays in b.c, and added a test for
804: overflow in penter. thanks to mark larsen.
1.1 tholo 805:
1.7 millert 806: Sep 24, 1991:
807: increased buffer in gsub. a very crude fix to a general problem.
808: and again on Sep 26.
1.1 tholo 809:
1.7 millert 810: Aug 18, 1991:
811: enforce variable name syntax for commandline variables: has to
812: start with letter or _.
1.1 tholo 813:
1.7 millert 814: Jul 27, 1991:
815: allow newline after ; in for statements.
1.1 tholo 816:
1.7 millert 817: Jul 21, 1991:
818: fixed so that in self-assignment like $1=$1, side effects
819: like recomputing $0 take place. (this is getting subtle.)
1.1 tholo 820:
1.7 millert 821: Jun 30, 1991:
822: better test for detecting too-long output record.
1.1 tholo 823:
1.7 millert 824: Jun 2, 1991:
825: better defense against very long printf strings.
826: made break and continue illegal outside of loops.
1.1 tholo 827:
1.7 millert 828: May 13, 1991:
829: removed extra arg on gettemp, tempfree. minor error message rewording.
1.1 tholo 830:
831: May 6, 1991:
832: fixed silly bug in hex parsing in hexstr().
833: removed an apparently unnecessary test in isnumber().
834: warn about weird printf conversions.
835: fixed unchecked array overwrite in relex().
836:
837: changed for (i in array) to access elements in sorted order.
838: then unchanged it -- it really does run slower in too many cases.
839: left the code in place, commented out.
840:
1.7 millert 841: Feb 10, 1991:
842: check error status on all writes, to avoid banging on full disks.
1.1 tholo 843:
1.7 millert 844: Jan 28, 1991:
845: awk -f - reads the program from stdin.
1.1 tholo 846:
1.7 millert 847: Jan 11, 1991:
848: failed to set numeric state on $0 in cmd|getline context in run.c.
1.1 tholo 849:
1.7 millert 850: Nov 2, 1990:
851: fixed sleazy test for integrality in getsval; use modf.
1.1 tholo 852:
1.7 millert 853: Oct 29, 1990:
854: fixed sleazy buggy code in lib.c that looked (incorrectly) for
855: too long input lines.
1.1 tholo 856:
1.7 millert 857: Oct 14, 1990:
858: fixed the bug on p. 198 in which it couldn't deduce that an
859: argument was an array in some contexts. replaced the error
860: message in intest() by code that damn well makes it an array.
1.1 tholo 861:
1.7 millert 862: Oct 8, 1990:
863: fixed horrible bug: types and values were not preserved in
864: some kinds of self-assignment. (in assign().)
1.1 tholo 865:
1.7 millert 866: Aug 24, 1990:
867: changed NCHARS to 256 to handle 8-bit characters in strings
868: presented to match(), etc.
1.1 tholo 869:
1.7 millert 870: Jun 26, 1990:
871: changed struct rrow (awk.h) to use long instead of int for lval,
872: since cfoll() stores a pointer in it. now works better when int's
873: are smaller than pointers!
1.1 tholo 874:
1.7 millert 875: May 6, 1990:
876: AVA fixed the grammar so that ! is uniformly of the same precedence as
877: unary + and -. This renders illegal some constructs like !x=y, which
878: now has to be parenthesized as !(x=y), and makes others work properly:
879: !x+y is (!x)+y, and x!y is x !y, not two pattern-action statements.
880: (These problems were pointed out by Bob Lenk of Posix.)
1.1 tholo 881:
1.7 millert 882: Added \x to regular expressions (already in strings).
883: Limited octal to octal digits; \8 and \9 are not octal.
884: Centralized the code for parsing escapes in regular expressions.
885: Added a bunch of tests to T.re and T.sub to verify some of this.
1.1 tholo 886:
1.7 millert 887: Feb 9, 1990:
888: fixed null pointer dereference bug in main.c: -F[nothing]. sigh.
1.1 tholo 889:
1.7 millert 890: restored srand behavior: it returns the current seed.
1.1 tholo 891:
1.7 millert 892: Jan 18, 1990:
893: srand now returns previous seed value (0 to start).
1.1 tholo 894:
1.7 millert 895: Jan 5, 1990:
896: fix potential problem in tran.c -- something was freed,
897: then used in freesymtab.
1.1 tholo 898:
1.7 millert 899: Oct 18, 1989:
900: another try to get the max number of open files set with
901: relatively machine-independent code.
1.1 tholo 902:
1.7 millert 903: small fix to input() in case of multiple reads after EOF.
1.1 tholo 904:
1.7 millert 905: Oct 11, 1989:
906: FILENAME is now defined in the BEGIN block -- too many old
907: programs broke.
1.1 tholo 908:
1.7 millert 909: "-" means stdin in getline as well as on the commandline.
1.1 tholo 910:
1.7 millert 911: added a bunch of casts to the code to tell the truth about
912: char * vs. unsigned char *, a right royal pain. added a
913: setlocale call to the front of main, though probably no one
914: has it usefully implemented yet.
1.1 tholo 915:
1.7 millert 916: Aug 24, 1989:
917: removed redundant relational tests against nullnode if parse
918: tree already had a relational at that point.
1.1 tholo 919:
1.7 millert 920: Aug 11, 1989:
921: fixed bug: commandline variable assignment has to look like
922: var=something. (consider the man page for =, in file =.1)
1.1 tholo 923:
1.7 millert 924: changed number of arguments to functions to static arrays
925: to avoid repeated malloc calls.
1.1 tholo 926:
1.7 millert 927: Aug 2, 1989:
928: restored -F (space) separator
1.1 tholo 929:
1.7 millert 930: Jul 30, 1989:
931: added -v x=1 y=2 ... for immediate commandline variable assignment;
932: done before the BEGIN block for sure. they have to precede the
933: program if the program is on the commandline.
934: Modified Aug 2 to require a separate -v for each assignment.
1.1 tholo 935:
1.7 millert 936: Jul 10, 1989:
937: fixed ref-thru-zero bug in environment code in tran.c
1.1 tholo 938:
1.7 millert 939: Jun 23, 1989:
940: add newline to usage message.
1.1 tholo 941:
1.7 millert 942: Jun 14, 1989:
943: added some missing ansi printf conversion letters: %i %X %E %G.
944: no sensible meaning for h or L, so they may not do what one expects.
1.1 tholo 945:
1.7 millert 946: made %* conversions work.
1.1 tholo 947:
1.7 millert 948: changed x^y so that if n is a positive integer, it's done
949: by explicit multiplication, thus achieving maximum accuracy.
950: (this should be done by pow() but it seems not to be locally.)
951: done to x ^= y as well.
1.1 tholo 952:
1.7 millert 953: Jun 4, 1989:
954: ENVIRON array contains environment: if shell variable V=thing,
955: ENVIRON["V"] is "thing"
1.1 tholo 956:
1.7 millert 957: multiple -f arguments permitted. error reporting is naive.
958: (they were permitted before, but only the last was used.)
1.1 tholo 959:
1.7 millert 960: fixed a really stupid botch in the debugging macro dprintf
1.1 tholo 961:
1.7 millert 962: fixed order of evaluation of commandline assignments to match
963: what the book claims: an argument of the form x=e is evaluated
964: at the time it would have been opened if it were a filename (p 63).
965: this invalidates the suggested answer to ex 4-1 (p 195).
1.1 tholo 966:
1.7 millert 967: removed some code that permitted -F (space) fieldseparator,
968: since it didn't quite work right anyway. (restored aug 2)
1.1 tholo 969:
1.7 millert 970: Apr 27, 1989:
971: Line number now accumulated correctly for comment lines.
1.1 tholo 972:
1.7 millert 973: Apr 26, 1989:
974: Debugging output now includes a version date,
975: if one compiles it into the source each time.
1.1 tholo 976:
1.7 millert 977: Apr 9, 1989:
978: Changed grammar to prohibit constants as 3rd arg of sub and gsub;
979: prevents class of overwriting-a-constant errors. (Last one?)
980: This invalidates the "banana" example on page 43 of the book.
1.1 tholo 981:
1.7 millert 982: Added \a ("alert"), \v (vertical tab), \xhhh (hexadecimal),
983: as in ANSI, for strings. Rescinded the sloppiness that permitted
984: non-octal digits in \ooo. Warning: not all compilers and libraries
985: will be able to deal with \x correctly.
1.1 tholo 986:
1.7 millert 987: Jan 9, 1989:
988: Fixed bug that caused tempcell list to contain a duplicate.
989: The fix is kludgy.
1.1 tholo 990:
1.7 millert 991: Dec 17, 1988:
992: Catches some more commandline errors in main.
993: Removed redundant decl of modf in run.c (confuses some compilers).
994: Warning: there's no single declaration of malloc, etc., in awk.h
995: that seems to satisfy all compilers.
1.1 tholo 996:
1.7 millert 997: Dec 7, 1988:
998: Added a bit of code to error printing to avoid printing nulls.
999: (Not clear that it actually would.)
1.1 tholo 1000:
1.7 millert 1001: Nov 27, 1988:
1002: With fear and trembling, modified the grammar to permit
1003: multiple pattern-action statements on one line without
1004: an explicit separator. By definition, this capitulation
1005: to the ghost of ancient implementations remains undefined
1006: and thus subject to change without notice or apology.
1007: DO NOT COUNT ON IT.
1.1 tholo 1008:
1.7 millert 1009: Oct 30, 1988:
1010: Fixed bug in call() that failed to recover storage.
1.1 tholo 1011:
1.7 millert 1012: A warning is now generated if there are more arguments
1013: in the call than in the definition (in lieu of fixing
1014: another storage leak).
1.2 millert 1015:
1.7 millert 1016: Oct 20, 1988:
1017: Fixed %c: if expr is numeric, use numeric value;
1018: otherwise print 1st char of string value. still
1019: doesn't work if the value is 0 -- won't print \0.
1.2 millert 1020:
1.7 millert 1021: Added a few more checks for running out of malloc.
1.2 millert 1022:
1.7 millert 1023: Oct 12, 1988:
1024: Fixed bug in call() that freed local arrays twice.
1.5 kstailey 1025:
1.7 millert 1026: Fixed to handle deletion of non-existent array right;
1027: complains about attempt to delete non-array element.
1.6 kstailey 1028:
1.7 millert 1029: Sep 30, 1988:
1030: Now guarantees to evaluate all arguments of built-in
1031: functions, as in C; the appearance is that arguments
1032: are evaluated before the function is called. Places
1033: affected are sub (gsub was ok), substr, printf, and
1034: all the built-in arithmetic functions in bltin().
1035: A warning is generated if a bltin() is called with
1036: the wrong number of arguments.
1.6 kstailey 1037:
1.7 millert 1038: This requires changing makeprof on p167 of the book.
1.6 kstailey 1039:
1.7 millert 1040: Aug 23, 1988:
1041: setting FILENAME in BEGIN caused core dump, apparently
1042: because it was freeing space not allocated by malloc.
1.6 kstailey 1043:
1.7 millert 1044: July 24, 1988:
1045: fixed egregious error in toupper/tolower functions.
1046: still subject to rescinding, however.
1.6 kstailey 1047:
1.7 millert 1048: July 2, 1988:
1049: flush stdout before opening file or pipe
1.6 kstailey 1050:
1.7 millert 1051: July 2, 1988:
1052: performance bug in b.c/cgoto(): not freeing some sets of states.
1053: partial fix only right now, and the number of states increased
1054: to make it less obvious.
1.6 kstailey 1055:
1.7 millert 1056: June 1, 1988:
1057: check error status on close
1.6 kstailey 1058:
1.7 millert 1059: May 28, 1988:
1060: srand returns seed value it's using.
1061: see 1/18/90
1.6 kstailey 1062:
1.7 millert 1063: May 22, 1988:
1064: Removed limit on depth of function calls.
1.6 kstailey 1065:
1.7 millert 1066: May 10, 1988:
1067: Fixed lib.c to permit _ in commandline variable names.
1.6 kstailey 1068:
1.7 millert 1069: Mar 25, 1988:
1070: main.c fixed to recognize -- as terminator of command-
1071: line options. Illegal options flagged.
1072: Error reporting slightly cleaned up.
1.6 kstailey 1073:
1.7 millert 1074: Dec 2, 1987:
1075: Newer C compilers apply a strict scope rule to extern
1076: declarations within functions. Two extern declarations in
1077: lib.c and tran.c have been moved to obviate this problem.
1.6 kstailey 1078:
1.7 millert 1079: Oct xx, 1987:
1080: Reluctantly added toupper and tolower functions.
1081: Subject to rescinding without notice.
1.6 kstailey 1082:
1.7 millert 1083: Sep 17, 1987:
1084: Error-message printer had printf(s) instead of
1085: printf("%s",s); got core dumps when the message
1086: included a %.
1.6 kstailey 1087:
1.7 millert 1088: Sep 12, 1987:
1089: Very long printf strings caused core dump;
1090: fixed aprintf, asprintf, format to catch them.
1091: Can still get a core dump in printf itself.
1.6 kstailey 1092:
1093: