Annotation of src/usr.bin/awk/FIXES, Revision 1.20
1.20 ! millert 1: /* $OpenBSD: FIXES,v 1.19 2020/06/10 21:00:01 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.
1.20 ! millert 28:
! 29: Aug 27, 2018:
! 30: Disallow '$' in printf formats; arguments evaluated in order
! 31: and printed in order.
! 32:
! 33: Added some casts to silence warnings on debugging printfs.
! 34: (Thanks, Arnold.)
1.18 millert 35:
1.19 millert 36: Aug 23, 2018:
37: A long list of fixes courtesy of Arnold Robbins,
38: to whom profound thanks.
39:
40: 1. ofs-rebuild: OFS value used to rebuild the record was incorrect.
41: Fixed August 19, 2014. Revised fix August 2018.
42:
43: 2. system-status: Instead of a floating-point division by 256, use
44: the wait(2) macros to create a reasonable exit status.
45: Fixed March 12, 2016.
46:
47: 3. space: Use provided xisblank() function instead of ispace() for
48: matching [[:blank:]].
49:
50: 4. a-format: Add POSIX standard %a and %A to supported formats. Check
51: at runtime that this format is available.
52:
53: 5. decr-NF: Decrementing NF did not change $0. This is a decades-old
54: bug. There are interactions with the old and new value of OFS as well.
55: Most of the fix came from the NetBSD awk.
56:
57: 6. string-conv: String conversions of scalars were sticky. Once a
58: conversion to string happened, even with OFMT, that value was used until
59: a new numeric value was assigned, even if OFMT differed from CONVFMT,
60: and also if CONVFMT changed.
61:
62: 7. unary-plus: Unary plus on a string constant returned the string.
63: Instead, it should convert the value to numeric and give that value.
64:
65: Also added Arnold's tests for these to awktest.tar as T.arnold.
66:
67: Aug 15, 2018:
68: fixed mangled awktest.tar (thanks, Arnold), posted all
69: current (very minor) fixes to github / onetrueawk
70:
1.18 millert 71: Jun 7, 2018:
72: (yes, a long layoff)
73: Updated some broken tests (beebe.tar, T.lilly)
74: [thanks to Arnold Robbins]
75:
76: Mar 26, 2015:
77: buffer overflow in error reporting; thanks to tobias ulmer
78: and john-mark gurney for spotting it and the fix.
79:
80: Feb 4, 2013:
81: cleaned up a handful of tests that didn't seem to actually
82: test for correct behavior: T.latin1, T.gawk.
83:
84: Jan 5, 2013:
85: added ,NULL initializer to static Cells in run.c; not really
86: needed but cleaner. Thanks to Michael Bombardieri.
1.17 millert 87:
88: Dec 20, 2012:
89: fiddled makefile to get correct yacc and bison flags. pick yacc
90: (linux) or bison (mac) as necessary.
91:
92: added __attribute__((__noreturn__)) to a couple of lines in
93: proto.h, to silence someone's enthusiastic checker.
94:
95: fixed obscure call by value bug in split(a[1],a) reported on
96: 9fans. the management of temporary values is just a mess; i
97: took a shortcut by making an extra string copy. thanks
98: to paul patience and arnold robbins for passing it on and for
99: proposed patches.
100:
101: tiny fiddle in setfval to eliminate -0 results in T.expr, which
102: has irritated me for 20+ years.
1.16 millert 103:
104: Aug 10, 2011:
105: another fix to avoid core dump with delete(ARGV); again, many thanks
106: to ruslan ermilov.
107:
108: Aug 7, 2011:
109: split(s, a, //) now behaves the same as split(s, a, "")
110:
111: Jun 12, 2011:
112: /pat/, \n /pat/ {...} is now legal, though bad style to use.
113:
114: added checks to new -v code that permits -vnospace; thanks to
115: ruslan ermilov for spotting this and providing the patch.
116:
117: removed fixed limit on number of open files; thanks to aleksey
118: cheusov and christos zoulos.
119:
120: fixed day 1 bug that resurrected deleted elements of ARGV when
121: used as filenames (in lib.c).
122:
123: minor type fiddles to make gcc -Wall -pedantic happier (but not
124: totally so); turned on -fno-strict-aliasing in makefile.
125:
126: May 6, 2011:
127: added #ifdef for isblank.
128: now allows -ffoo as well as -f foo arguments.
129: (thanks, ruslan)
130:
131: May 1, 2011:
132: after advice from todd miller, kevin lo, ruslan ermilov,
133: and arnold robbins, changed srand() to return the previous
134: seed (which is 1 on the first call of srand). the seed is
135: an Awkfloat internally though converted to unsigned int to
136: pass to the library srand(). thanks, everyone.
137:
138: fixed a subtle (and i hope low-probability) overflow error
139: in fldbld, by adding space for one extra \0. thanks to
140: robert bassett for spotting this one and providing a fix.
141:
142: removed the files related to compilation on windows. i no
143: longer have anything like a current windows environment, so
144: i can't test any of it.
1.15 millert 145:
146: May 23, 2010:
147: fixed long-standing overflow bug in run.c; many thanks to
148: nelson beebe for spotting it and providing the fix.
149:
150: fixed bug that didn't parse -vd=1 properly; thanks to santiago
151: vila for spotting it.
152:
153: Feb 8, 2010:
154: i give up. replaced isblank with isspace in b.c; there are
155: no consistent header files.
156:
157: Nov 26, 2009:
158: fixed a long-standing issue with when FS takes effect. a
159: change to FS is now noticed immediately for subsequent splits.
160:
161: changed the name getline() to awkgetline() to avoid yet another
162: name conflict somewhere.
163:
164: Feb 11, 2009:
165: temporarily for now defined HAS_ISBLANK, since that seems to
166: be the best way through the thicket. isblank arrived in C99,
167: but seems to be arriving at different systems at different
168: times.
169:
170: Oct 8, 2008:
171: fixed typo in b.c that set tmpvec wrongly. no one had ever
172: run into the problem, apparently. thanks to alistair crooks.
1.14 millert 173:
174: Oct 23, 2007:
175: minor fix in lib.c: increase inputFS to 100, change malloc
176: for fields to n+1.
177:
178: fixed memory fault caused by out of order test in setsval.
179:
180: thanks to david o'brien, freebsd, for both fixes.
181:
182: May 1, 2007:
183: fiddle in makefile to fix for BSD make; thanks to igor sobrado.
184:
185: Mar 31, 2007:
186: fixed some null pointer refs calling adjbuf.
187:
188: Feb 21, 2007:
189: fixed a bug in matching the null RE in sub and gsub. thanks to al aho
190: who actually did the fix (in b.c), and to wolfgang seeberg for finding
191: it and providing a very compact test case.
192:
193: fixed quotation in b.c; thanks to Hal Pratt and the Princeton Dante
194: Project.
195:
196: removed some no-effect asserts in run.c.
197:
198: fiddled maketab.c to not complain about bison-generated values.
199:
200: removed the obsolete -V argument; fixed --version to print the
201: version and exit.
202:
203: fixed wording and an outright error in the usage message; thanks to igor
204: sobrado and jason mcintyre.
205:
206: fixed a bug in -d that caused core dump if no program followed.
207:
208: Jan 1, 2007:
209: dropped mac.code from makefile; there are few non-MacOSX
210: mac's these days.
211:
212: Jan 17, 2006:
213: system() not flagged as unsafe in the unadvertised -safe option.
214: found it while enhancing tests before shipping the ;login: article.
215: practice what you preach.
216:
217: removed the 9-years-obsolete -mr and -mf flags.
218:
219: added -version and --version options.
220:
221: core dump on linux with BEGIN {nextfile}, now fixed.
222:
223: removed some #ifdef's in run.c and lex.c that appear to no
224: longer be necessary.
225:
226: Apr 24, 2005:
227: modified lib.c so that values of $0 et al are preserved in the END
228: block, apparently as required by posix. thanks to havard eidnes
229: for the report and code.
230:
231: Jan 14, 2005:
232: fixed infinite loop in parsing, originally found by brian tsang.
233: thanks to arnold robbins for a suggestion that started me
234: rethinking it.
235:
236: Dec 31, 2004:
237: prevent overflow of -f array in main, head off potential error in
238: call of SYNTAX(), test malloc return in lib.c, all with thanks to
239: todd miller.
1.13 millert 240:
241: Dec 22, 2004:
242: cranked up size of NCHARS; coverity thinks it can be overrun with
243: smaller size, and i think that's right. added some assertions to b.c
244: to catch places where it might overrun. the RE code is still fragile.
245:
246: Dec 5, 2004:
247: fixed a couple of overflow problems with ridiculous field numbers:
248: e.g., print $(2^32-1). thanks to ruslan ermilov, giorgos keramidas
249: and david o'brien at freebsd.org for patches. this really should
250: be re-done from scratch.
251:
252: Nov 21, 2004:
253: fixed another 25-year-old RE bug, in split. it's another failure
254: to (re-)initialize. thanks to steve fisher for spotting this and
255: providing a good test case.
256:
257: Nov 22, 2003:
258: fixed a bug in regular expressions that dates (so help me) from 1977;
259: it's been there from the beginning. an anchored longest match that
260: was longer than the number of states triggered a failure to initialize
261: the machine properly. many thanks to moinak ghosh for not only finding
262: this one but for providing a fix, in some of the most mysterious
263: code known to man.
264:
265: fixed a storage leak in call() that appears to have been there since
266: 1983 or so -- a function without an explicit return that assigns a
267: string to a parameter leaked a Cell. thanks to moinak ghosh for
268: spotting this very subtle one.
269:
270: Jul 31, 2003:
271: fixed, thanks to andrey chernov and ruslan ermilov, a bug in lex.c
272: that mis-handled the character 255 in input. (it was being compared
273: to EOF with a signed comparison.)
274:
275: Jul 29, 2003:
276: fixed (i think) the long-standing botch that included the beginning of
277: line state ^ for RE's in the set of valid characters; this led to a
278: variety of odd problems, including failure to properly match certain
279: regular expressions in non-US locales. thanks to ruslan for keeping
280: at this one.
281:
282: Jul 28, 2003:
283: n-th try at getting internationalization right, with thanks to volker
284: kiefel, arnold robbins and ruslan ermilov for advice, though they
285: should not be blamed for the outcome. according to posix, "." is the
286: radix character in programs and command line arguments regardless of
287: the locale; otherwise, the locale should prevail for input and output
288: of numbers. so it's intended to work that way.
289:
290: i have rescinded the attempt to use strcoll in expanding shorthands in
291: regular expressions (cclenter). its properties are much too
292: surprising; for example [a-c] matches aAbBc in locale en_US but abBcC
293: in locale fr_CA. i can see how this might arise by implementation
294: but i cannot explain it to a human user. (this behavior can be seen
295: in gawk as well; we're leaning on the same library.)
296:
297: the issue appears to be that strcoll is meant for sorting, where
298: merging upper and lower case may make sense (though note that unix
299: sort does not do this by default either). it is not appropriate
300: for regular expressions, where the goal is to match specific
301: patterns of characters. in any case, the notations [:lower:], etc.,
302: are available in awk, and they are more likely to work correctly in
303: most locales.
304:
305: a moratorium is hereby declared on internationalization changes.
306: i apologize to friends and colleagues in other parts of the world.
307: i would truly like to get this "right", but i don't know what
308: that is, and i do not want to keep making changes until it's clear.
309:
310: Jul 4, 2003:
311: fixed bug that permitted non-terminated RE, as in "awk /x".
312:
313: Jun 1, 2003:
314: subtle change to split: if source is empty, number of elems
315: is always 0 and the array is not set.
316:
317: Mar 21, 2003:
318: added some parens to isblank, in another attempt to make things
319: internationally portable.
320:
321: Mar 14, 2003:
322: the internationalization changes, somewhat modified, are now
323: reinstated. in theory awk will now do character comparisons
324: and case conversions in national language, but "." will always
325: be the decimal point separator on input and output regardless
326: of national language. isblank(){} has an #ifndef.
327:
328: this no longer compiles on windows: LC_MESSAGES isn't defined
329: in vc6++.
330:
331: fixed subtle behavior in field and record splitting: if FS is
332: a single character and RS is not empty, \n is NOT a separator.
333: this tortuous reading is found in the awk book; behavior now
334: matches gawk and mawk.
1.11 millert 335:
336: Dec 13, 2002:
337: for the moment, the internationalization changes of nov 29 are
338: rolled back -- programs like x = 1.2 don't work in some locales,
339: because the parser is expecting x = 1,2. until i understand this
340: better, this will have to wait.
341:
342: Nov 29, 2002:
343: modified b.c (with tiny changes in main and run) to support
344: locales, using strcoll and iswhatever tests for posix character
345: classes. thanks to ruslan ermilov (ru@freebsd.org) for code.
346: the function isblank doesn't seem to have propagated to any
347: header file near me, so it's there explicitly. not properly
348: tested on non-ascii character sets by me.
349:
350: Jun 28, 2002:
351: modified run/format() and tran/getsval() to do a slightly better
352: job on using OFMT for output from print and CONVFMT for other
353: number->string conversions, as promised by posix and done by
354: gawk and mawk. there are still places where it doesn't work
355: right if CONVFMT is changed; by then the STR attribute of the
356: variable has been irrevocably set. thanks to arnold robbins for
357: code and examples.
358:
359: fixed subtle bug in format that could get core dump. thanks to
360: Jaromir Dolecek <jdolecek@NetBSD.org> for finding and fixing.
361: minor cleanup in run.c / format() at the same time.
362:
363: added some tests for null pointers to debugging printf's, which
364: were never intended for external consumption. thanks to dave
365: kerns (dkerns@lucent.com) for pointing this out.
366:
367: GNU compatibility: an empty regexp matches anything (thanks to
368: dag-erling smorgrav, des@ofug.org). subject to reversion if
369: this does more harm than good.
370:
371: pervasive small changes to make things more const-correct, as
372: reported by gcc's -Wwrite-strings. as it says in the gcc manual,
373: this may be more nuisance than useful. provoked by a suggestion
374: and code from arnaud desitter, arnaud@nimbus.geog.ox.ac.uk
375:
376: minor documentation changes to note that this now compiles out
377: of the box on Mac OS X.
378:
379: Feb 10, 2002:
380: changed types in posix chars structure to quiet solaris cc.
381:
382: Jan 1, 2002:
383: fflush() or fflush("") flushes all files and pipes.
384:
385: length(arrayname) returns number of elements; thanks to
386: arnold robbins for suggestion.
387:
388: added a makefile.win to make it easier to build on windows.
389: based on dan allen's buildwin.bat.
390:
391: Nov 16, 2001:
392: added support for posix character class names like [:digit:],
393: which are not exactly shorter than [0-9] and perhaps no more
394: portable. thanks to dag-erling smorgrav for code.
395:
396: Feb 16, 2001:
397: removed -m option; no longer needed, and it was actually
398: broken (noted thanks to volker kiefel).
399:
400: Feb 10, 2001:
401: fixed an appalling bug in gettok: any sequence of digits, +,-, E, e,
402: and period was accepted as a valid number if it started with a period.
403: this would never have happened with the lex version.
404:
405: other 1-character botches, now fixed, include a bare $ and a
406: bare " at the end of the input.
407:
408: Feb 7, 2001:
409: more (const char *) casts in b.c and tran.c to silence warnings.
1.8 millert 410:
1.10 millert 411: Nov 15, 2000:
412: fixed a bug introduced in august 1997 that caused expressions
413: like $f[1] to be syntax errors. thanks to arnold robbins for
414: noticing this and providing a fix.
415:
416: Oct 30, 2000:
417: fixed some nextfile bugs: not handling all cases. thanks to
418: arnold robbins for pointing this out. new regressions added.
419:
420: close() is now a function. it returns whatever the library
421: fclose returns, and -1 for closing a file or pipe that wasn't
422: opened.
423:
424: Sep 24, 2000:
425: permit \n explicitly in character classes; won't work right
426: if comes in as "[\n]" but ok as /[\n]/, because of multiple
427: processing of \'s. thanks to arnold robbins.
428:
429: July 5, 2000:
430: minor fiddles in tran.c to keep compilers happy about uschar.
431: thanks to norman wilson.
432:
433: May 25, 2000:
434: yet another attempt at making 8-bit input work, with another
435: band-aid in b.c (member()), and some (uschar) casts to head
436: off potential errors in subscripts (like isdigit). also
437: changed HAT to NCHARS-2. thanks again to santiago vila.
438:
439: changed maketab.c to ignore apparently out of range definitions
440: instead of halting; new freeBSD generates one. thanks to
441: jon snader <jsnader@ix.netcom.com> for pointing out the problem.
442:
443: May 2, 2000:
444: fixed an 8-bit problem in b.c by making several char*'s into
445: unsigned char*'s. not clear i have them all yet. thanks to
446: Santiago Vila <sanvila@unex.es> for the bug report.
447:
448: Apr 21, 2000:
449: finally found and fixed a memory leak in function call; it's
450: been there since functions were added ~1983. thanks to
451: jon bentley for the test case that found it.
452:
453: added test in envinit to catch environment "variables" with
1.12 otto 454: names beginning with '='; thanks to Berend Hasselman.
1.10 millert 455:
456: Jul 28, 1999:
457: added test in defn() to catch function foo(foo), which
458: otherwise recurses until core dump. thanks to arnold
459: robbins for noticing this.
460:
1.9 millert 461: Jun 20, 1999:
462: added *bp in gettok in lex.c; appears possible to exit function
463: without terminating the string. thanks to russ cox.
464:
465: Jun 2, 1999:
466: added function stdinit() to run to initialize files[] array,
467: in case stdin, etc., are not constants; some compilers care.
468:
469: May 10, 1999:
470: replaced the ERROR ... FATAL, etc., macros with functions
471: based on vprintf, to avoid problems caused by overrunning
472: fixed-size errbuf array. thanks to ralph corderoy for the
473: impetus, and for pointing out a string termination bug in
474: qstring as well.
475:
476: Apr 21, 1999:
477: fixed bug that caused occasional core dumps with commandline
478: variable with value ending in \. (thanks to nelson beebe for
479: the test case.)
480:
1.8 millert 481: Apr 16, 1999:
482: with code kindly provided by Bruce Lilly, awk now parses
483: /=/ and similar constructs more sensibly in more places.
1.10 millert 484: Bruce also provided some helpful test cases.
1.8 millert 485:
486: Apr 5, 1999:
487: changed true/false to True/False in run.c to make it
488: easier to compile with C++. Added some casts on malloc
489: and realloc to be honest about casts; ditto. changed
490: ltype int to long in struct rrow to reduce some 64-bit
491: complaints; other changes scattered throughout for the
492: same purpose. thanks to Nelson Beebe for these portability
493: improvements.
494:
495: removed some horrible pointer-int casting in b.c and elsewhere
496: by adding ptoi and itonp to localize the casts, which are
497: all benign. fixed one incipient bug that showed up on sgi
498: in 64-bit mode.
499:
500: reset lineno for new source file; include filename in error
501: message. also fixed line number error in continuation lines.
502: (thanks to Nelson Beebe for both of these.)
503:
504: Mar 24, 1999:
505: Nelson Beebe notes that irix 5.3 yacc dies with a bogus
506: error; use a newer version or switch to bison, since sgi
507: is unlikely to fix it.
1.1 tholo 508:
1.7 millert 509: Mar 5, 1999:
1.9 millert 510: changed isnumber to is_number to avoid the problem caused by
511: versions of ctype.h that include the name isnumber.
1.7 millert 512:
513: distribution now includes a script for building on a Mac,
514: thanks to Dan Allen.
515:
516: Feb 20, 1999:
517: fixed memory leaks in run.c (call) and tran.c (setfval).
518: thanks to Stephen Nutt for finding these and providing the fixes.
519:
520: Jan 13, 1999:
521: replaced srand argument by (unsigned int) in run.c;
522: avoids problem on Mac and potentially on Unix & Windows.
523: thanks to Dan Allen.
524:
525: added a few (int) casts to silence useless compiler warnings.
526: e.g., errorflag= in run.c jump().
527:
528: added proctab.c to the bundle outout; one less thing
529: to have to compile out of the box.
530:
531: added calls to _popen and _pclose to the win95 stub for
532: pipes (thanks to Steve Adams for this helpful suggestion).
533: seems to work, though properties are not well understood
534: by me, and it appears that under some circumstances the
535: pipe output is truncated. Be careful.
536:
537: Oct 19, 1998:
538: fixed a couple of bugs in getrec: could fail to update $0
539: after a getline var; because inputFS wasn't initialized,
540: could split $0 on every character, a misleading diversion.
541:
542: fixed caching bug in makedfa: LRU was actually removing
543: least often used.
544:
545: thanks to ross ridge for finding these, and for providing
546: great bug reports.
547:
548: May 12, 1998:
549: fixed potential bug in readrec: might fail to update record
550: pointer after growing. thanks to dan levy for spotting this
551: and suggesting the fix.
552:
553: Mar 12, 1998:
554: added -V to print version number and die.
1.19 millert 555:
556: [notify dave kerns, dkerns@dacsoup.ih.lucent.com]
1.7 millert 557:
558: Feb 11, 1998:
559: subtle silent bug in lex.c: if the program ended with a number
560: longer than 1 digit, part of the input would be pushed back and
561: parsed again because token buffer wasn't terminated right.
562: example: awk 'length($0) > 10'. blush. at least i found it
563: myself.
564:
565: Aug 31, 1997:
566: s/adelete/awkdelete/: SGI uses this in malloc.h.
567: thanks to nelson beebe for pointing this one out.
568:
569: Aug 21, 1997:
570: fixed some bugs in sub and gsub when replacement includes \\.
571: this is a dark, horrible corner, but at least now i believe that
572: the behavior is the same as gawk and the intended posix standard.
573: thanks to arnold robbins for advice here.
574:
575: Aug 9, 1997:
576: somewhat regretfully, replaced the ancient lex-based lexical
577: analyzer with one written in C. it's longer, generates less code,
578: and more portable; the old one depended too much on mysterious
579: properties of lex that were not preserved in other environments.
580: in theory these recognize the same language.
581:
582: now using strtod to test whether a string is a number, instead of
583: the convoluted original function. should be more portable and
584: reliable if strtod is implemented right.
585:
586: removed now-pointless optimization in makefile that tries to avoid
587: recompilation when awkgram.y is changed but symbols are not.
588:
589: removed most fixed-size arrays, though a handful remain, some
590: of which are unchecked. you have been warned.
591:
592: Aug 4, 1997:
593: with some trepidation, replaced the ancient code that managed
594: fields and $0 in fixed-size arrays with arrays that grow on
595: demand. there is still some tension between trying to make this
596: run fast and making it clean; not sure it's right yet.
597:
598: the ill-conceived -mr and -mf arguments are now useful only
599: for debugging. previous dynamic string code removed.
600:
601: numerous other minor cleanups along the way.
1.1 tholo 602:
1.7 millert 603: Jul 30, 1997:
604: using code provided by dan levy (to whom profuse thanks), replaced
605: fixed-size arrays and awkward kludges by a fairly uniform mechanism
606: to grow arrays as needed for printf, sub, gsub, etc.
1.1 tholo 607:
1.7 millert 608: Jul 23, 1997:
609: falling off the end of a function returns "" and 0, not 0.
610: thanks to arnold robbins.
1.1 tholo 611:
1.7 millert 612: Jun 17, 1997:
613: replaced several fixed-size arrays by dynamically-created ones
614: in run.c; added overflow tests to some previously unchecked cases.
615: getline, toupper, tolower.
1.1 tholo 616:
1.7 millert 617: getline code is still broken in that recursive calls may wind
618: up using the same space. [fixed later]
1.1 tholo 619:
1.7 millert 620: increased RECSIZE to 8192 to push problems further over the horizon.
1.1 tholo 621:
1.7 millert 622: added \r to \n as input line separator for programs, not data.
623: damn CRLFs.
1.1 tholo 624:
1.7 millert 625: modified format() to permit explicit printf("%c", 0) to include
626: a null byte in output. thanks to ken stailey for the fix.
1.1 tholo 627:
1.7 millert 628: added a "-safe" argument that disables file output (print >,
629: print >>), process creation (cmd|getline, print |, system), and
630: access to the environment (ENVIRON). this is a first approximation
631: to a "safe" version of awk, but don't rely on it too much. thanks
632: to joan feigenbaum and matt blaze for the inspiration long ago.
1.1 tholo 633:
1.7 millert 634: Jul 8, 1996:
635: fixed long-standing bug in sub, gsub(/a/, "\\\\&"); thanks to
636: ralph corderoy.
1.1 tholo 637:
1.7 millert 638: Jun 29, 1996:
639: fixed awful bug in new field splitting; didn't get all the places
640: where input was done.
1.1 tholo 641:
1.7 millert 642: Jun 28, 1996:
643: changed field-splitting to conform to posix definition: fields are
644: split using the value of FS at the time of input; it used to be
645: the value when the field or NF was first referred to, a much less
646: predictable definition. thanks to arnold robbins for encouragement
647: to do the right thing.
1.1 tholo 648:
1.7 millert 649: May 28, 1996:
650: fixed appalling but apparently unimportant bug in parsing octal
651: numbers in reg exprs.
1.1 tholo 652:
1.7 millert 653: explicit hex in reg exprs now limited to 2 chars: \xa, \xaa.
1.1 tholo 654:
1.7 millert 655: May 27, 1996:
656: cleaned up some declarations so gcc -Wall is now almost silent.
1.1 tholo 657:
1.7 millert 658: makefile now includes backup copies of ytab.c and lexyy.c in case
659: one makes before looking; it also avoids recreating lexyy.c unless
660: really needed.
1.1 tholo 661:
1.7 millert 662: s/aprintf/awkprint, s/asprintf/awksprintf/ to avoid some name clashes
663: with unwisely-written header files.
1.1 tholo 664:
1.7 millert 665: thanks to jeffrey friedl for several of these.
1.1 tholo 666:
1.7 millert 667: May 26, 1996:
668: an attempt to rationalize the (unsigned) char issue. almost all
669: instances of unsigned char have been removed; the handful of places
670: in b.c where chars are used as table indices have been hand-crafted.
671: added some latin-1 tests to the regression, but i'm not confident;
672: none of my compilers seem to care much. thanks to nelson beebe for
673: pointing out some others that do care.
1.1 tholo 674:
1.7 millert 675: May 2, 1996:
676: removed all register declarations.
1.1 tholo 677:
1.7 millert 678: enhanced split(), as in gawk, etc: split(s, a, "") splits s into
679: a[1]...a[length(s)] with each character a single element.
1.1 tholo 680:
1.7 millert 681: made the same changes for field-splitting if FS is "".
1.1 tholo 682:
1.7 millert 683: added nextfile, as in gawk: causes immediate advance to next
684: input file. (thanks to arnold robbins for inspiration and code).
1.1 tholo 685:
1.7 millert 686: small fixes to regexpr code: can now handle []], [[], and
687: variants; [] is now a syntax error, rather than matching
688: everything; [z-a] is now empty, not z. far from complete
689: or correct, however. (thanks to jeffrey friedl for pointing out
690: some awful behaviors.)
1.1 tholo 691:
1.7 millert 692: Apr 29, 1996:
1.12 otto 693: replaced uchar by uschar everywhere; apparently some compilers
1.7 millert 694: usurp this name and this causes conflicts.
1.1 tholo 695:
1.7 millert 696: fixed call to time in run.c (bltin); arg is time_t *.
1.1 tholo 697:
1.7 millert 698: replaced horrible pointer/long punning in b.c by a legitimate
699: union. should be safer on 64-bit machines and cleaner everywhere.
700: (thanks to nelson beebe for pointing out some of these problems.)
1.1 tholo 701:
1.7 millert 702: replaced nested comments by #if 0...#endif in run.c, lib.c.
1.1 tholo 703:
1.7 millert 704: removed getsval, setsval, execute macros from run.c and lib.c.
705: machines are 100x faster than they were when these macros were
706: first used.
1.1 tholo 707:
1.7 millert 708: revised filenames: awk.g.y => awkgram.y, awk.lx.l => awklex.l,
709: y.tab.[ch] => ytab.[ch], lex.yy.c => lexyy.c, all in the aid of
710: portability to nameless systems.
1.1 tholo 711:
1.7 millert 712: "make bundle" now includes yacc and lex output files for recipients
713: who don't have yacc or lex.
1.1 tholo 714:
1.7 millert 715: Aug 15, 1995:
716: initialized Cells in setsymtab more carefully; some fields
717: were not set. (thanks to purify, all of whose complaints i
718: think i now understand.)
1.1 tholo 719:
1.7 millert 720: fixed at least one error in gsub that looked at -1-th element
721: of an array when substituting for a null match (e.g., $).
1.1 tholo 722:
1.7 millert 723: delete arrayname is now legal; it clears the elements but leaves
724: the array, which may not be the right behavior.
1.1 tholo 725:
1.7 millert 726: modified makefile: my current make can't cope with the test used
727: to avoid unnecessary yacc invocations.
1.1 tholo 728:
1.7 millert 729: Jul 17, 1995:
730: added dynamically growing strings to awk.lx.l and b.c
731: to permit regular expressions to be much bigger.
732: the state arrays can still overflow.
1.1 tholo 733:
1.7 millert 734: Aug 24, 1994:
735: detect duplicate arguments in function definitions (mdm).
1.1 tholo 736:
1.7 millert 737: May 11, 1994:
738: trivial fix to printf to limit string size in sub().
1.1 tholo 739:
1.7 millert 740: Apr 22, 1994:
741: fixed yet another subtle self-assignment problem:
742: $1 = $2; $1 = $1 clobbered $1.
1.1 tholo 743:
1.7 millert 744: Regression tests now use private echo, to avoid quoting problems.
1.1 tholo 745:
1.7 millert 746: Feb 2, 1994:
747: changed error() to print line number as %d, not %g.
1.1 tholo 748:
1.7 millert 749: Jul 23, 1993:
750: cosmetic changes: increased sizes of some arrays,
751: reworded some error messages.
1.1 tholo 752:
1.7 millert 753: added CONVFMT as in posix (just replaced OFMT in getsval)
1.1 tholo 754:
1.7 millert 755: FILENAME is now "" until the first thing that causes a file
756: to be opened.
1.1 tholo 757:
1.7 millert 758: Nov 28, 1992:
759: deleted yyunput and yyoutput from proto.h;
760: different versions of lex give these different declarations.
1.1 tholo 761:
1.7 millert 762: May 31, 1992:
763: added -mr N and -mf N options: more record and fields.
764: these really ought to adjust automatically.
1.1 tholo 765:
1.7 millert 766: cleaned up some error messages; "out of space" now means
767: malloc returned NULL in all cases.
1.1 tholo 768:
1.7 millert 769: changed rehash so that if it runs out, it just returns;
770: things will continue to run slow, but maybe a bit longer.
1.1 tholo 771:
1.7 millert 772: Apr 24, 1992:
773: remove redundant close of stdin when using -f -.
1.1 tholo 774:
1.7 millert 775: got rid of core dump with -d; awk -d just prints date.
1.1 tholo 776:
1.7 millert 777: Apr 12, 1992:
778: added explicit check for /dev/std(in,out,err) in redirection.
779: unlike gawk, no /dev/fd/n yet.
1.1 tholo 780:
1.7 millert 781: added (file/pipe) builtin. hard to test satisfactorily.
782: not posix.
1.1 tholo 783:
1.7 millert 784: Feb 20, 1992:
785: recompile after abortive changes; should be unchanged.
1.1 tholo 786:
1.7 millert 787: Dec 2, 1991:
788: die-casting time: converted to ansi C, installed that.
1.1 tholo 789:
1.7 millert 790: Nov 30, 1991:
791: fixed storage leak in freefa, failing to recover [N]CCL.
792: thanks to Bill Jones (jones@cs.usask.ca)
1.1 tholo 793:
1.7 millert 794: Nov 19, 1991:
795: use RAND_MAX instead of literal in builtin().
1.1 tholo 796:
1.7 millert 797: Nov 12, 1991:
798: cranked up some fixed-size arrays in b.c, and added a test for
799: overflow in penter. thanks to mark larsen.
1.1 tholo 800:
1.7 millert 801: Sep 24, 1991:
802: increased buffer in gsub. a very crude fix to a general problem.
803: and again on Sep 26.
1.1 tholo 804:
1.7 millert 805: Aug 18, 1991:
806: enforce variable name syntax for commandline variables: has to
807: start with letter or _.
1.1 tholo 808:
1.7 millert 809: Jul 27, 1991:
810: allow newline after ; in for statements.
1.1 tholo 811:
1.7 millert 812: Jul 21, 1991:
813: fixed so that in self-assignment like $1=$1, side effects
814: like recomputing $0 take place. (this is getting subtle.)
1.1 tholo 815:
1.7 millert 816: Jun 30, 1991:
817: better test for detecting too-long output record.
1.1 tholo 818:
1.7 millert 819: Jun 2, 1991:
820: better defense against very long printf strings.
821: made break and continue illegal outside of loops.
1.1 tholo 822:
1.7 millert 823: May 13, 1991:
824: removed extra arg on gettemp, tempfree. minor error message rewording.
1.1 tholo 825:
826: May 6, 1991:
827: fixed silly bug in hex parsing in hexstr().
828: removed an apparently unnecessary test in isnumber().
829: warn about weird printf conversions.
830: fixed unchecked array overwrite in relex().
831:
832: changed for (i in array) to access elements in sorted order.
833: then unchanged it -- it really does run slower in too many cases.
834: left the code in place, commented out.
835:
1.7 millert 836: Feb 10, 1991:
837: check error status on all writes, to avoid banging on full disks.
1.1 tholo 838:
1.7 millert 839: Jan 28, 1991:
840: awk -f - reads the program from stdin.
1.1 tholo 841:
1.7 millert 842: Jan 11, 1991:
843: failed to set numeric state on $0 in cmd|getline context in run.c.
1.1 tholo 844:
1.7 millert 845: Nov 2, 1990:
846: fixed sleazy test for integrality in getsval; use modf.
1.1 tholo 847:
1.7 millert 848: Oct 29, 1990:
849: fixed sleazy buggy code in lib.c that looked (incorrectly) for
850: too long input lines.
1.1 tholo 851:
1.7 millert 852: Oct 14, 1990:
853: fixed the bug on p. 198 in which it couldn't deduce that an
854: argument was an array in some contexts. replaced the error
855: message in intest() by code that damn well makes it an array.
1.1 tholo 856:
1.7 millert 857: Oct 8, 1990:
858: fixed horrible bug: types and values were not preserved in
859: some kinds of self-assignment. (in assign().)
1.1 tholo 860:
1.7 millert 861: Aug 24, 1990:
862: changed NCHARS to 256 to handle 8-bit characters in strings
863: presented to match(), etc.
1.1 tholo 864:
1.7 millert 865: Jun 26, 1990:
866: changed struct rrow (awk.h) to use long instead of int for lval,
867: since cfoll() stores a pointer in it. now works better when int's
868: are smaller than pointers!
1.1 tholo 869:
1.7 millert 870: May 6, 1990:
871: AVA fixed the grammar so that ! is uniformly of the same precedence as
872: unary + and -. This renders illegal some constructs like !x=y, which
873: now has to be parenthesized as !(x=y), and makes others work properly:
874: !x+y is (!x)+y, and x!y is x !y, not two pattern-action statements.
875: (These problems were pointed out by Bob Lenk of Posix.)
1.1 tholo 876:
1.7 millert 877: Added \x to regular expressions (already in strings).
878: Limited octal to octal digits; \8 and \9 are not octal.
879: Centralized the code for parsing escapes in regular expressions.
880: Added a bunch of tests to T.re and T.sub to verify some of this.
1.1 tholo 881:
1.7 millert 882: Feb 9, 1990:
883: fixed null pointer dereference bug in main.c: -F[nothing]. sigh.
1.1 tholo 884:
1.7 millert 885: restored srand behavior: it returns the current seed.
1.1 tholo 886:
1.7 millert 887: Jan 18, 1990:
888: srand now returns previous seed value (0 to start).
1.1 tholo 889:
1.7 millert 890: Jan 5, 1990:
891: fix potential problem in tran.c -- something was freed,
892: then used in freesymtab.
1.1 tholo 893:
1.7 millert 894: Oct 18, 1989:
895: another try to get the max number of open files set with
896: relatively machine-independent code.
1.1 tholo 897:
1.7 millert 898: small fix to input() in case of multiple reads after EOF.
1.1 tholo 899:
1.7 millert 900: Oct 11, 1989:
901: FILENAME is now defined in the BEGIN block -- too many old
902: programs broke.
1.1 tholo 903:
1.7 millert 904: "-" means stdin in getline as well as on the commandline.
1.1 tholo 905:
1.7 millert 906: added a bunch of casts to the code to tell the truth about
907: char * vs. unsigned char *, a right royal pain. added a
908: setlocale call to the front of main, though probably no one
909: has it usefully implemented yet.
1.1 tholo 910:
1.7 millert 911: Aug 24, 1989:
912: removed redundant relational tests against nullnode if parse
913: tree already had a relational at that point.
1.1 tholo 914:
1.7 millert 915: Aug 11, 1989:
916: fixed bug: commandline variable assignment has to look like
917: var=something. (consider the man page for =, in file =.1)
1.1 tholo 918:
1.7 millert 919: changed number of arguments to functions to static arrays
920: to avoid repeated malloc calls.
1.1 tholo 921:
1.7 millert 922: Aug 2, 1989:
923: restored -F (space) separator
1.1 tholo 924:
1.7 millert 925: Jul 30, 1989:
926: added -v x=1 y=2 ... for immediate commandline variable assignment;
927: done before the BEGIN block for sure. they have to precede the
928: program if the program is on the commandline.
929: Modified Aug 2 to require a separate -v for each assignment.
1.1 tholo 930:
1.7 millert 931: Jul 10, 1989:
932: fixed ref-thru-zero bug in environment code in tran.c
1.1 tholo 933:
1.7 millert 934: Jun 23, 1989:
935: add newline to usage message.
1.1 tholo 936:
1.7 millert 937: Jun 14, 1989:
938: added some missing ansi printf conversion letters: %i %X %E %G.
939: no sensible meaning for h or L, so they may not do what one expects.
1.1 tholo 940:
1.7 millert 941: made %* conversions work.
1.1 tholo 942:
1.7 millert 943: changed x^y so that if n is a positive integer, it's done
944: by explicit multiplication, thus achieving maximum accuracy.
945: (this should be done by pow() but it seems not to be locally.)
946: done to x ^= y as well.
1.1 tholo 947:
1.7 millert 948: Jun 4, 1989:
949: ENVIRON array contains environment: if shell variable V=thing,
950: ENVIRON["V"] is "thing"
1.1 tholo 951:
1.7 millert 952: multiple -f arguments permitted. error reporting is naive.
953: (they were permitted before, but only the last was used.)
1.1 tholo 954:
1.7 millert 955: fixed a really stupid botch in the debugging macro dprintf
1.1 tholo 956:
1.7 millert 957: fixed order of evaluation of commandline assignments to match
958: what the book claims: an argument of the form x=e is evaluated
959: at the time it would have been opened if it were a filename (p 63).
960: this invalidates the suggested answer to ex 4-1 (p 195).
1.1 tholo 961:
1.7 millert 962: removed some code that permitted -F (space) fieldseparator,
963: since it didn't quite work right anyway. (restored aug 2)
1.1 tholo 964:
1.7 millert 965: Apr 27, 1989:
966: Line number now accumulated correctly for comment lines.
1.1 tholo 967:
1.7 millert 968: Apr 26, 1989:
969: Debugging output now includes a version date,
970: if one compiles it into the source each time.
1.1 tholo 971:
1.7 millert 972: Apr 9, 1989:
973: Changed grammar to prohibit constants as 3rd arg of sub and gsub;
974: prevents class of overwriting-a-constant errors. (Last one?)
975: This invalidates the "banana" example on page 43 of the book.
1.1 tholo 976:
1.7 millert 977: Added \a ("alert"), \v (vertical tab), \xhhh (hexadecimal),
978: as in ANSI, for strings. Rescinded the sloppiness that permitted
979: non-octal digits in \ooo. Warning: not all compilers and libraries
980: will be able to deal with \x correctly.
1.1 tholo 981:
1.7 millert 982: Jan 9, 1989:
983: Fixed bug that caused tempcell list to contain a duplicate.
984: The fix is kludgy.
1.1 tholo 985:
1.7 millert 986: Dec 17, 1988:
987: Catches some more commandline errors in main.
988: Removed redundant decl of modf in run.c (confuses some compilers).
989: Warning: there's no single declaration of malloc, etc., in awk.h
990: that seems to satisfy all compilers.
1.1 tholo 991:
1.7 millert 992: Dec 7, 1988:
993: Added a bit of code to error printing to avoid printing nulls.
994: (Not clear that it actually would.)
1.1 tholo 995:
1.7 millert 996: Nov 27, 1988:
997: With fear and trembling, modified the grammar to permit
998: multiple pattern-action statements on one line without
999: an explicit separator. By definition, this capitulation
1000: to the ghost of ancient implementations remains undefined
1001: and thus subject to change without notice or apology.
1002: DO NOT COUNT ON IT.
1.1 tholo 1003:
1.7 millert 1004: Oct 30, 1988:
1005: Fixed bug in call() that failed to recover storage.
1.1 tholo 1006:
1.7 millert 1007: A warning is now generated if there are more arguments
1008: in the call than in the definition (in lieu of fixing
1009: another storage leak).
1.2 millert 1010:
1.7 millert 1011: Oct 20, 1988:
1012: Fixed %c: if expr is numeric, use numeric value;
1013: otherwise print 1st char of string value. still
1014: doesn't work if the value is 0 -- won't print \0.
1.2 millert 1015:
1.7 millert 1016: Added a few more checks for running out of malloc.
1.2 millert 1017:
1.7 millert 1018: Oct 12, 1988:
1019: Fixed bug in call() that freed local arrays twice.
1.5 kstailey 1020:
1.7 millert 1021: Fixed to handle deletion of non-existent array right;
1022: complains about attempt to delete non-array element.
1.6 kstailey 1023:
1.7 millert 1024: Sep 30, 1988:
1025: Now guarantees to evaluate all arguments of built-in
1026: functions, as in C; the appearance is that arguments
1027: are evaluated before the function is called. Places
1028: affected are sub (gsub was ok), substr, printf, and
1029: all the built-in arithmetic functions in bltin().
1030: A warning is generated if a bltin() is called with
1031: the wrong number of arguments.
1.6 kstailey 1032:
1.7 millert 1033: This requires changing makeprof on p167 of the book.
1.6 kstailey 1034:
1.7 millert 1035: Aug 23, 1988:
1036: setting FILENAME in BEGIN caused core dump, apparently
1037: because it was freeing space not allocated by malloc.
1.6 kstailey 1038:
1.7 millert 1039: July 24, 1988:
1040: fixed egregious error in toupper/tolower functions.
1041: still subject to rescinding, however.
1.6 kstailey 1042:
1.7 millert 1043: July 2, 1988:
1044: flush stdout before opening file or pipe
1.6 kstailey 1045:
1.7 millert 1046: July 2, 1988:
1047: performance bug in b.c/cgoto(): not freeing some sets of states.
1048: partial fix only right now, and the number of states increased
1049: to make it less obvious.
1.6 kstailey 1050:
1.7 millert 1051: June 1, 1988:
1052: check error status on close
1.6 kstailey 1053:
1.7 millert 1054: May 28, 1988:
1055: srand returns seed value it's using.
1056: see 1/18/90
1.6 kstailey 1057:
1.7 millert 1058: May 22, 1988:
1059: Removed limit on depth of function calls.
1.6 kstailey 1060:
1.7 millert 1061: May 10, 1988:
1062: Fixed lib.c to permit _ in commandline variable names.
1.6 kstailey 1063:
1.7 millert 1064: Mar 25, 1988:
1065: main.c fixed to recognize -- as terminator of command-
1066: line options. Illegal options flagged.
1067: Error reporting slightly cleaned up.
1.6 kstailey 1068:
1.7 millert 1069: Dec 2, 1987:
1070: Newer C compilers apply a strict scope rule to extern
1071: declarations within functions. Two extern declarations in
1072: lib.c and tran.c have been moved to obviate this problem.
1.6 kstailey 1073:
1.7 millert 1074: Oct xx, 1987:
1075: Reluctantly added toupper and tolower functions.
1076: Subject to rescinding without notice.
1.6 kstailey 1077:
1.7 millert 1078: Sep 17, 1987:
1079: Error-message printer had printf(s) instead of
1080: printf("%s",s); got core dumps when the message
1081: included a %.
1.6 kstailey 1082:
1.7 millert 1083: Sep 12, 1987:
1084: Very long printf strings caused core dump;
1085: fixed aprintf, asprintf, format to catch them.
1086: Can still get a core dump in printf itself.
1.6 kstailey 1087:
1088: