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