2000-06-05 Michael Kifer <kifer@cs.sunysb.edu>
[emacs.git] / lib-src / rcs2log
blobf41552e110de498cba4ab84b5c81bce249b8411f
1 #! /bin/sh
3 # RCS to ChangeLog generator
5 # Generate a change log prefix from RCS files (perhaps in the CVS repository)
6 # and the ChangeLog (if any).
7 # Output the new prefix to standard output.
8 # You can edit this prefix by hand, and then prepend it to ChangeLog.
10 # Ignore log entries that start with `#'.
11 # Clump together log entries that start with `{topic} ',
12 # where `topic' contains neither white space nor `}'.
14 Help='The default FILEs are the files registered under the working directory.
15 Options:
17 -c CHANGELOG Output a change log prefix to CHANGELOG (default ChangeLog).
18 -h HOSTNAME Use HOSTNAME in change log entries (default current host).
19 -i INDENT Indent change log lines by INDENT spaces (default 8).
20 -l LENGTH Try to limit log lines to LENGTH characters (default 79).
21 -R If no FILEs are given and RCS is used, recurse through working directory.
22 -r OPTION Pass OPTION to subsidiary log command.
23 -t TABWIDTH Tab stops are every TABWIDTH characters (default 8).
24 -u "LOGIN<tab>FULLNAME<tab>MAILADDR" Assume LOGIN has FULLNAME and MAILADDR.
25 -v Append RCS revision to file names in log lines.
26 --help Output help.
27 --version Output version number.
29 Report bugs to <bug-gnu-emacs@gnu.org>.'
31 Id='$Id: rcs2log,v 1.44 1998/08/12 14:22:14 eggert Exp eggert $'
33 # Copyright 1992, 93, 94, 95, 96, 97, 1998 Free Software Foundation, Inc.
35 # This program is free software; you can redistribute it and/or modify
36 # it under the terms of the GNU General Public License as published by
37 # the Free Software Foundation; either version 2, or (at your option)
38 # any later version.
40 # This program is distributed in the hope that it will be useful,
41 # but WITHOUT ANY WARRANTY; without even the implied warranty of
42 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
43 # GNU General Public License for more details.
45 # You should have received a copy of the GNU General Public License
46 # along with this program; see the file COPYING. If not, write to the
47 # Free Software Foundation, Inc., 59 Temple Place - Suite 330,
48 # Boston, MA 02111-1307, USA.
50 Copyright='Copyright 1998 Free Software Foundation, Inc.
51 This program comes with NO WARRANTY, to the extent permitted by law.
52 You may redistribute copies of this program
53 under the terms of the GNU General Public License.
54 For more information about these matters, see the files named COPYING.
55 Author: Paul Eggert <eggert@twinsun.com>'
57 tab=' '
58 nl='
61 # Parse options.
63 # defaults
64 : ${AWK=awk}
65 : ${TMPDIR=/tmp}
66 changelog=ChangeLog # change log file name
67 datearg= # rlog date option
68 hostname= # name of local host (if empty, will deduce it later)
69 indent=8 # indent of log line
70 length=79 # suggested max width of log line
71 logins= # login names for people we know fullnames and mailaddrs of
72 loginFullnameMailaddrs= # login<tab>fullname<tab>mailaddr triplets
73 logTZ= # time zone for log dates (if empty, use local time)
74 recursive= # t if we want recursive rlog
75 revision= # t if we want revision numbers
76 rlog_options= # options to pass to rlog
77 tabwidth=8 # width of horizontal tab
79 while :
81 case $1 in
82 -c) changelog=${2?}; shift;;
83 -i) indent=${2?}; shift;;
84 -h) hostname=${2?}; shift;;
85 -l) length=${2?}; shift;;
86 -[nu]) # -n is obsolescent; it is replaced by -u.
87 case $1 in
88 -n) case ${2?}${3?}${4?} in
89 *"$tab"* | *"$nl"*)
90 echo >&2 "$0: -n '$2' '$3' '$4': tabs, newlines not allowed"
91 exit 1
92 esac
93 case $loginFullnameMailaddrs in
94 '') loginFullnameMailaddrs=$2$tab$3$tab$4;;
95 ?*) loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2$tab$3$tab$4
96 esac
97 shift; shift; shift;;
98 -u)
99 # If $2 is not tab-separated, use colon for separator.
100 case ${2?} in
101 *"$nl"*)
102 echo >&2 "$0: -u '$2': newlines not allowed"
103 exit 1;;
104 *"$tab"*)
105 t=$tab;;
108 esac
109 case $2 in
110 *"$t"*"$t"*"$t"*)
111 echo >&2 "$0: -u '$2': too many fields"
112 exit 1;;
113 *"$t"*"$t"*)
116 echo >&2 "$0: -u '$2': not enough fields"
117 exit 1
118 esac
119 case $loginFullnameMailaddrs in
120 '') loginFullnameMailaddrs=$2;;
121 ?*) loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2
122 esac
123 shift
124 esac
125 case $logins in
126 '') logins=$login;;
127 ?*) logins=$logins$nl$login
128 esac
131 case $rlog_options in
132 '') rlog_options=${2?};;
133 ?*) rlog_options=$rlog_options$nl${2?}
134 esac
135 shift;;
136 -R) recursive=t;;
137 -t) tabwidth=${2?}; shift;;
138 -v) revision=t;;
139 --version)
140 set $Id
141 rcs2logVersion=$3
142 echo >&2 "rcs2log (GNU Emacs) $rcs2logVersion$nl$Copyright"
143 exit 0;;
144 -*) echo >&2 "Usage: $0 [OPTION]... [FILE ...]$nl$Help"
145 case $1 in
146 --help) exit 0;;
147 *) exit 1
148 esac;;
149 *) break
150 esac
151 shift
152 done
154 month_data='
155 m[0]="Jan"; m[1]="Feb"; m[2]="Mar"
156 m[3]="Apr"; m[4]="May"; m[5]="Jun"
157 m[6]="Jul"; m[7]="Aug"; m[8]="Sep"
158 m[9]="Oct"; m[10]="Nov"; m[11]="Dec"
162 # Put rlog output into $rlogout.
164 # If no rlog options are given,
165 # log the revisions checked in since the first ChangeLog entry.
166 # Since ChangeLog is only by date, some of these revisions may be duplicates of
167 # what's already in ChangeLog; it's the user's responsibility to remove them.
168 case $rlog_options in
170 if test -s "$changelog"
171 then
173 /^[0-9]+-[0-9][0-9]-[0-9][0-9]/{
174 # ISO 8601 date
175 print $1
176 exit
178 /^... ... [ 0-9][0-9] [ 0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9]+ /{
179 # old-fashioned date and time (Emacs 19.31 and earlier)
180 '"$month_data"'
181 year = $5
182 for (i=0; i<=11; i++) if (m[i] == $2) break
183 dd = $3
184 printf "%d-%02d-%02d\n", year, i+1, dd
185 exit
188 d=`$AWK "$e" <"$changelog"` || exit
189 case $d in
190 ?*) datearg="-d>$d"
191 esac
193 esac
195 # Use TZ specified by ChangeLog local variable, if any.
196 if test -s "$changelog"
197 then
198 extractTZ='
199 /^.*change-log-time-zone-rule['"$tab"' ]*:['"$tab"' ]*"\([^"]*\)".*/{
200 s//\1/; p; q
202 /^.*change-log-time-zone-rule['"$tab"' ]*:['"$tab"' ]*t.*/{
203 s//UTC0/; p; q
206 logTZ=`tail "$changelog" | sed -n "$extractTZ"`
207 case $logTZ in
208 ?*) TZ=$logTZ; export TZ
209 esac
212 # If CVS is in use, examine its repository, not the normal RCS files.
213 if test ! -f CVS/Repository
214 then
215 rlog=rlog
216 repository=
217 else
218 rlog='cvs -q log'
219 repository=`sed 1q <CVS/Repository` || exit
220 test ! -f CVS/Root || CVSROOT=`cat <CVS/Root` || exit
221 case $CVSROOT in
222 *:/*)
223 # remote repository
226 # local repository
227 case $repository in
228 /*) ;;
229 *) repository=${CVSROOT?}/$repository
230 esac
231 if test ! -d "$repository"
232 then
233 echo >&2 "$0: $repository: bad repository (see CVS/Repository)"
234 exit 1
236 esac
239 # Use $rlog's -zLT option, if $rlog supports it.
240 case `$rlog -zLT 2>&1` in
241 *' option'*) ;;
243 case $rlog_options in
244 '') rlog_options=-zLT;;
245 ?*) rlog_options=-zLT$nl$rlog_options
246 esac
247 esac
249 # With no arguments, examine all files under the RCS directory.
250 case $# in
252 case $repository in
254 oldIFS=$IFS
255 IFS=$nl
256 case $recursive in
258 RCSdirs=`find . -name RCS -type d -print`
259 filesFromRCSfiles='s|,v$||; s|/RCS/|/|; s|^\./||'
260 files=`
262 case $RCSdirs in
263 ?*) find $RCSdirs \
264 -type f \
265 ! -name '*_' \
266 ! -name ',*,' \
267 ! -name '.*_' \
268 ! -name .rcsfreeze.log \
269 ! -name .rcsfreeze.ver \
270 -print
271 esac
272 find . -name '*,v' -print
274 sort -u |
275 sed "$filesFromRCSfiles"
278 files=
279 for file in RCS/.* RCS/* .*,v *,v
281 case $file in
282 RCS/. | RCS/.. | RCS/,*, | RCS/*_) continue;;
283 RCS/.rcsfreeze.log | RCS/.rcsfreeze.ver) continue;;
284 RCS/.\* | RCS/\* | .\*,v | \*,v) test -f "$file" || continue;;
285 RCS/*,v | RCS/.*,v) ;;
286 RCS/* | RCS/.*) test -f "$file" || continue
287 esac
288 case $files in
289 '') files=$file;;
290 ?*) files=$files$nl$file
291 esac
292 done
293 case $files in
294 '') exit 0
295 esac
296 esac
297 set x $files
298 shift
299 IFS=$oldIFS
300 esac
301 esac
303 llogout=$TMPDIR/rcs2log$$l
304 rlogout=$TMPDIR/rcs2log$$r
305 trap exit 1 2 13 15
306 trap "rm -f $llogout $rlogout; exit 1" 0
308 case $datearg in
309 ?*) $rlog $rlog_options "$datearg" ${1+"$@"} >$rlogout;;
310 '') $rlog $rlog_options ${1+"$@"} >$rlogout
311 esac || exit
314 # Get the full name of each author the logs mention, and set initialize_fullname
315 # to awk code that initializes the `fullname' awk associative array.
316 # Warning: foreign authors (i.e. not known in the passwd file) are mishandled;
317 # you have to fix the resulting output by hand.
319 initialize_fullname=
320 initialize_mailaddr=
322 case $loginFullnameMailaddrs in
324 case $loginFullnameMailaddrs in
325 *\"* | *\\*)
326 sed 's/["\\]/\\&/g' >$llogout <<EOF || exit
327 $loginFullnameMailaddrs
329 loginFullnameMailaddrs=`cat $llogout`
330 esac
332 oldIFS=$IFS
333 IFS=$nl
334 for loginFullnameMailaddr in $loginFullnameMailaddrs
336 case $loginFullnameMailaddr in
337 *"$tab"*) IFS=$tab;;
338 *) IFS=:
339 esac
340 set x $loginFullnameMailaddr
341 login=$2
342 fullname=$3
343 mailaddr=$4
344 initialize_fullname="$initialize_fullname
345 fullname[\"$login\"] = \"$fullname\""
346 initialize_mailaddr="$initialize_mailaddr
347 mailaddr[\"$login\"] = \"$mailaddr\""
348 done
349 IFS=$oldIFS
350 esac
352 case $llogout in
353 ?*) sort -u -o $llogout <<EOF || exit
354 $logins
356 esac
357 output_authors='/^date: / {
358 if ($2 ~ /^[0-9]*[-\/][0-9][0-9][-\/][0-9][0-9]$/ && $3 ~ /^[0-9][0-9]:[0-9][0-9]:[0-9][0-9][-+0-9:]*;$/ && $4 == "author:" && $5 ~ /^[^;]*;$/) {
359 print substr($5, 1, length($5)-1)
362 authors=`
363 $AWK "$output_authors" <$rlogout |
364 case $llogout in
365 '') sort -u;;
366 ?*) sort -u | comm -23 - $llogout
367 esac
369 case $authors in
371 cat >$llogout <<EOF || exit
372 $authors
374 initialize_author_script='s/["\\]/\\&/g; s/.*/author[\"&\"] = 1/'
375 initialize_author=`sed -e "$initialize_author_script" <$llogout`
376 awkscript='
377 BEGIN {
378 alphabet = "abcdefghijklmnopqrstuvwxyz"
379 ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
380 '"$initialize_author"'
383 if (author[$1]) {
384 fullname = $5
385 if (fullname ~ /[0-9]+-[^(]*\([0-9]+\)$/) {
386 # Remove the junk from fullnames like "0000-Admin(0000)".
387 fullname = substr(fullname, index(fullname, "-") + 1)
388 fullname = substr(fullname, 1, index(fullname, "(") - 1)
390 if (fullname ~ /,[^ ]/) {
391 # Some sites put comma-separated junk after the fullname.
392 # Remove it, but leave "Bill Gates, Jr" alone.
393 fullname = substr(fullname, 1, index(fullname, ",") - 1)
395 abbr = index(fullname, "&")
396 if (abbr) {
397 a = substr($1, 1, 1)
398 A = a
399 i = index(alphabet, a)
400 if (i) A = substr(ALPHABET, i, 1)
401 fullname = substr(fullname, 1, abbr-1) A substr($1, 2) substr(fullname, abbr+1)
404 # Quote quotes and backslashes properly in full names.
405 # Do not use gsub; traditional awk lacks it.
406 quoted = ""
407 rest = fullname
408 for (;;) {
409 p = index(rest, "\\")
410 q = index(rest, "\"")
411 if (p) {
412 if (q && q<p) p = q
413 } else {
414 if (!q) break
415 p = q
417 quoted = quoted substr(rest, 1, p-1) "\\" substr(rest, p, 1)
418 rest = substr(rest, p+1)
421 printf "fullname[\"%s\"] = \"%s%s\"\n", $1, quoted, rest
422 author[$1] = 0
427 initialize_fullname=`
429 (getent passwd $authors) ||
431 cat /etc/passwd
432 for author in $authors
433 do NIS_PATH= nismatch $author passwd.org_dir
434 done
435 ypmatch $authors passwd
437 } 2>/dev/null |
438 $AWK -F: "$awkscript"
439 `$initialize_fullname
440 esac
443 # Function to print a single log line.
444 # We don't use awk functions, to stay compatible with old awk versions.
445 # `Log' is the log message (with \n replaced by \001).
446 # `files' contains the affected files.
447 printlogline='{
449 # Following the GNU coding standards, rewrite
450 # * file: (function): comment
451 # to
452 # * file (function): comment
453 if (Log ~ /^\([^)]*\): /) {
454 i = index(Log, ")")
455 files = files " " substr(Log, 1, i)
456 Log = substr(Log, i+3)
459 # If "label: comment" is too long, break the line after the ":".
460 sep = " "
461 if ('"$length"' <= '"$indent"' + 1 + length(files) + index(Log, SOH)) sep = "\n" indent_string
463 # Print the label.
464 printf "%s*%s:", indent_string, files
466 # Print each line of the log, transliterating \001 to \n.
467 while ((i = index(Log, SOH)) != 0) {
468 logline = substr(Log, 1, i-1)
469 if (logline ~ /[^'"$tab"' ]/) {
470 printf "%s%s\n", sep, logline
471 } else {
472 print ""
474 sep = indent_string
475 Log = substr(Log, i+1)
479 # Pattern to match the `revision' line of rlog output.
480 rlog_revision_pattern='^revision [0-9]+\.[0-9]+(\.[0-9]+\.[0-9]+)*(['"$tab"' ]+locked by: [^'"$tab"' $,.0-9:;@]*[^'"$tab"' $,:;@][^'"$tab"' $,.0-9:;@]*;)?['"$tab"' ]*$'
482 case $hostname in
484 hostname=`(
485 hostname || uname -n || uuname -l || cat /etc/whoami
486 ) 2>/dev/null` || {
487 echo >&2 "$0: cannot deduce hostname"
488 exit 1
491 case $hostname in
492 *.*) ;;
494 domainname=`(domainname) 2>/dev/null` &&
495 case $domainname in
496 *.*) hostname=$hostname.$domainname
497 esac
498 esac
499 esac
502 # Process the rlog output, generating ChangeLog style entries.
504 # First, reformat the rlog output so that each line contains one log entry.
505 # Transliterate \n to \001 so that multiline entries fit on a single line.
506 # Discard irrelevant rlog output.
507 $AWK <$rlogout '
508 BEGIN { repository = "'"$repository"'" }
509 /^RCS file:/ {
510 if (repository != "") {
511 filename = $3
512 if (substr(filename, 1, length(repository) + 1) == repository "/") {
513 filename = substr(filename, length(repository) + 2)
515 if (filename ~ /,v$/) {
516 filename = substr(filename, 1, length(filename) - 2)
518 if (filename ~ /(^|\/)Attic\/[^\/]*$/) {
519 i = length(filename)
520 while (substr(filename, i, 1) != "/") i--
521 filename = substr(filename, 1, i - 6) substr(filename, i + 1)
524 rev = "?"
526 /^Working file:/ { if (repository == "") filename = $3 }
527 /'"$rlog_revision_pattern"'/, /^(-----------*|===========*)$/ {
528 line = $0
529 if (line ~ /'"$rlog_revision_pattern"'/) {
530 rev = $2
531 next
533 if (line ~ /^date: [0-9][- +\/0-9:]*;/) {
534 date = $2
535 if (date ~ /\//) {
536 # This is a traditional RCS format date YYYY/MM/DD.
537 # Replace "/"s with "-"s to get ISO format.
538 newdate = ""
539 while ((i = index(date, "/")) != 0) {
540 newdate = newdate substr(date, 1, i-1) "-"
541 date = substr(date, i+1)
543 date = newdate date
545 time = substr($3, 1, length($3) - 1)
546 author = substr($5, 1, length($5)-1)
547 printf "%s %s %s %s %s %c", filename, rev, date, time, author, 1
548 rev = "?"
549 next
551 if (line ~ /^branches: /) { next }
552 if (line ~ /^(-----------*|===========*)$/) { print ""; next }
553 if (line == "Initial revision" || line ~ /^file .+ was initially added on branch .+\.$/) {
554 line = "New file."
556 printf "%s%c", line, 1
560 # Now each line is of the form
561 # FILENAME REVISION YYYY-MM-DD HH:MM:SS[+-TIMEZONE] AUTHOR \001LOG
562 # where \001 stands for a carriage return,
563 # and each line of the log is terminated by \001 instead of \n.
564 # Sort the log entries, first by date+time (in reverse order),
565 # then by author, then by log entry, and finally by file name and revision
566 # (just in case).
567 sort +2 -4r +4 +0 |
569 # Finally, reformat the sorted log entries.
570 $AWK '
571 BEGIN {
572 logTZ = "'"$logTZ"'"
573 revision = "'"$revision"'"
575 # Some awk variants do not understand "\001", so we have to
576 # put the char directly in the file.
577 SOH="\x01" # <-- There is a single SOH (octal code 001) here.
579 # Initialize the fullname and mailaddr associative arrays.
580 '"$initialize_fullname"'
581 '"$initialize_mailaddr"'
583 # Initialize indent string.
584 indent_string = ""
585 i = '"$indent"'
586 if (0 < '"$tabwidth"')
587 for (; '"$tabwidth"' <= i; i -= '"$tabwidth"')
588 indent_string = indent_string "\t"
589 while (1 <= i--)
590 indent_string = indent_string " "
594 newlog = substr($0, 1 + index($0, SOH))
596 # Ignore log entries prefixed by "#".
597 if (newlog ~ /^#/) { next }
599 if (Log != newlog || date != $3 || author != $5) {
601 # The previous log and this log differ.
603 # Print the old log.
604 if (date != "") '"$printlogline"'
606 # Logs that begin with "{clumpname} " should be grouped together,
607 # and the clumpname should be removed.
608 # Extract the new clumpname from the log header,
609 # and use it to decide whether to output a blank line.
610 newclumpname = ""
611 sep = "\n"
612 if (date == "") sep = ""
613 if (newlog ~ /^\{[^'"$tab"' }]*}['"$tab"' ]/) {
614 i = index(newlog, "}")
615 newclumpname = substr(newlog, 1, i)
616 while (substr(newlog, i+1) ~ /^['"$tab"' ]/) i++
617 newlog = substr(newlog, i+1)
618 if (clumpname == newclumpname) sep = ""
620 printf sep
621 clumpname = newclumpname
623 # Get ready for the next log.
624 Log = newlog
625 if (files != "")
626 for (i in filesknown)
627 filesknown[i] = 0
628 files = ""
630 if (date != $3 || author != $5) {
631 # The previous date+author and this date+author differ.
632 # Print the new one.
633 date = $3
634 time = $4
635 author = $5
637 zone = ""
638 if (logTZ && ((i = index(time, "-")) || (i = index(time, "+"))))
639 zone = " " substr(time, i)
641 # Print "date[ timezone] fullname <email address>".
642 # Get fullname and email address from associative arrays;
643 # default to author and author@hostname if not in arrays.
644 if (fullname[author])
645 auth = fullname[author]
646 else
647 auth = author
648 printf "%s%s %s ", date, zone, auth
649 if (mailaddr[author])
650 printf "<%s>\n\n", mailaddr[author]
651 else
652 printf "<%s@%s>\n\n", author, "'"$hostname"'"
654 if (! filesknown[$1]) {
655 filesknown[$1] = 1
656 if (files == "") files = " " $1
657 else files = files ", " $1
658 if (revision && $2 != "?") files = files " " $2
661 END {
662 # Print the last log.
663 if (date != "") {
664 '"$printlogline"'
665 printf "\n"
668 ' &&
671 # Exit successfully.
673 exec rm -f $llogout $rlogout
675 # Local Variables:
676 # tab-width:4
677 # End: