lib-src/rcs2log

   1 #! /bin/sh
   2
   3 # RCS to ChangeLog generator
   4
   5 # Generate a change log prefix from RCS files and the ChangeLog (if any).
   6 # Output the new prefix to standard output.
   7 # You can edit this prefix by hand, and then prepend it to ChangeLog.
   8
   9 # Ignore log entries that start with `#'.
  10 # Clump together log entries that start with `{topic} ',
  11 # where `topic' contains neither white space nor `}'.
  12
  13 # Author: Paul Eggert <eggert@twinsun.com>
  14
  15 # $Id: rcs2log,v 1.29 1996/08/24 21:11:14 erik Exp eggert $
  16
  17 # Copyright 1992, 1993, 1994, 1995, 1996 Free Software Foundation, Inc.
  18
  19 # This program is free software; you can redistribute it and/or modify
  20 # it under the terms of the GNU General Public License as published by
  21 # the Free Software Foundation; either version 2, or (at your option)
  22 # any later version.
  23 #
  24 # This program is distributed in the hope that it will be useful,
  25 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  26 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  27 # GNU General Public License for more details.
  28 #
  29 # You should have received a copy of the GNU General Public License
  30 # along with this program; see the file COPYING.  If not, write to the
  31 # Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  32 # Boston, MA 02111-1307, USA.
  33
  34 tab='   '
  35 nl='
  36 '
  37
  38 # Parse options.
  39
  40 # defaults
  41 : ${AWK=awk}
  42 : ${TMPDIR=/tmp}
  43 changelog=ChangeLog # change log file name
  44 datearg= # rlog date option
  45 hostname= # name of local host (if empty, will deduce it later)
  46 indent=8 # indent of log line
  47 length=79 # suggested max width of log line
  48 logins= # login names for people we know fullnames and mailaddrs of
  49 loginFullnameMailaddrs= # login<tab>fullname<tab>mailaddr triplets
  50 logTZ= # time zone for log dates (if empty, use local time)
  51 recursive= # t if we want recursive rlog
  52 revision= # t if we want revision numbers
  53 rlog_options= # options to pass to rlog
  54 tabwidth=8 # width of horizontal tab
  55
  56 while :
  57 do
  58         case $1 in
  59         -c)     changelog=${2?}; shift;;
  60         -i)     indent=${2?}; shift;;
  61         -h)     hostname=${2?}; shift;;
  62         -l)     length=${2?}; shift;;
  63         -[nu])  # -n is obsolescent; it is replaced by -u.
  64                 case $1 in
  65                 -n)     case ${2?}${3?}${4?} in
  66                         *"$tab"* | *"$nl"*)
  67                                 echo >&2 "$0: -n '$2' '$3' '$4': tabs, newlines not allowed"
  68                                 exit 1
  69                         esac
  70                         loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2$tab$3$tab$4
  71                         shift; shift; shift;;
  72                 -u)
  73                         # If $2 is not tab-separated, use colon for separator.
  74                         case ${2?} in
  75                         *"$nl"*)
  76                                 echo >&2 "$0: -u '$2': newlines not allowed"
  77                                 exit 1;;
  78                         *"$tab"*)
  79                                 t=$tab;;
  80                         *)
  81                                 t=:
  82                         esac
  83                         case $2 in
  84                         *"$t"*"$t"*"$t"*)
  85                                 echo >&2 "$0: -u '$2': too many fields"
  86                                 exit 1;;
  87                         *"$t"*"$t"*)
  88                                 ;;
  89                         *)
  90                                 echo >&2 "$0: -u '$2': not enough fields"
  91                                 exit 1
  92                         esac
  93                         loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2
  94                         shift
  95                 esac
  96                 logins=$logins$nl$login
  97                 ;;
  98         -r)     rlog_options=$rlog_options$nl${2?}; shift;;
  99         -R)     recursive=t;;
 100         -t)     tabwidth=${2?}; shift;;
 101         -v)     revision=t;;
 102         -*)     echo >&2 "$0: usage: $0 [options] [file ...]
 103 Options:
 104         [-c changelog] [-h hostname] [-i indent] [-l length] [-R]
 105         [-r rlog_option] [-t tabwidth] [-v]
 106         [-u 'login<TAB>fullname<TAB>mailaddr']..."
 107                 exit 1;;
 108         *)      break
 109         esac
 110         shift
 111 done
 112
 113 month_data='
 114         m[0]="Jan"; m[1]="Feb"; m[2]="Mar"
 115         m[3]="Apr"; m[4]="May"; m[5]="Jun"
 116         m[6]="Jul"; m[7]="Aug"; m[8]="Sep"
 117         m[9]="Oct"; m[10]="Nov"; m[11]="Dec"
 118 '
 119
 120
 121 # Put rlog output into $rlogout.
 122
 123 # If no rlog options are given,
 124 # log the revisions checked in since the first ChangeLog entry.
 125 # Since ChangeLog is only by date, some of these revisions may be duplicates of
 126 # what's already in ChangeLog; it's the user's responsibility to remove them.
 127 case $rlog_options in
 128 '')
 129         date=1970
 130         if test -s "$changelog"
 131         then
 132                 e='
 133                         /^[0-9]+-[0-9][0-9]-[0-9][0-9]/{
 134                                 # ISO 8601 date
 135                                 print $1
 136                                 exit
 137                         }
 138                         /^... ... [ 0-9][0-9] [ 0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9]+ /{
 139                                 # old-fashioned date and time (Emacs 19.31 and earlier)
 140                                 '"$month_data"'
 141                                 year = $5
 142                                 for (i=0; i<=11; i++) if (m[i] == $2) break
 143                                 dd = $3
 144                                 printf "%d-%02d-%02d\n", year, i+1, dd
 145                                 exit
 146                         }
 147                 '
 148                 d=`$AWK "$e" <"$changelog"` || exit
 149                 case $d in
 150                 ?*) date=$d
 151                 esac
 152         fi
 153         datearg="-d>$date"
 154 esac
 155
 156 # Use rlog's -zLT option, if rlog supports it.
 157 case `rlog -zLT 2>&1` in
 158 *'unknown option'*) ;;
 159 *) rlog_options=-zLT$nl$rlog_options
 160 esac
 161
 162 # Use TZ specified by ChangeLog local variable, if any.
 163 if test -s "$changelog"
 164 then
 165         extractTZ='
 166                 /^.*change-log-time-zone-rule['"$tab"' ]*:['"$tab"' ]*"\([^"]*\)".*/{
 167                         s//\1/; p; q
 168                 }
 169                 /^.*change-log-time-zone-rule['"$tab"' ]*:['"$tab"' ]*t.*/{
 170                         s//UTC0/; p; q
 171                 }
 172         '
 173         logTZ=`tail "$changelog" | sed -n "$extractTZ"`
 174         case $logTZ in
 175         ?*) TZ=$logTZ; export TZ
 176         esac
 177 fi
 178
 179 # If CVS is in use, examine its repository, not the normal RCS files.
 180 if test ! -f CVS/Repository
 181 then
 182         rlog=rlog
 183         repository=
 184 else
 185         rlog='cvs log'
 186         repository=`sed 1q <CVS/Repository` || exit
 187         test ! -f CVS/Root || CVSROOT=`cat <CVS/Root` || exit
 188         case $CVSROOT in
 189         *:/*)
 190                 # remote repository
 191                 ;;
 192         *)
 193                 # local repository
 194                 case $repository in
 195                 /*) ;;
 196                 *) repository=${CVSROOT?}/$repository
 197                 esac
 198                 if test ! -d "$repository"
 199                 then
 200                         echo >&2 "$0: $repository: bad repository (see CVS/Repository)"
 201                         exit 1
 202                 fi
 203         esac
 204 fi
 205
 206 # With no arguments, examine all files under the RCS directory.
 207 case $# in
 208 0)
 209         case $repository in
 210         '')
 211                 oldIFS=$IFS
 212                 IFS=$nl
 213                 case $recursive in
 214                 t)
 215                         RCSdirs=`find . -name RCS -type d -print`
 216                         filesFromRCSfiles='s|,v$||; s|/RCS/|/|; s|^\./||'
 217                         files=`
 218                                 {
 219                                         case $RCSdirs in
 220                                         ?*) find $RCSdirs -type f -print
 221                                         esac
 222                                         find . -name '*,v' -print
 223                                 } |
 224                                 sort -u |
 225                                 sed "$filesFromRCSfiles"
 226                         `;;
 227                 *)
 228                         files=
 229                         for file in RCS/.* RCS/* .*,v *,v
 230                         do
 231                                 case $file in
 232                                 RCS/. | RCS/..) continue;;
 233                                 RCS/.\* | RCS/\* | .\*,v | \*,v) test -f "$file" || continue
 234                                 esac
 235                                 files=$files$nl$file
 236                         done
 237                         case $files in
 238                         '') exit 0
 239                         esac
 240                 esac
 241                 set x $files
 242                 shift
 243                 IFS=$oldIFS
 244         esac
 245 esac
 246
 247 llogout=$TMPDIR/rcs2log$$l
 248 rlogout=$TMPDIR/rcs2log$$r
 249 trap exit 1 2 13 15
 250 trap "rm -f $llogout $rlogout; exit 1" 0
 251
 252 case $datearg in
 253 ?*) $rlog $rlog_options "$datearg" ${1+"$@"} >$rlogout;;
 254 '') $rlog $rlog_options ${1+"$@"} >$rlogout
 255 esac || exit
 256
 257
 258 # Get the full name of each author the logs mention, and set initialize_fullname
 259 # to awk code that initializes the `fullname' awk associative array.
 260 # Warning: foreign authors (i.e. not known in the passwd file) are mishandled;
 261 # you have to fix the resulting output by hand.
 262
 263 initialize_fullname=
 264 initialize_mailaddr=
 265
 266 case $loginFullnameMailaddrs in
 267 ?*)
 268         case $loginFullnameMailaddrs in
 269         *\"* | *\\*)
 270                 sed 's/["\\]/\\&/g' >$llogout <<EOF || exit
 271 $loginFullnameMailaddrs
 272 EOF
 273                 loginFullnameMailaddrs=`cat $llogout`
 274         esac
 275
 276         oldIFS=$IFS
 277         IFS=$nl
 278         for loginFullnameMailaddr in $loginFullnameMailaddrs
 279         do
 280                 case $loginFullnameMailaddr in
 281                 *"$tab"*) IFS=$tab;;
 282                 *) IFS=:
 283                 esac
 284                 set x $loginFullnameMailaddr
 285                 login=$2
 286                 fullname=$3
 287                 mailaddr=$4
 288                 initialize_fullname="$initialize_fullname
 289                         fullname[\"$login\"] = \"$fullname\""
 290                 initialize_mailaddr="$initialize_mailaddr
 291                         mailaddr[\"$login\"] = \"$mailaddr\""
 292         done
 293         IFS=$oldIFS
 294 esac
 295
 296 case $llogout in
 297 ?*) sort -u -o $llogout <<EOF || exit
 298 $logins
 299 EOF
 300 esac
 301 output_authors='/^date: / {
 302         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:" && $5 ~ /^[^;]*;$/) {
 303                 print substr($5, 1, length($5)-1)
 304         }
 305 }'
 306 authors=`
 307         $AWK "$output_authors" <$rlogout |
 308         case $llogout in
 309         '') sort -u;;
 310         ?*) sort -u | comm -23 - $llogout
 311         esac
 312 `
 313 case $authors in
 314 ?*)
 315         cat >$llogout <<EOF || exit
 316 $authors
 317 EOF
 318         initialize_author_script='s/["\\]/\\&/g; s/.*/author[\"&\"] = 1/'
 319         initialize_author=`sed -e "$initialize_author_script" <$llogout`
 320         awkscript='
 321                 BEGIN {
 322                         alphabet = "abcdefghijklmnopqrstuvwxyz"
 323                         ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 324                         '"$initialize_author"'
 325                 }
 326                 {
 327                         if (author[$1]) {
 328                                 fullname = $5
 329                                 if (fullname ~ /[0-9]+-[^(]*\([0-9]+\)$/) {
 330                                         # Remove the junk from fullnames like "0000-Admin(0000)".
 331                                         fullname = substr(fullname, index(fullname, "-") + 1)
 332                                         fullname = substr(fullname, 1, index(fullname, "(") - 1)
 333                                 }
 334                                 if (fullname ~ /,[^ ]/) {
 335                                         # Some sites put comma-separated junk after the fullname.
 336                                         # Remove it, but leave "Bill Gates, Jr" alone.
 337                                         fullname = substr(fullname, 1, index(fullname, ",") - 1)
 338                                 }
 339                                 abbr = index(fullname, "&")
 340                                 if (abbr) {
 341                                         a = substr($1, 1, 1)
 342                                         A = a
 343                                         i = index(alphabet, a)
 344                                         if (i) A = substr(ALPHABET, i, 1)
 345                                         fullname = substr(fullname, 1, abbr-1) A substr($1, 2) substr(fullname, abbr+1)
 346                                 }
 347
 348                                 # Quote quotes and backslashes properly in full names.
 349                                 # Do not use gsub; traditional awk lacks it.
 350                                 quoted = ""
 351                                 rest = fullname
 352                                 for (;;) {
 353                                         p = index(rest, "\\")
 354                                         q = index(rest, "\"")
 355                                         if (p) {
 356                                                 if (q && q<p) p = q
 357                                         } else {
 358                                                 if (!q) break
 359                                                 p = q
 360                                         }
 361                                         quoted = quoted substr(rest, 1, p-1) "\\" substr(rest, p, 1)
 362                                         rest = substr(rest, p+1)
 363                                 }
 364
 365                                 printf "fullname[\"%s\"] = \"%s%s\"\n", $1, quoted, rest
 366                                 author[$1] = 0
 367                         }
 368                 }
 369         '
 370
 371         initialize_fullname=`
 372                 (
 373                         cat /etc/passwd
 374                         for author in $authors
 375                         do nismatch $author passwd.org_dir
 376                         done
 377                         ypmatch $authors passwd
 378                 ) 2>/dev/null |
 379                 $AWK -F: "$awkscript"
 380         `$initialize_fullname
 381 esac
 382
 383
 384 # Function to print a single log line.
 385 # We don't use awk functions, to stay compatible with old awk versions.
 386 # `Log' is the log message (with \n replaced by \r).
 387 # `files' contains the affected files.
 388 printlogline='{
 389
 390         # Following the GNU coding standards, rewrite
 391         #       * file: (function): comment
 392         # to
 393         #       * file (function): comment
 394         if (Log ~ /^\([^)]*\): /) {
 395                 i = index(Log, ")")
 396                 files = files " " substr(Log, 1, i)
 397                 Log = substr(Log, i+3)
 398         }
 399
 400         # If "label: comment" is too long, break the line after the ":".
 401         sep = " "
 402         if ('"$length"' <= '"$indent"' + 1 + length(files) + index(Log, CR)) sep = "\n" indent_string
 403
 404         # Print the label.
 405         printf "%s*%s:", indent_string, files
 406
 407         # Print each line of the log, transliterating \r to \n.
 408         while ((i = index(Log, CR)) != 0) {
 409                 logline = substr(Log, 1, i-1)
 410                 if (logline ~ /[^'"$tab"' ]/) {
 411                         printf "%s%s\n", sep, logline
 412                 } else {
 413                         print ""
 414                 }
 415                 sep = indent_string
 416                 Log = substr(Log, i+1)
 417         }
 418 }'
 419
 420 # Pattern to match the `revision' line of rlog output.
 421 rlog_revision_pattern='^revision [0-9]+\.[0-9]+(\.[0-9]+\.[0-9]+)*(['"$tab"' ]+locked by: [^'"$tab"' $,.0-9:;@]*[^'"$tab"' $,:;@][^'"$tab"' $,.0-9:;@]*;)?['"$tab"' ]*$'
 422
 423 case $hostname in
 424 '')
 425         hostname=`(
 426                 hostname || uname -n || uuname -l || cat /etc/whoami
 427         ) 2>/dev/null` || {
 428                 echo >&2 "$0: cannot deduce hostname"
 429                 exit 1
 430         }
 431
 432         case $hostname in
 433         *.*) ;;
 434         *)
 435                 domainname=`(domainname) 2>/dev/null` &&
 436                 case $domainname in
 437                 *.*) hostname=$hostname.$domainname
 438                 esac
 439         esac
 440 esac
 441
 442
 443 # Process the rlog output, generating ChangeLog style entries.
 444
 445 # First, reformat the rlog output so that each line contains one log entry.
 446 # Transliterate \n to \r so that multiline entries fit on a single line.
 447 # Discard irrelevant rlog output.
 448 $AWK <$rlogout '
 449         BEGIN { repository = "'"$repository"'" }
 450         /^RCS file:/ {
 451                 if (repository != "") {
 452                         filename = $3
 453                         if (substr(filename, 1, length(repository) + 1) == repository "/") {
 454                                 filename = substr(filename, length(repository) + 2)
 455                         }
 456                         if (filename ~ /,v$/) {
 457                                 filename = substr(filename, 1, length(filename) - 2)
 458                         }
 459                 }
 460                 rev = "?"
 461         }
 462         /^Working file:/ { if (repository == "") filename = $3 }
 463         /'"$rlog_revision_pattern"'/, /^(-----------*|===========*)$/ {
 464                 if ($0 ~ /'"$rlog_revision_pattern"'/) {
 465                         rev = $2
 466                         next
 467                 }
 468                 if ($0 ~ /^date: [0-9][- +\/0-9:]*;/) {
 469                         date = $2
 470                         if (date ~ /\//) {
 471                                 # This is a traditional RCS format date YYYY/MM/DD.
 472                                 # Replace "/"s with "-"s to get ISO format.
 473                                 newdate = ""
 474                                 while ((i = index(date, "/")) != 0) {
 475                                         newdate = newdate substr(date, 1, i-1) "-"
 476                                         date = substr(date, i+1)
 477                                 }
 478                                 date = newdate date
 479                         }
 480                         time = substr($3, 1, length($3) - 1)
 481                         author = substr($5, 1, length($5)-1)
 482                         printf "%s %s %s %s %s %c", filename, rev, date, time, author, 13
 483                         rev = "?"
 484                         next
 485                 }
 486                 if ($0 ~ /^branches: /) { next }
 487                 if ($0 ~ /^(-----------*|===========*)$/) { print ""; next }
 488                 printf "%s%c", $0, 13
 489         }
 490 ' |
 491
 492 # Now each line is of the form
 493 # FILENAME REVISION YYYY-MM-DD HH:MM:SS[+-TIMEZONE] AUTHOR \rLOG
 494 #       where \r stands for a carriage return,
 495 #       and each line of the log is terminated by \r instead of \n.
 496 # Sort the log entries, first by date+time (in reverse order),
 497 # then by author, then by log entry, and finally by file name and revision
 498 # (just in case).
 499 sort +2 -4r +4 +0 |
 500
 501 # Finally, reformat the sorted log entries.
 502 $AWK '
 503         BEGIN {
 504                 logTZ = "'"$logTZ"'"
 505                 revision = "'"$revision"'"
 506
 507                 # Some awk variants do not understand "\r" or "\013", so we have to
 508                 # put a carriage return directly in the file.
 509                 CR="
" # <-- There is a single CR between the " chars here.
 510
 511                 # Initialize the fullname and mailaddr associative arrays.
 512                 '"$initialize_fullname"'
 513                 '"$initialize_mailaddr"'
 514
 515                 # Initialize indent string.
 516                 indent_string = ""
 517                 i = '"$indent"'
 518                 if (0 < '"$tabwidth"')
 519                         for (;  '"$tabwidth"' <= i;  i -= '"$tabwidth"')
 520                                 indent_string = indent_string "\t"
 521                 while (1 <= i--)
 522                         indent_string = indent_string " "
 523         }
 524
 525         {
 526                 newlog = substr($0, 1 + index($0, CR))
 527
 528                 # Ignore log entries prefixed by "#".
 529                 if (newlog ~ /^#/) { next }
 530
 531                 if (Log != newlog || date != $3 || author != $5) {
 532
 533                         # The previous log and this log differ.
 534
 535                         # Print the old log.
 536                         if (date != "") '"$printlogline"'
 537
 538                         # Logs that begin with "{clumpname} " should be grouped together,
 539                         # and the clumpname should be removed.
 540                         # Extract the new clumpname from the log header,
 541                         # and use it to decide whether to output a blank line.
 542                         newclumpname = ""
 543                         sep = "\n"
 544                         if (date == "") sep = ""
 545                         if (newlog ~ /^\{[^'"$tab"' }]*}['"$tab"' ]/) {
 546                                 i = index(newlog, "}")
 547                                 newclumpname = substr(newlog, 1, i)
 548                                 while (substr(newlog, i+1) ~ /^['"$tab"' ]/) i++
 549                                 newlog = substr(newlog, i+1)
 550                                 if (clumpname == newclumpname) sep = ""
 551                         }
 552                         printf sep
 553                         clumpname = newclumpname
 554
 555                         # Get ready for the next log.
 556                         Log = newlog
 557                         if (files != "")
 558                                 for (i in filesknown)
 559                                         filesknown[i] = 0
 560                         files = ""
 561                 }
 562                 if (date != $3  ||  author != $5) {
 563                         # The previous date+author and this date+author differ.
 564                         # Print the new one.
 565                         date = $3
 566                         time = $4
 567                         author = $5
 568
 569                         zone = ""
 570                         if (logTZ && ((i = index(time, "-")) || (i = index(time, "+"))))
 571                                 zone = " " substr(time, i)
 572
 573                         # Print "date[ timezone]  fullname  <email address>".
 574                         # Get fullname and email address from associative arrays;
 575                         # default to author and author@hostname if not in arrays.
 576                         if (fullname[author])
 577                                 auth = fullname[author]
 578                         else
 579                                 auth = author
 580                         printf "%s%s  %s  ", date, zone, auth
 581                         if (mailaddr[author])
 582                                 printf "<%s>\n\n", mailaddr[author]
 583                         else
 584                                 printf "<%s@%s>\n\n", author, "'"$hostname"'"
 585                 }
 586                 if (! filesknown[$1]) {
 587                         filesknown[$1] = 1
 588                         if (files == "") files = " " $1
 589                         else files = files ", " $1
 590                         if (revision && $2 != "?") files = files " " $2
 591                 }
 592         }
 593         END {
 594                 # Print the last log.
 595                 if (date != "") {
 596                         '"$printlogline"'
 597                         printf "\n"
 598                 }
 599         }
 600 ' &&
 601
 602
 603 # Exit successfully.
 604
 605 exec rm -f $llogout $rlogout
 606
 607 # Local Variables:
 608 # tab-width:4
 609 # End: