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 1.27 1996/01/15 01:17:03 eggert Exp $
  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
  31 # the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
  32
  33 tab='   '
  34 nl='
  35 '
  36
  37 # Parse options.
  38
  39 # defaults
  40 : ${AWK=awk}
  41 : ${TMPDIR=/tmp}
  42 hostname= # name of local host (if empty, will deduce it later)
  43 indent=8 # indent of log line
  44 length=79 # suggested max width of log line
  45 logins= # login names for people we know fullnames and mailaddrs of
  46 loginFullnameMailaddrs= # login<tab>fullname<tab>mailaddr triplets
  47 recursive= # t if we want recursive rlog
  48 rlog_options= # options to pass to rlog
  49 tabwidth=8 # width of horizontal tab
  50
  51 while :
  52 do
  53         case $1 in
  54         -i)     indent=${2?}; shift;;
  55         -h)     hostname=${2?}; shift;;
  56         -l)     length=${2?}; shift;;
  57         -[nu])  # -n is obsolescent; it is replaced by -u.
  58                 case $1 in
  59                 -n)     case ${2?}${3?}${4?} in
  60                         *"$tab"* | *"$nl"*)
  61                                 echo >&2 "$0: -n '$2' '$3' '$4': tabs, newlines not allowed"
  62                                 exit 1
  63                         esac
  64                         loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2$tab$3$tab$4
  65                         shift; shift; shift;;
  66                 -u)
  67                         # If $2 is not tab-separated, use colon for separator.
  68                         case ${2?} in
  69                         *"$nl"*)
  70                                 echo >&2 "$0: -u '$2': newlines not allowed"
  71                                 exit 1;;
  72                         *"$tab"*)
  73                                 t=$tab;;
  74                         *)
  75                                 t=:
  76                         esac
  77                         case $2 in
  78                         *"$t"*"$t"*"$t"*)
  79                                 echo >&2 "$0: -u '$2': too many fields"
  80                                 exit 1;;
  81                         *"$t"*"$t"*)
  82                                 ;;
  83                         *)
  84                                 echo >&2 "$0: -u '$2': not enough fields"
  85                                 exit 1
  86                         esac
  87                         loginFullnameMailaddrs=$loginFullnameMailaddrs$nl$2
  88                         shift
  89                 esac
  90                 logins=$logins$nl$login
  91                 ;;
  92         -r)     rlog_options=$rlog_options$nl${2?}; shift;;
  93         -R)     recursive=t;;
  94         -t)     tabwidth=${2?}; shift;;
  95         -*)     echo >&2 "$0: usage: $0 [options] [file ...]
  96 Options:
  97         [-h hostname] [-i indent] [-l length] [-R] [-r rlog_option]
  98         [-t tabwidth] [-u 'login<TAB>fullname<TAB>mailaddr']..."
  99                 exit 1;;
 100         *)      break
 101         esac
 102         shift
 103 done
 104
 105 month_data='
 106         m[0]="Jan"; m[1]="Feb"; m[2]="Mar"
 107         m[3]="Apr"; m[4]="May"; m[5]="Jun"
 108         m[6]="Jul"; m[7]="Aug"; m[8]="Sep"
 109         m[9]="Oct"; m[10]="Nov"; m[11]="Dec"
 110
 111         # days in non-leap year thus far, indexed by month (0-12)
 112         mo[0]=0; mo[1]=31; mo[2]=59; mo[3]=90
 113         mo[4]=120; mo[5]=151; mo[6]=181; mo[7]=212
 114         mo[8]=243; mo[9]=273; mo[10]=304; mo[11]=334
 115         mo[12]=365
 116 '
 117
 118
 119 # Put rlog output into $rlogout.
 120
 121 # If no rlog options are given,
 122 # log the revisions checked in since the first ChangeLog entry.
 123 case $rlog_options in
 124 '')
 125         date=1970
 126         if test -s ChangeLog
 127         then
 128                 # Add 1 to seconds to avoid duplicating most recent log.
 129                 e='
 130                         /^... ... [ 0-9][0-9] [ 0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9]+ /{
 131                                 '"$month_data"'
 132                                 year = $5
 133                                 for (i=0; i<=11; i++) if (m[i] == $2) break
 134                                 dd = $3
 135                                 hh = substr($0,12,2)
 136                                 mm = substr($0,15,2)
 137                                 ss = substr($0,18,2)
 138                                 ss++
 139                                 if (ss == 60) {
 140                                         ss = 0
 141                                         mm++
 142                                         if (mm == 60) {
 143                                                 mm = 0
 144                                                 hh++
 145                                                 if (hh == 24) {
 146                                                         hh = 0
 147                                                         dd++
 148                                                         monthdays = mo[i+1] - mo[i]
 149                                                         if (i == 1 && year%4 == 0 && (year%100 != 0 || year%400 == 0)) monthdays++
 150                                                         if (dd == monthdays + 1) {
 151                                                                 dd = 1
 152                                                                 i++
 153                                                                 if (i == 12) {
 154                                                                         i = 0
 155                                                                         year++
 156                                                                 }
 157                                                         }
 158                                                 }
 159                                         }
 160                                 }
 161                                 # Output comma instead of space to avoid CVS 1.5 bug.
 162                                 printf "%d/%02d/%02d,%02d:%02d:%02d\n", year,i+1,dd,hh,mm,ss
 163                                 exit
 164                         }
 165                 '
 166                 d=`$AWK "$e" <ChangeLog` || exit
 167                 case $d in
 168                 ?*) date=$d
 169                 esac
 170         fi
 171         datearg="-d>$date"
 172 esac
 173
 174 # If CVS is in use, examine its repository, not the normal RCS files.
 175 if test ! -f CVS/Repository
 176 then
 177         rlog=rlog
 178         repository=
 179 else
 180         rlog='cvs log'
 181         repository=`sed 1q <CVS/Repository` || exit
 182         test ! -f CVS/Root || CVSROOT=`cat <CVS/Root` || exit
 183         case $CVSROOT in
 184         *:/*)
 185                 # remote repository
 186                 ;;
 187         *)
 188                 # local repository
 189                 case $repository in
 190                 /*) ;;
 191                 *) repository=${CVSROOT?}/$repository
 192                 esac
 193                 if test ! -d "$repository"
 194                 then
 195                         echo >&2 "$0: $repository: bad repository (see CVS/Repository)"
 196                         exit 1
 197                 fi
 198         esac
 199 fi
 200
 201 # With no arguments, examine all files under the RCS directory.
 202 case $# in
 203 0)
 204         case $repository in
 205         '')
 206                 oldIFS=$IFS
 207                 IFS=$nl
 208                 case $recursive in
 209                 t)
 210                         RCSdirs=`find . -name RCS -type d -print`
 211                         filesFromRCSfiles='s|,v$||; s|/RCS/|/|; s|^\./||'
 212                         files=`
 213                                 {
 214                                         case $RCSdirs in
 215                                         ?*) find $RCSdirs -type f -print
 216                                         esac
 217                                         find . -name '*,v' -print
 218                                 } |
 219                                 sort -u |
 220                                 sed "$filesFromRCSfiles"
 221                         `;;
 222                 *)
 223                         files=
 224                         for file in RCS/.* RCS/* .*,v *,v
 225                         do
 226                                 case $file in
 227                                 RCS/. | RCS/..) continue;;
 228                                 RCS/.\* | RCS/\* | .\*,v | \*,v) test -f "$file" || continue
 229                                 esac
 230                                 files=$files$nl$file
 231                         done
 232                         case $files in
 233                         '') exit 0
 234                         esac
 235                 esac
 236                 set x $files
 237                 shift
 238                 IFS=$oldIFS
 239         esac
 240 esac
 241
 242 llogout=$TMPDIR/rcs2log$$l
 243 rlogout=$TMPDIR/rcs2log$$r
 244 trap exit 1 2 13 15
 245 trap "rm -f $llogout $rlogout; exit 1" 0
 246
 247 case $rlog_options in
 248 ?*) $rlog $rlog_options ${1+"$@"} >$rlogout;;
 249 '') $rlog "$datearg" ${1+"$@"} >$rlogout
 250 esac || exit
 251
 252
 253 # Get the full name of each author the logs mention, and set initialize_fullname
 254 # to awk code that initializes the `fullname' awk associative array.
 255 # Warning: foreign authors (i.e. not known in the passwd file) are mishandled;
 256 # you have to fix the resulting output by hand.
 257
 258 initialize_fullname=
 259 initialize_mailaddr=
 260
 261 case $loginFullnameMailaddrs in
 262 ?*)
 263         case $loginFullnameMailaddrs in
 264         *\"* | *\\*)
 265                 sed 's/["\\]/\\&/g' >$llogout <<EOF || exit
 266 $loginFullnameMailaddrs
 267 EOF
 268                 loginFullnameMailaddrs=`cat $llogout`
 269         esac
 270
 271         oldIFS=$IFS
 272         IFS=$nl
 273         for loginFullnameMailaddr in $loginFullnameMailaddrs
 274         do
 275                 case $loginFullnameMailaddr in
 276                 *"$tab"*) IFS=$tab;;
 277                 *) IFS=:
 278                 esac
 279                 set x $loginFullnameMailaddr
 280                 login=$2
 281                 fullname=$3
 282                 mailaddr=$4
 283                 initialize_fullname="$initialize_fullname
 284                         fullname[\"$login\"] = \"$fullname\""
 285                 initialize_mailaddr="$initialize_mailaddr
 286                         mailaddr[\"$login\"] = \"$mailaddr\""
 287         done
 288         IFS=$oldIFS
 289 esac
 290
 291 case $llogout in
 292 ?*) sort -u -o $llogout <<EOF || exit
 293 $logins
 294 EOF
 295 esac
 296 output_authors='/^date: / {
 297         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:" && $5 ~ /^[^;]*;$/) {
 298                 print substr($5, 1, length($5)-1)
 299         }
 300 }'
 301 authors=`
 302         $AWK "$output_authors" <$rlogout |
 303         case $llogout in
 304         '') sort -u;;
 305         ?*) sort -u | comm -23 - $llogout
 306         esac
 307 `
 308 case $authors in
 309 ?*)
 310         cat >$llogout <<EOF || exit
 311 $authors
 312 EOF
 313         initialize_author_script='s/["\\]/\\&/g; s/.*/author[\"&\"] = 1/'
 314         initialize_author=`sed -e "$initialize_author_script" <$llogout`
 315         awkscript='
 316                 BEGIN {
 317                         alphabet = "abcdefghijklmnopqrstuvwxyz"
 318                         ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 319                         '"$initialize_author"'
 320                 }
 321                 {
 322                         if (author[$1]) {
 323                                 fullname = $5
 324                                 if (fullname ~ /[0-9]+-[^(]*\([0-9]+\)$/) {
 325                                         # Remove the junk from fullnames like "0000-Admin(0000)".
 326                                         fullname = substr(fullname, index(fullname, "-") + 1)
 327                                         fullname = substr(fullname, 1, index(fullname, "(") - 1)
 328                                 }
 329                                 if (fullname ~ /,[^ ]/) {
 330                                         # Some sites put comma-separated junk after the fullname.
 331                                         # Remove it, but leave "Bill Gates, Jr" alone.
 332                                         fullname = substr(fullname, 1, index(fullname, ",") - 1)
 333                                 }
 334                                 abbr = index(fullname, "&")
 335                                 if (abbr) {
 336                                         a = substr($1, 1, 1)
 337                                         A = a
 338                                         i = index(alphabet, a)
 339                                         if (i) A = substr(ALPHABET, i, 1)
 340                                         fullname = substr(fullname, 1, abbr-1) A substr($1, 2) substr(fullname, abbr+1)
 341                                 }
 342
 343                                 # Quote quotes and backslashes properly in full names.
 344                                 # Do not use gsub; traditional awk lacks it.
 345                                 quoted = ""
 346                                 rest = fullname
 347                                 for (;;) {
 348                                         p = index(rest, "\\")
 349                                         q = index(rest, "\"")
 350                                         if (p) {
 351                                                 if (q && q<p) p = q
 352                                         } else {
 353                                                 if (!q) break
 354                                                 p = q
 355                                         }
 356                                         quoted = quoted substr(rest, 1, p-1) "\\" substr(rest, p, 1)
 357                                         rest = substr(rest, p+1)
 358                                 }
 359
 360                                 printf "fullname[\"%s\"] = \"%s%s\"\n", $1, quoted, rest
 361                                 author[$1] = 0
 362                         }
 363                 }
 364         '
 365
 366         initialize_fullname=`
 367                 (
 368                         cat /etc/passwd
 369                         for author in $authors
 370                         do nismatch $author passwd.org_dir
 371                         done
 372                         ypmatch $authors passwd
 373                 ) 2>/dev/null |
 374                 $AWK -F: "$awkscript"
 375         `$initialize_fullname
 376 esac
 377
 378
 379 # Function to print a single log line.
 380 # We don't use awk functions, to stay compatible with old awk versions.
 381 # `Log' is the log message (with \n replaced by \r).
 382 # `files' contains the affected files.
 383 printlogline='{
 384
 385         # Following the GNU coding standards, rewrite
 386         #       * file: (function): comment
 387         # to
 388         #       * file (function): comment
 389         if (Log ~ /^\([^)]*\): /) {
 390                 i = index(Log, ")")
 391                 files = files " " substr(Log, 1, i)
 392                 Log = substr(Log, i+3)
 393         }
 394
 395         # If "label: comment" is too long, break the line after the ":".
 396         sep = " "
 397         if ('"$length"' <= '"$indent"' + 1 + length(files) + index(Log, CR)) sep = "\n" indent_string
 398
 399         # Print the label.
 400         printf "%s*%s:", indent_string, files
 401
 402         # Print each line of the log, transliterating \r to \n.
 403         while ((i = index(Log, CR)) != 0) {
 404                 logline = substr(Log, 1, i-1)
 405                 if (logline ~ /[^'"$tab"' ]/) {
 406                         printf "%s%s\n", sep, logline
 407                 } else {
 408                         print ""
 409                 }
 410                 sep = indent_string
 411                 Log = substr(Log, i+1)
 412         }
 413 }'
 414
 415 case $hostname in
 416 '')
 417         hostname=`(
 418                 hostname || uname -n || uuname -l || cat /etc/whoami
 419         ) 2>/dev/null` || {
 420                 echo >&2 "$0: cannot deduce hostname"
 421                 exit 1
 422         }
 423
 424         case $hostname in
 425         *.*) ;;
 426         *)
 427                 domainname=`(domainname) 2>/dev/null` &&
 428                 case $domainname in
 429                 *.*) hostname=$hostname.$domainname
 430                 esac
 431         esac
 432 esac
 433
 434
 435 # Process the rlog output, generating ChangeLog style entries.
 436
 437 # First, reformat the rlog output so that each line contains one log entry.
 438 # Transliterate \n to \r so that multiline entries fit on a single line.
 439 # Discard irrelevant rlog output.
 440 $AWK <$rlogout '
 441         BEGIN { repository = "'"$repository"'" }
 442         /^RCS file:/ {
 443                 if (repository != "") {
 444                         filename = $3
 445                         if (substr(filename, 1, length(repository) + 1) == repository "/") {
 446                                 filename = substr(filename, length(repository) + 2)
 447                         }
 448                         if (filename ~ /,v$/) {
 449                                 filename = substr(filename, 1, length(filename) - 2)
 450                         }
 451                 }
 452         }
 453         /^Working file:/ { if (repository == "") filename = $3 }
 454         /^date: /, /^(-----------*|===========*)$/ {
 455                 if ($0 ~ /^branches: /) { next }
 456                 if ($0 ~ /^date: [0-9][- +\/0-9:]*;/) {
 457                         date = $2
 458                         if (date ~ /-/) {
 459                                 # An ISO format date.  Replace all "-"s with "/"s.
 460                                 newdate = ""
 461                                 while ((i = index(date, "-")) != 0) {
 462                                         newdate = newdate substr(date, 1, i-1) "/"
 463                                         date = substr(date, i+1)
 464                                 }
 465                                 date = newdate date
 466                         }
 467                         # Ignore any time zone; ChangeLog has no room for it.
 468                         time = substr($3, 1, 8)
 469                         author = substr($5, 1, length($5)-1)
 470                         printf "%s %s %s %s %c", filename, date, time, author, 13
 471                         next
 472                 }
 473                 if ($0 ~ /^(-----------*|===========*)$/) { print ""; next }
 474                 printf "%s%c", $0, 13
 475         }
 476 ' |
 477
 478 # Now each line is of the form
 479 # FILENAME YYYY/MM/DD HH:MM:SS AUTHOR \rLOG
 480 #       where \r stands for a carriage return,
 481 #       and each line of the log is terminated by \r instead of \n.
 482 # Sort the log entries, first by date+time (in reverse order),
 483 # then by author, then by log entry, and finally by file name (just in case).
 484 sort +1 -3r +3 +0 |
 485
 486 # Finally, reformat the sorted log entries.
 487 $AWK '
 488         BEGIN {
 489                 # Some awk variants do not understand "\r" or "\013", so we have to
 490                 # put a carriage return directly in the file.
 491                 CR="
" # <-- There is a single CR between the " chars here.
 492
 493                 # Initialize the fullname and mailaddr associative arrays.
 494                 '"$initialize_fullname"'
 495                 '"$initialize_mailaddr"'
 496
 497                 # Initialize indent string.
 498                 indent_string = ""
 499                 i = '"$indent"'
 500                 if (0 < '"$tabwidth"')
 501                         for (;  '"$tabwidth"' <= i;  i -= '"$tabwidth"')
 502                                 indent_string = indent_string "\t"
 503                 while (1 <= i--)
 504                         indent_string = indent_string " "
 505
 506                 # Set up date conversion tables.
 507                 # RCS uses a nice, clean, sortable format,
 508                 # but ChangeLog wants the traditional, ugly ctime format.
 509
 510                 # January 1, 0 AD (Gregorian) was Saturday = 6
 511                 EPOCH_WEEKDAY = 6
 512                 # Of course, there was no 0 AD, but the algorithm works anyway.
 513
 514                 w[0]="Sun"; w[1]="Mon"; w[2]="Tue"; w[3]="Wed"
 515                 w[4]="Thu"; w[5]="Fri"; w[6]="Sat"
 516
 517                 '"$month_data"'
 518         }
 519
 520         {
 521                 newlog = substr($0, 1 + index($0, CR))
 522
 523                 # Ignore log entries prefixed by "#".
 524                 if (newlog ~ /^#/) { next }
 525
 526                 if (Log != newlog || date != $2 || author != $4) {
 527
 528                         # The previous log and this log differ.
 529
 530                         # Print the old log.
 531                         if (date != "") '"$printlogline"'
 532
 533                         # Logs that begin with "{clumpname} " should be grouped together,
 534                         # and the clumpname should be removed.
 535                         # Extract the new clumpname from the log header,
 536                         # and use it to decide whether to output a blank line.
 537                         newclumpname = ""
 538                         sep = "\n"
 539                         if (date == "") sep = ""
 540                         if (newlog ~ /^\{[^'"$tab"' }]*}['"$tab"' ]/) {
 541                                 i = index(newlog, "}")
 542                                 newclumpname = substr(newlog, 1, i)
 543                                 while (substr(newlog, i+1) ~ /^['"$tab"' ]/) i++
 544                                 newlog = substr(newlog, i+1)
 545                                 if (clumpname == newclumpname) sep = ""
 546                         }
 547                         printf sep
 548                         clumpname = newclumpname
 549
 550                         # Get ready for the next log.
 551                         Log = newlog
 552                         if (files != "")
 553                                 for (i in filesknown)
 554                                         filesknown[i] = 0
 555                         files = ""
 556                 }
 557                 if (date != $2  ||  author != $4) {
 558                         # The previous date+author and this date+author differ.
 559                         # Print the new one.
 560                         date = $2
 561                         author = $4
 562
 563                         # Convert nice RCS date like "1992/01/03 00:03:44"
 564                         # into ugly ctime date like "Fri Jan  3 00:03:44 1992".
 565                         # Calculate day of week from Gregorian calendar.
 566                         i = index($2, "/")
 567                         year = substr($2, 1, i-1) + 0
 568                         monthday = substr($2, i+1)
 569                         i = index(monthday, "/")
 570                         month = substr(monthday, 1, i-1) + 0
 571                         day = substr(monthday, i+1) + 0
 572                         leap = 0
 573                         if (2 < month && year%4 == 0 && (year%100 != 0 || year%400 == 0)) leap = 1
 574                         days_since_Sunday_before_epoch = EPOCH_WEEKDAY + year * 365 + int((year + 3) / 4) - int((year + 99) / 100) + int((year + 399) / 400) + mo[month-1] + leap + day - 1
 575
 576                         # Print "date  fullname  (email address)".
 577                         # Get fullname and email address from associative arrays;
 578                         # default to author and author@hostname if not in arrays.
 579                         if (fullname[author])
 580                                 auth = fullname[author]
 581                         else
 582                                 auth = author
 583                         printf "%s %s %2d %s %d  %s  ", w[days_since_Sunday_before_epoch%7], m[month-1], day, $3, year, auth
 584                         if (mailaddr[author])
 585                                 printf "<%s>\n\n", mailaddr[author]
 586                         else
 587                                 printf "<%s@%s>\n\n", author, "'"$hostname"'"
 588                 }
 589                 if (! filesknown[$1]) {
 590                         filesknown[$1] = 1
 591                         if (files == "") files = " " $1
 592                         else files = files ", " $1
 593                 }
 594         }
 595         END {
 596                 # Print the last log.
 597                 if (date != "") {
 598                         '"$printlogline"'
 599                         printf "\n"
 600                 }
 601         }
 602 ' &&
 603
 604
 605 # Exit successfully.
 606
 607 exec rm -f $llogout $rlogout