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 1.22 1995/04/30 15:34:52 eggert Exp $
  16
  17 # Copyright 1992, 1993, 1994, 1995 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($0,18,2)
 138                                 ss++
 139                                 if (ss == 60) {
 140                                         ss = 0
 141                                         mm++
 142                                         if (mm == 60) {
 143                                                 mm = 0
 144                                                 hh++
 145                                                 if (hh == 24) {
 146                                                         hh = 0
 147                                                         dd++
 148                                                         monthdays = mo[i+1] - mo[i]
 149                                                         if (i == 1 && year%4 == 0 && (year%100 != 0 || year%400 == 0)) monthdays++
 150                                                         if (dd == monthdays + 1) {
 151                                                                 dd = 1
 152                                                                 i++
 153                                                                 if (i == 12) {
 154                                                                         i = 0
 155                                                                         year++
 156                                                                 }
 157                                                         }
 158                                                 }
 159                                         }
 160                                 }
 161                                 printf "%d/%02d/%02d %02d:%02d:%02d\n", year,i+1,dd,hh,mm,ss
 162                                 exit
 163                         }
 164                 '
 165                 d=`$AWK "$e" <ChangeLog` || exit
 166                 case $d in
 167                 ?*) date=$d
 168                 esac
 169         fi
 170         datearg="-d>$date"
 171 esac
 172
 173 # If CVS is in use, examine its repository, not the normal RCS files.
 174 if test ! -f CVS/Repository
 175 then
 176         rlog=rlog
 177         repository=
 178 else
 179         rlog='cvs log'
 180         repository=`sed 1q <CVS/Repository` || exit
 181         case $repository in
 182         /*) ;;
 183         *) repository=${CVSROOT?}/$repository
 184         esac
 185         if test ! -d "$repository"
 186         then
 187                 echo >&2 "$0: $repository: bad repository (see CVS/Repository)"
 188                 exit 1
 189         fi
 190 fi
 191
 192 # With no arguments, examine all files under the RCS directory.
 193 case $# in
 194 0)
 195         case $repository in
 196         '')
 197                 oldIFS=$IFS
 198                 IFS=$nl
 199                 case $recursive in
 200                 t)
 201                         RCSdirs=`find . -name RCS -type d -print`
 202                         filesFromRCSfiles='s|,v$||; s|/RCS/|/|; s|^\./||'
 203                         files=`
 204                                 {
 205                                         case $RCSdirs in
 206                                         ?*) find $RCSdirs -type f -print
 207                                         esac
 208                                         find . -name '*,v' -print
 209                                 } |
 210                                 sort -u |
 211                                 sed "$filesFromRCSfiles"
 212                         `;;
 213                 *)
 214                         files=
 215                         for file in RCS/.* RCS/* .*,v *,v
 216                         do
 217                                 case $file in
 218                                 RCS/. | RCS/..) continue;;
 219                                 RCS/.\* | RCS/\* | .\*,v | \*,v) test -f "$file" || continue
 220                                 esac
 221                                 files=$files$nl$file
 222                         done
 223                         case $files in
 224                         '') exit 0
 225                         esac
 226                 esac
 227                 set x $files
 228                 shift
 229                 IFS=$oldIFS
 230         esac
 231 esac
 232
 233 llogout=$TMPDIR/rcs2log$$l
 234 rlogout=$TMPDIR/rcs2log$$r
 235 trap exit 1 2 13 15
 236 trap "rm -f $llogout $rlogout; exit 1" 0
 237
 238 case $rlog_options in
 239 ?*) $rlog $rlog_options ${1+"$@"} >$rlogout;;
 240 '') $rlog "$datearg" ${1+"$@"} >$rlogout
 241 esac || exit
 242
 243
 244 # Get the full name of each author the logs mention, and set initialize_fullname
 245 # to awk code that initializes the `fullname' awk associative array.
 246 # Warning: foreign authors (i.e. not known in the passwd file) are mishandled;
 247 # you have to fix the resulting output by hand.
 248
 249 initialize_fullname=
 250 initialize_mailaddr=
 251
 252 case $loginFullnameMailaddrs in
 253 ?*)
 254         case $loginFullnameMailaddrs in
 255         *\"* | *\\*)
 256                 sed 's/["\\]/\\&/g' >$llogout <<EOF || exit
 257 $loginFullnameMailaddrs
 258 EOF
 259                 loginFullnameMailaddrs=`cat $llogout`
 260         esac
 261
 262         oldIFS=$IFS
 263         IFS=$nl
 264         for loginFullnameMailaddr in $loginFullnameMailaddrs
 265         do
 266                 case $loginFullnameMailaddr in
 267                 *"$tab"*) IFS=$tab;;
 268                 *) IFS=:
 269                 esac
 270                 set x $loginFullnameMailaddr
 271                 login=$2
 272                 fullname=$3
 273                 mailaddr=$4
 274                 initialize_fullname="$initialize_fullname
 275                         fullname[\"$login\"] = \"$fullname\""
 276                 initialize_mailaddr="$initialize_mailaddr
 277                         mailaddr[\"$login\"] = \"$mailaddr\""
 278         done
 279         IFS=$oldIFS
 280 esac
 281
 282 case $llogout in
 283 ?*) sort -u -o $llogout <<EOF || exit
 284 $logins
 285 EOF
 286 esac
 287 output_authors='/^date: / {
 288         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:" && $5 ~ /^[^;]*;$/) {
 289                 print substr($5, 1, length($5)-1)
 290         }
 291 }'
 292 authors=`
 293         $AWK "$output_authors" <$rlogout |
 294         case $llogout in
 295         '') sort -u;;
 296         ?*) sort -u | comm -23 - $llogout
 297         esac
 298 `
 299 case $authors in
 300 ?*)
 301         cat >$llogout <<EOF || exit
 302 $authors
 303 EOF
 304         initialize_author_script='s/["\\]/\\&/g; s/.*/author[\"&\"] = 1/'
 305         initialize_author=`sed -e "$initialize_author_script" <$llogout`
 306         awkscript='
 307                 BEGIN {
 308                         alphabet = "abcdefghijklmnopqrstuvwxyz"
 309                         ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 310                         '"$initialize_author"'
 311                 }
 312                 {
 313                         if (author[$1]) {
 314                                 fullname = $5
 315                                 if (fullname ~ /[0-9]+-[^(]*\([0-9]+\)$/) {
 316                                         # Remove the junk from fullnames like "0000-Admin(0000)".
 317                                         fullname = substr(fullname, index(fullname, "-") + 1)
 318                                         fullname = substr(fullname, 1, index(fullname, "(") - 1)
 319                                 }
 320                                 if (fullname ~ /,[^ ]/) {
 321                                         # Some sites put comma-separated junk after the fullname.
 322                                         # Remove it, but leave "Bill Gates, Jr" alone.
 323                                         fullname = substr(fullname, 1, index(fullname, ",") - 1)
 324                                 }
 325                                 abbr = index(fullname, "&")
 326                                 if (abbr) {
 327                                         a = substr($1, 1, 1)
 328                                         A = a
 329                                         i = index(alphabet, a)
 330                                         if (i) A = substr(ALPHABET, i, 1)
 331                                         fullname = substr(fullname, 1, abbr-1) A substr($1, 2) substr(fullname, abbr+1)
 332                                 }
 333
 334                                 # Quote quotes and backslashes properly in full names.
 335                                 # Do not use gsub; traditional awk lacks it.
 336                                 quoted = ""
 337                                 rest = fullname
 338                                 for (;;) {
 339                                         p = index(rest, "\\")
 340                                         q = index(rest, "\"")
 341                                         if (p) {
 342                                                 if (q && q<p) p = q
 343                                         } else {
 344                                                 if (!q) break
 345                                                 p = q
 346                                         }
 347                                         quoted = quoted substr(rest, 1, p-1) "\\" substr(rest, p, 1)
 348                                         rest = substr(rest, p+1)
 349                                 }
 350
 351                                 printf "fullname[\"%s\"] = \"%s%s\"\n", $1, quoted, rest
 352                                 author[$1] = 0
 353                         }
 354                 }
 355         '
 356
 357         initialize_fullname=`
 358                 (cat /etc/passwd; ypmatch $authors passwd) 2>/dev/null |
 359                 $AWK -F: "$awkscript"
 360         `$initialize_fullname
 361 esac
 362
 363
 364 # Function to print a single log line.
 365 # We don't use awk functions, to stay compatible with old awk versions.
 366 # `Log' is the log message (with \n replaced by \r).
 367 # `files' contains the affected files.
 368 printlogline='{
 369
 370         # Following the GNU coding standards, rewrite
 371         #       * file: (function): comment
 372         # to
 373         #       * file (function): comment
 374         if (Log ~ /^\([^)]*\): /) {
 375                 i = index(Log, ")")
 376                 files = files " " substr(Log, 1, i)
 377                 Log = substr(Log, i+3)
 378         }
 379
 380         # If "label: comment" is too long, break the line after the ":".
 381         sep = " "
 382         if ('"$length"' <= '"$indent"' + 1 + length(files) + index(Log, CR)) sep = "\n" indent_string
 383
 384         # Print the label.
 385         printf "%s*%s:", indent_string, files
 386
 387         # Print each line of the log, transliterating \r to \n.
 388         while ((i = index(Log, CR)) != 0) {
 389                 logline = substr(Log, 1, i-1)
 390                 if (logline ~ /[^'"$tab"' ]/) {
 391                         printf "%s%s\n", sep, logline
 392                 } else {
 393                         print ""
 394                 }
 395                 sep = indent_string
 396                 Log = substr(Log, i+1)
 397         }
 398 }'
 399
 400 case $hostname in
 401 '')
 402         hostname=`(
 403                 hostname || uname -n || uuname -l || cat /etc/whoami
 404         ) 2>/dev/null` || {
 405                 echo >&2 "$0: cannot deduce hostname"
 406                 exit 1
 407         }
 408 esac
 409
 410
 411 # Process the rlog output, generating ChangeLog style entries.
 412
 413 # First, reformat the rlog output so that each line contains one log entry.
 414 # Transliterate \n to \r so that multiline entries fit on a single line.
 415 # Discard irrelevant rlog output.
 416 $AWK <$rlogout '
 417         BEGIN { repository = "'"$repository"'" }
 418         /^RCS file:/ {
 419                 if (repository != "") {
 420                         filename = $3
 421                         if (substr(filename, 1, length(repository) + 1) == repository "/") {
 422                                 filename = substr(filename, length(repository) + 2)
 423                         }
 424                         if (filename ~ /,v$/) {
 425                                 filename = substr(filename, 1, length(filename) - 2)
 426                         }
 427                 }
 428         }
 429         /^Working file:/ { if (repository == "") filename = $3 }
 430         /^date: /, /^(-----------*|===========*)$/ {
 431                 if ($0 ~ /^branches: /) { next }
 432                 if ($0 ~ /^date: [0-9][- +\/0-9:]*;/) {
 433                         date = $2
 434                         if (date ~ /-/) {
 435                                 # An ISO format date.  Replace all "-"s with "/"s.
 436                                 newdate = ""
 437                                 while ((i = index(date, "-")) != 0) {
 438                                         newdate = newdate substr(date, 1, i-1) "/"
 439                                         date = substr(date, i+1)
 440                                 }
 441                                 date = newdate date
 442                         }
 443                         # Ignore any time zone; ChangeLog has no room for it.
 444                         time = substr($3, 1, 8)
 445                         author = substr($5, 1, length($5)-1)
 446                         printf "%s %s %s %s %c", filename, date, time, author, 13
 447                         next
 448                 }
 449                 if ($0 ~ /^(-----------*|===========*)$/) { print ""; next }
 450                 printf "%s%c", $0, 13
 451         }
 452 ' |
 453
 454 # Now each line is of the form
 455 # FILENAME YYYY/MM/DD HH:MM:SS AUTHOR \rLOG
 456 #       where \r stands for a carriage return,
 457 #       and each line of the log is terminated by \r instead of \n.
 458 # Sort the log entries, first by date+time (in reverse order),
 459 # then by author, then by log entry, and finally by file name (just in case).
 460 sort +1 -3r +3 +0 |
 461
 462 # Finally, reformat the sorted log entries.
 463 $AWK '
 464         BEGIN {
 465                 # Some awk variants do not understand "\r" or "\013", so we have to
 466                 # put a carriage return directly in the file.
 467                 CR="
" # <-- There is a single CR between the " chars here.
 468
 469                 # Initialize the fullname and mailaddr associative arrays.
 470                 '"$initialize_fullname"'
 471                 '"$initialize_mailaddr"'
 472
 473                 # Initialize indent string.
 474                 indent_string = ""
 475                 i = '"$indent"'
 476                 if (0 < '"$tabwidth"')
 477                         for (;  '"$tabwidth"' <= i;  i -= '"$tabwidth"')
 478                                 indent_string = indent_string "\t"
 479                 while (1 <= i--)
 480                         indent_string = indent_string " "
 481
 482                 # Set up date conversion tables.
 483                 # RCS uses a nice, clean, sortable format,
 484                 # but ChangeLog wants the traditional, ugly ctime format.
 485
 486                 # January 1, 0 AD (Gregorian) was Saturday = 6
 487                 EPOCH_WEEKDAY = 6
 488                 # Of course, there was no 0 AD, but the algorithm works anyway.
 489
 490                 w[0]="Sun"; w[1]="Mon"; w[2]="Tue"; w[3]="Wed"
 491                 w[4]="Thu"; w[5]="Fri"; w[6]="Sat"
 492
 493                 '"$month_data"'
 494         }
 495
 496         {
 497                 newlog = substr($0, 1 + index($0, CR))
 498
 499                 # Ignore log entries prefixed by "#".
 500                 if (newlog ~ /^#/) { next }
 501
 502                 if (Log != newlog || date != $2 || author != $4) {
 503
 504                         # The previous log and this log differ.
 505
 506                         # Print the old log.
 507                         if (date != "") '"$printlogline"'
 508
 509                         # Logs that begin with "{clumpname} " should be grouped together,
 510                         # and the clumpname should be removed.
 511                         # Extract the new clumpname from the log header,
 512                         # and use it to decide whether to output a blank line.
 513                         newclumpname = ""
 514                         sep = "\n"
 515                         if (date == "") sep = ""
 516                         if (newlog ~ /^\{[^'"$tab"' }]*}['"$tab"' ]/) {
 517                                 i = index(newlog, "}")
 518                                 newclumpname = substr(newlog, 1, i)
 519                                 while (substr(newlog, i+1) ~ /^['"$tab"' ]/) i++
 520                                 newlog = substr(newlog, i+1)
 521                                 if (clumpname == newclumpname) sep = ""
 522                         }
 523                         printf sep
 524                         clumpname = newclumpname
 525
 526                         # Get ready for the next log.
 527                         Log = newlog
 528                         if (files != "")
 529                                 for (i in filesknown)
 530                                         filesknown[i] = 0
 531                         files = ""
 532                 }
 533                 if (date != $2  ||  author != $4) {
 534                         # The previous date+author and this date+author differ.
 535                         # Print the new one.
 536                         date = $2
 537                         author = $4
 538
 539                         # Convert nice RCS date like "1992/01/03 00:03:44"
 540                         # into ugly ctime date like "Fri Jan  3 00:03:44 1992".
 541                         # Calculate day of week from Gregorian calendar.
 542                         i = index($2, "/")
 543                         year = substr($2, 1, i-1) + 0
 544                         monthday = substr($2, i+1)
 545                         i = index(monthday, "/")
 546                         month = substr(monthday, 1, i-1) + 0
 547                         day = substr(monthday, i+1) + 0
 548                         leap = 0
 549                         if (2 < month && year%4 == 0 && (year%100 != 0 || year%400 == 0)) leap = 1
 550                         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
 551
 552                         # Print "date  fullname  (email address)".
 553                         # Get fullname and email address from associative arrays;
 554                         # default to author and author@hostname if not in arrays.
 555                         if (fullname[author])
 556                                 auth = fullname[author]
 557                         else
 558                                 auth = author
 559                         printf "%s %s %2d %s %d  %s  ", w[days_since_Sunday_before_epoch%7], m[month-1], day, $3, year, auth
 560                         if (mailaddr[author])
 561                                 printf "<%s>\n\n", mailaddr[author]
 562                         else
 563                                 printf "<%s@%s>\n\n", author, "'"$hostname"'"
 564                 }
 565                 if (! filesknown[$1]) {
 566                         filesknown[$1] = 1
 567                         if (files == "") files = " " $1
 568                         else files = files ", " $1
 569                 }
 570         }
 571         END {
 572                 # Print the last log.
 573                 if (date != "") {
 574                         '"$printlogline"'
 575                         printf "\n"
 576                 }
 577         }
 578 ' &&
 579
 580
 581 # Exit successfully.
 582
 583 exec rm -f $llogout $rlogout