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 1.20 1995/03/21 23:14:41 eggert Exp eggert $
  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 mailaddresses 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 # Log into $rlogout the revisions checked in since the first ChangeLog entry.
 120
 121 date=1970
 122 if test -s ChangeLog
 123 then
 124         # Add 1 to seconds to avoid duplicating most recent log.
 125         e='
 126                 /^... ... [ 0-9][0-9] [ 0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9]+ /{
 127                         '"$month_data"'
 128                         year = $5
 129                         for (i=0; i<=11; i++) if (m[i] == $2) break
 130                         dd = $3
 131                         hh = substr($0,12,2)
 132                         mm = substr($0,15,2)
 133                         ss = substr($0,18,2)
 134                         ss++
 135                         if (ss == 60) {
 136                                 ss = 0
 137                                 mm++
 138                                 if (mm == 60) {
 139                                         mm = 0
 140                                         hh++
 141                                         if (hh == 24) {
 142                                                 hh = 0
 143                                                 dd++
 144                                                 monthdays = mo[i+1] - mo[i]
 145                                                 if (i == 1 && year%4 == 0 && (year%100 != 0 || year%400 == 0)) monthdays++
 146                                                 if (dd == monthdays + 1) {
 147                                                         dd = 1
 148                                                         i++
 149                                                         if (i == 12) {
 150                                                                 i = 0
 151                                                                 year++
 152                                                         }
 153                                                 }
 154                                         }
 155                                 }
 156                         }
 157                         printf "%d/%02d/%02d %02d:%02d:%02d\n", year, i+1, dd, hh, mm, ss
 158                         exit
 159                 }
 160         '
 161         d=`$AWK "$e" <ChangeLog` || exit
 162         case $d in
 163         ?*) date=$d
 164         esac
 165 fi
 166 datearg="-d>$date"
 167
 168 repository=
 169 rlog=rlog
 170 case $CVSROOT in
 171 ?*)
 172         if test -d "$CVSROOT" && test -f CVS/Repository
 173         then
 174                 r=`cat <CVS/Repository` || exit
 175                 if test -d "$CVSROOT/$r"
 176                 then
 177                         repository=$CVSROOT/$r
 178                         rlog='cvs log'
 179                 fi
 180         fi
 181 esac
 182
 183 # With no arguments, examine all files under the RCS directory.
 184 case $# in
 185 0)
 186         case $repository in
 187         '')
 188                 oldIFS=$IFS
 189                 IFS=$nl
 190                 case $recursive in
 191                 t)
 192                         RCSdirs=`find . -name RCS -type d -print`
 193                         filesFromRCSfiles='s|,v$||; s|/RCS/|/|; s|^\./||'
 194                         files=`
 195                                 {
 196                                         case $RCSdirs in
 197                                         ?*) find $RCSdirs -type f -print
 198                                         esac
 199                                         find . -name '*,v' -print
 200                                 } |
 201                                 sort -u |
 202                                 sed "$filesFromRCSfiles"
 203                         `;;
 204                 *)
 205                         files=
 206                         for file in RCS/.* RCS/* .*,v *,v
 207                         do
 208                                 case $file in
 209                                 RCS/. | RCS/..) continue;;
 210                                 RCS/.\* | RCS/\* | .\*,v | \*,v) test -f "$file" || continue
 211                                 esac
 212                                 files=$files$nl$file
 213                         done
 214                         case $files in
 215                         '') exit 0
 216                         esac
 217                 esac
 218                 set x $files
 219                 shift
 220                 IFS=$oldIFS
 221         esac
 222 esac
 223
 224 llogout=$TMPDIR/rcs2log$$l
 225 rlogout=$TMPDIR/rcs2log$$r
 226 trap exit 1 2 13 15
 227 trap "rm -f $llogout $rlogout; exit 1" 0
 228
 229 $rlog "$datearg" $rlog_options ${1+"$@"} >$rlogout || exit
 230
 231
 232 # Get the full name of each author the logs mention, and set initialize_fullname
 233 # to awk code that initializes the `fullname' awk associative array.
 234 # Warning: foreign authors (i.e. not known in the passwd file) are mishandled;
 235 # you have to fix the resulting output by hand.
 236
 237 initialize_fullname=
 238 initialize_mailaddr=
 239
 240 case $loginFullnameMailaddrs in
 241 ?*)
 242         case $loginFullnameMailaddrs in
 243         *\"* | *\\*)
 244                 sed 's/["\\]/\\&/g' >$llogout <<EOF || exit
 245 $loginFullnameMailaddrs
 246 EOF
 247                 loginFullnameMailaddrs=`cat $llogout`
 248         esac
 249
 250         oldIFS=$IFS
 251         IFS=$nl
 252         for loginFullnameMailaddr in $loginFullnameMailaddrs
 253         do
 254                 case $loginFullnameMailaddr in
 255                 *"$tab"*) IFS=$tab;;
 256                 *) IFS=:
 257                 esac
 258                 set x $loginFullnameMailaddr
 259                 login=$2
 260                 fullname=$3
 261                 mailaddr=$4
 262                 initialize_fullname="$initialize_fullname
 263                         fullname[\"$login\"] = \"$fullname\""
 264                 initialize_mailaddr="$initialize_mailaddr
 265                         mailaddr[\"$login\"] = \"$mailaddr\""
 266         done
 267         IFS=$oldIFS
 268 esac
 269
 270 case $llogout in
 271 ?*) sort -u -o $llogout <<EOF || exit
 272 $logins
 273 EOF
 274 esac
 275 output_authors='/^date: / {
 276         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:" && $5 ~ /^[^;]*;$/) {
 277                 print substr($5, 1, length($5)-1)
 278         }
 279 }'
 280 authors=`
 281         $AWK "$output_authors" <$rlogout |
 282         case $llogout in
 283         '') sort -u;;
 284         ?*) sort -u | comm -23 - $llogout
 285         esac
 286 `
 287 case $authors in
 288 ?*)
 289         cat >$llogout <<EOF || exit
 290 $authors
 291 EOF
 292         initialize_author_script='s/["\\]/\\&/g; s/.*/author[\"&\"] = 1/'
 293         initialize_author=`sed -e "$initialize_author_script" <$llogout`
 294         awkscript='
 295                 BEGIN {
 296                         alphabet = "abcdefghijklmnopqrstuvwxyz"
 297                         ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 298                         '"$initialize_author"'
 299                 }
 300                 {
 301                         if (author[$1]) {
 302                                 fullname = $5
 303                                 if (fullname ~ /[0-9]+-[^(]*\([0-9]+\)$/) {
 304                                         # Remove the junk from fullnames like "0000-Admin(0000)".
 305                                         fullname = substr(fullname, index(fullname, "-") + 1)
 306                                         fullname = substr(fullname, 1, index(fullname, "(") - 1)
 307                                 }
 308                                 if (fullname ~ /,[^ ]/) {
 309                                         # Some sites put comma-separated junk after the fullname.
 310                                         # Remove it, but leave "Bill Gates, Jr" alone.
 311                                         fullname = substr(fullname, 1, index(fullname, ",") - 1)
 312                                 }
 313                                 abbr = index(fullname, "&")
 314                                 if (abbr) {
 315                                         a = substr($1, 1, 1)
 316                                         A = a
 317                                         i = index(alphabet, a)
 318                                         if (i) A = substr(ALPHABET, i, 1)
 319                                         fullname = substr(fullname, 1, abbr-1) A substr($1, 2) substr(fullname, abbr+1)
 320                                 }
 321
 322                                 # Quote quotes and backslashes properly in full names.
 323                                 # Do not use gsub; traditional awk lacks it.
 324                                 quoted = ""
 325                                 rest = fullname
 326                                 for (;;) {
 327                                         p = index(rest, "\\")
 328                                         q = index(rest, "\"")
 329                                         if (p) {
 330                                                 if (q && q<p) p = q
 331                                         } else {
 332                                                 if (!q) break
 333                                                 p = q
 334                                         }
 335                                         quoted = quoted substr(rest, 1, p-1) "\\" substr(rest, p, 1)
 336                                         rest = substr(rest, p+1)
 337                                 }
 338
 339                                 printf "fullname[\"%s\"] = \"%s%s\"\n", $1, quoted, rest
 340                                 author[$1] = 0
 341                         }
 342                 }
 343         '
 344
 345         initialize_fullname=`
 346                 (cat /etc/passwd; ypmatch $authors passwd) 2>/dev/null |
 347                 $AWK -F: "$awkscript"
 348         `$initialize_fullname
 349 esac
 350
 351
 352 # Function to print a single log line.
 353 # We don't use awk functions, to stay compatible with old awk versions.
 354 # `Log' is the log message (with \n replaced by \r).
 355 # `files' contains the affected files.
 356 printlogline='{
 357
 358         # Following the GNU coding standards, rewrite
 359         #       * file: (function): comment
 360         # to
 361         #       * file (function): comment
 362         if (Log ~ /^\([^)]*\): /) {
 363                 i = index(Log, ")")
 364                 files = files " " substr(Log, 1, i)
 365                 Log = substr(Log, i+3)
 366         }
 367
 368         # If "label: comment" is too long, break the line after the ":".
 369         sep = " "
 370         if ('"$length"' <= '"$indent"' + 1 + length(files) + index(Log, CR)) sep = "\n" indent_string
 371
 372         # Print the label.
 373         printf "%s*%s:", indent_string, files
 374
 375         # Print each line of the log, transliterating \r to \n.
 376         while ((i = index(Log, CR)) != 0) {
 377                 logline = substr(Log, 1, i-1)
 378                 if (logline ~ /[^'"$tab"' ]/) {
 379                         printf "%s%s\n", sep, logline
 380                 } else {
 381                         print ""
 382                 }
 383                 sep = indent_string
 384                 Log = substr(Log, i+1)
 385         }
 386 }'
 387
 388 case $hostname in
 389 '')
 390         hostname=`(
 391                 hostname || uname -n || uuname -l || cat /etc/whoami
 392         ) 2>/dev/null` || {
 393                 echo >&2 "$0: cannot deduce hostname"
 394                 exit 1
 395         }
 396 esac
 397
 398
 399 # Process the rlog output, generating ChangeLog style entries.
 400
 401 # First, reformat the rlog output so that each line contains one log entry.
 402 # Transliterate \n to \r so that multiline entries fit on a single line.
 403 # Discard irrelevant rlog output.
 404 $AWK <$rlogout '
 405         BEGIN { repository = "'"$repository"'" }
 406         /^RCS file:/ {
 407                 if (repository != "") {
 408                         filename = $3
 409                         if (substr(filename, 1, length(repository) + 1) == repository "/") {
 410                                 filename = substr(filename, length(repository) + 2)
 411                         }
 412                         if (filename ~ /,v$/) {
 413                                 filename = substr(filename, 1, length(filename) - 2)
 414                         }
 415                 }
 416         }
 417         /^Working file:/ { if (repository == "") filename = $3 }
 418         /^date: /, /^(-----------*|===========*)$/ {
 419                 if ($0 ~ /^branches: /) { next }
 420                 if ($0 ~ /^date: [0-9][- +\/0-9:]*;/) {
 421                         date = $2
 422                         if (date ~ /-/) {
 423                                 # An ISO format date.  Replace all "-"s with "/"s.
 424                                 newdate = ""
 425                                 while ((i = index(date, "-")) != 0) {
 426                                         newdate = newdate substr(date, 1, i-1) "/"
 427                                         date = substr(date, i+1)
 428                                 }
 429                                 date = newdate date
 430                         }
 431                         # Ignore any time zone; ChangeLog has no room for it.
 432                         time = substr($3, 1, 8)
 433                         author = substr($5, 1, length($5)-1)
 434                         printf "%s %s %s %s %c", filename, date, time, author, 13
 435                         next
 436                 }
 437                 if ($0 ~ /^(-----------*|===========*)$/) { print ""; next }
 438                 printf "%s%c", $0, 13
 439         }
 440 ' |
 441
 442 # Now each line is of the form
 443 # FILENAME YYYY/MM/DD HH:MM:SS AUTHOR \rLOG
 444 #       where \r stands for a carriage return,
 445 #       and each line of the log is terminated by \r instead of \n.
 446 # Sort the log entries, first by date+time (in reverse order),
 447 # then by author, then by log entry, and finally by file name (just in case).
 448 sort +1 -3r +3 +0 |
 449
 450 # Finally, reformat the sorted log entries.
 451 $AWK '
 452         BEGIN {
 453                 # Some awks do not understand "\r" or "\013", so we have to
 454                 # put a carriage return directly in the file.
 455                 CR="
" # <-- There is a single CR between the " chars here.
 456
 457                 # Initialize the fullname and mailaddr associative arrays.
 458                 '"$initialize_fullname"'
 459                 '"$initialize_mailaddr"'
 460
 461                 # Initialize indent string.
 462                 indent_string = ""
 463                 i = '"$indent"'
 464                 if (0 < '"$tabwidth"')
 465                         for (;  '"$tabwidth"' <= i;  i -= '"$tabwidth"')
 466                                 indent_string = indent_string "\t"
 467                 while (1 <= i--)
 468                         indent_string = indent_string " "
 469
 470                 # Set up date conversion tables.
 471                 # RCS uses a nice, clean, sortable format,
 472                 # but ChangeLog wants the traditional, ugly ctime format.
 473
 474                 # January 1, 0 AD (Gregorian) was Saturday = 6
 475                 EPOCH_WEEKDAY = 6
 476                 # Of course, there was no 0 AD, but the algorithm works anyway.
 477
 478                 w[0]="Sun"; w[1]="Mon"; w[2]="Tue"; w[3]="Wed"
 479                 w[4]="Thu"; w[5]="Fri"; w[6]="Sat"
 480
 481                 '"$month_data"'
 482         }
 483
 484         {
 485                 newlog = substr($0, 1 + index($0, CR))
 486
 487                 # Ignore log entries prefixed by "#".
 488                 if (newlog ~ /^#/) { next }
 489
 490                 if (Log != newlog || date != $2 || author != $4) {
 491
 492                         # The previous log and this log differ.
 493
 494                         # Print the old log.
 495                         if (date != "") '"$printlogline"'
 496
 497                         # Logs that begin with "{clumpname} " should be grouped together,
 498                         # and the clumpname should be removed.
 499                         # Extract the new clumpname from the log header,
 500                         # and use it to decide whether to output a blank line.
 501                         newclumpname = ""
 502                         sep = "\n"
 503                         if (date == "") sep = ""
 504                         if (newlog ~ /^\{[^'"$tab"' }]*}['"$tab"' ]/) {
 505                                 i = index(newlog, "}")
 506                                 newclumpname = substr(newlog, 1, i)
 507                                 while (substr(newlog, i+1) ~ /^['"$tab"' ]/) i++
 508                                 newlog = substr(newlog, i+1)
 509                                 if (clumpname == newclumpname) sep = ""
 510                         }
 511                         printf sep
 512                         clumpname = newclumpname
 513
 514                         # Get ready for the next log.
 515                         Log = newlog
 516                         if (files != "")
 517                                 for (i in filesknown)
 518                                         filesknown[i] = 0
 519                         files = ""
 520                 }
 521                 if (date != $2  ||  author != $4) {
 522                         # The previous date+author and this date+author differ.
 523                         # Print the new one.
 524                         date = $2
 525                         author = $4
 526
 527                         # Convert nice RCS date like "1992/01/03 00:03:44"
 528                         # into ugly ctime date like "Fri Jan  3 00:03:44 1992".
 529                         # Calculate day of week from Gregorian calendar.
 530                         i = index($2, "/")
 531                         year = substr($2, 1, i-1) + 0
 532                         monthday = substr($2, i+1)
 533                         i = index(monthday, "/")
 534                         month = substr(monthday, 1, i-1) + 0
 535                         day = substr(monthday, i+1) + 0
 536                         leap = 0
 537                         if (2 < month && year%4 == 0 && (year%100 != 0 || year%400 == 0)) leap = 1
 538                         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
 539
 540                         # Print "date  fullname  (email address)".
 541                         # Get fullname and email address from associative arrays;
 542                         # default to author and author@hostname if not in arrays.
 543                         if (fullname[author])
 544                                 auth = fullname[author]
 545                         else
 546                                 auth = author
 547                         printf "%s %s %2d %s %d  %s  ", w[days_since_Sunday_before_epoch%7], m[month-1], day, $3, year, auth
 548                         if (mailaddr[author])
 549                                 printf "<%s>\n\n", mailaddr[author]
 550                         else
 551                                 printf "<%s@%s>\n\n", author, "'"$hostname"'"
 552                 }
 553                 if (! filesknown[$1]) {
 554                         filesknown[$1] = 1
 555                         if (files == "") files = " " $1
 556                         else files = files ", " $1
 557                 }
 558         }
 559         END {
 560                 # Print the last log.
 561                 if (date != "") {
 562                         '"$printlogline"'
 563                         printf "\n"
 564                 }
 565         }
 566 ' &&
 567
 568
 569 # Exit successfully.
 570
 571 exec rm -f $llogout $rlogout