1#!./perl -Ilib -w 2 3# This file should really be extracted from a .PL file 4 5use strict; 6use Config; # for config options in the makefile 7use File::Spec; 8use Getopt::Long; # for command-line parsing 9use Cwd; 10use Pod::Html 'anchorify'; 11 12=head1 NAME 13 14installhtml - converts a collection of POD pages to HTML format. 15 16=head1 SYNOPSIS 17 18 installhtml [--help] [--podpath=<name>:...:<name>] [--podroot=<name>] 19 [--htmldir=<name>] [--htmlroot=<name>] [--norecurse] [--recurse] 20 [--splithead=<name>,...,<name>] [--splititem=<name>,...,<name>] 21 [--libpods=<name>,...,<name>] [--ignore=<name>,...,<name>] 22 [--verbose] 23 24=head1 DESCRIPTION 25 26I<installhtml> converts a collection of POD pages to a corresponding 27collection of HTML pages. This is primarily used to convert the pod 28pages found in the perl distribution. 29 30=head1 OPTIONS 31 32=over 4 33 34=item B<--help> help 35 36Displays the usage. 37 38=item B<--podroot> POD search path base directory 39 40The base directory to search for all .pod and .pm files to be converted. 41Default is current directory. 42 43=item B<--podpath> POD search path 44 45The list of directories to search for .pod and .pm files to be converted. 46Default is `podroot/.'. 47 48=item B<--recurse> recurse on subdirectories 49 50Whether or not to convert all .pm and .pod files found in subdirectories 51too. Default is to not recurse. 52 53=item B<--htmldir> HTML destination directory 54 55The base directory which all HTML files will be written to. This should 56be a path relative to the filesystem, not the resulting URL. 57 58=item B<--htmlroot> URL base directory 59 60The base directory which all resulting HTML files will be visible at in 61a URL. The default is `/'. 62 63=item B<--splithead> POD files to split on =head directive 64 65Comma-separated list of pod files to split by the =head directive. The 66.pod suffix is optional. These files should have names specified 67relative to podroot. 68 69=item B<--splititem> POD files to split on =item directive 70 71Comma-separated list of all pod files to split by the =item directive. 72The .pod suffix is optional. I<installhtml> does not do the actual 73split, rather it invokes I<splitpod> to do the dirty work. As with 74--splithead, these files should have names specified relative to podroot. 75 76=item B<--splitpod> Directory containing the splitpod program 77 78The directory containing the splitpod program. The default is `podroot/pod'. 79 80=item B<--libpods> library PODs for LE<lt>E<gt> links 81 82Comma-separated list of "library" pod files. This is the same list that 83will be passed to pod2html when any pod is converted. 84 85=item B<--ignore> files to be ignored 86 87Comma-separated of files that shouldn't be installed, given relative 88to podroot. 89 90=item B<--verbose> verbose output 91 92Self-explanatory. 93 94=back 95 96=head1 EXAMPLE 97 98The following command-line is an example of the one we use to convert 99perl documentation: 100 101 ./installhtml --podpath=lib:ext:pod:vms \ 102 --podroot=/usr/src/perl \ 103 --htmldir=/perl/nmanual \ 104 --htmlroot=/perl/nmanual \ 105 --splithead=pod/perlipc \ 106 --splititem=pod/perlfunc \ 107 --libpods=perlfunc,perlguts,perlvar,perlrun,perlop \ 108 --recurse \ 109 --verbose 110 111=head1 AUTHOR 112 113Chris Hall E<lt>hallc@cs.colorado.eduE<gt> 114 115=head1 TODO 116 117=cut 118 119my $usage; 120 121$usage =<<END_OF_USAGE; 122Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name> 123 --htmldir=<name> --htmlroot=<name> --norecurse --recurse 124 --splithead=<name>,...,<name> --splititem=<name>,...,<name> 125 --libpods=<name>,...,<name> --ignore=<name>,...,<name> --verbose 126 127 --help - this message 128 --podpath - colon-separated list of directories containing .pod and 129 .pm files to be converted (. by default). 130 --podroot - filesystem base directory from which all relative paths in 131 podpath stem (default is .). 132 --htmldir - directory to store resulting html files in relative 133 to the filesystem (\$podroot/html by default). 134 --htmlroot - http-server base directory from which all relative paths 135 in podpath stem (default is /). 136 --libpods - comma-separated list of files to search for =item pod 137 directives in as targets of C<> and implicit links (empty 138 by default). 139 --norecurse - don't recurse on those subdirectories listed in podpath. 140 (default behavior). 141 --recurse - recurse on those subdirectories listed in podpath 142 --splithead - comma-separated list of .pod or .pm files to split. will 143 split each file into several smaller files at every occurrence 144 of a pod =head[1-6] directive. 145 --splititem - comma-separated list of .pod or .pm files to split using 146 splitpod. 147 --splitpod - directory where the program splitpod can be found 148 (\$podroot/pod by default). 149 --ignore - comma-separated list of files that shouldn't be installed. 150 --verbose - self-explanatory. 151 152END_OF_USAGE 153 154my (@libpods, @podpath, $podroot, $htmldir, $htmlroot, $recurse, @splithead, 155 @splititem, $splitpod, $verbose, $pod2html, @ignore); 156 157@libpods = (); 158@podpath = ( "." ); # colon-separated list of directories containing .pod 159 # and .pm files to be converted. 160$podroot = "."; # assume the pods we want are here 161$htmldir = ""; # nothing for now... 162$htmlroot = "/"; # default value 163$recurse = 0; # default behavior 164@splithead = (); # don't split any files by default 165@splititem = (); # don't split any files by default 166$splitpod = ""; # nothing for now. 167 168$verbose = 0; # whether or not to print debugging info 169 170$pod2html = "pod/pod2html"; 171 172usage("") unless @ARGV; 173 174# Overcome shell's p1,..,p8 limitation. 175# See vms/descrip_mms.template -> descrip.mms for invokation. 176if ( $^O eq 'VMS' ) { @ARGV = split(/\s+/,$ARGV[0]); } 177 178use vars qw( %Options ); 179 180# parse the command-line 181my $result = GetOptions( \%Options, qw( 182 help 183 podpath=s 184 podroot=s 185 htmldir=s 186 htmlroot=s 187 libpods=s 188 ignore=s 189 recurse! 190 splithead=s 191 splititem=s 192 splitpod=s 193 verbose 194)); 195usage("invalid parameters") unless $result; 196parse_command_line(); 197 198 199# set these variables to appropriate values if the user didn't specify 200# values for them. 201$htmldir = "$htmlroot/html" unless $htmldir; 202$splitpod = "$podroot/pod" unless $splitpod; 203 204 205# make sure that the destination directory exists 206(mkdir($htmldir, 0755) || 207 die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir; 208 209 210# the following array will eventually contain files that are to be 211# ignored in the conversion process. these are files that have been 212# process by splititem or splithead and should not be converted as a 213# result. 214my @splitdirs; 215 216# split pods. It's important to do this before convert ANY pods because 217# it may affect some of the links 218@splitdirs = (); # files in these directories won't get an index 219split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead); 220split_on_item($podroot, \@splitdirs, \@ignore, @splititem); 221 222 223# convert the pod pages found in @poddirs 224#warn "converting files\n" if $verbose; 225#warn "\@ignore\t= @ignore\n" if $verbose; 226foreach my $dir (@podpath) { 227 installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore); 228} 229 230 231# now go through and create master indices for each pod we split 232foreach my $dir (@splititem) { 233 print "creating index $htmldir/$dir.html\n" if $verbose; 234 create_index("$htmldir/$dir.html", "$htmldir/$dir"); 235} 236 237foreach my $dir (@splithead) { 238 (my $pod = $dir) =~ s,^.*/,,; 239 $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/; 240 # let pod2html create the file 241 runpod2html($dir, 1); 242 243 # now go through and truncate after the index 244 $dir =~ /^(.*?)(\.pod|\.pm)?$/sm; 245 my $file = "$htmldir/$1"; 246 print "creating index $file.html\n" if $verbose; 247 248 # read in everything until what would have been the first =head 249 # directive, patching the index as we go. 250 open(H, "<$file.html") || 251 die "$0: error opening $file.html for input: $!\n"; 252 $/ = ""; 253 my @data = (); 254 while (<H>) { 255 last if /name="name"/i; 256 $_ =~ s{href="#(.*)">}{ 257 my $url = "$pod/$1.html" ; 258 $url = Pod::Html::relativize_url( $url, "$file.html" ) 259 if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ); 260 "href=\"$url\">" ; 261 }egi; 262 push @data, $_; 263 } 264 close(H); 265 266 # now rewrite the file 267 open(H, ">$file.html") || 268 die "$0: error opening $file.html for output: $!\n"; 269 print H "@data", "\n"; 270 close(H); 271} 272 273############################################################################## 274 275 276sub usage { 277 warn "$0: @_\n" if @_; 278 die $usage; 279} 280 281 282sub parse_command_line { 283 usage() if defined $Options{help}; 284 $Options{help} = ""; # make -w shut up 285 286 # list of directories 287 @podpath = split(":", $Options{podpath}) if defined $Options{podpath}; 288 289 # lists of files 290 @splithead = split(",", $Options{splithead}) if defined $Options{splithead}; 291 @splititem = split(",", $Options{splititem}) if defined $Options{splititem}; 292 @libpods = split(",", $Options{libpods}) if defined $Options{libpods}; 293 294 $htmldir = $Options{htmldir} if defined $Options{htmldir}; 295 $htmlroot = $Options{htmlroot} if defined $Options{htmlroot}; 296 $podroot = $Options{podroot} if defined $Options{podroot}; 297 $splitpod = $Options{splitpod} if defined $Options{splitpod}; 298 299 $recurse = $Options{recurse} if defined $Options{recurse}; 300 $verbose = $Options{verbose} if defined $Options{verbose}; 301 302 @ignore = map "$podroot/$_", split(",", $Options{ignore}) if defined $Options{ignore}; 303} 304 305 306sub create_index { 307 my($html, $dir) = @_; 308 (my $pod = $dir) =~ s,^.*/,,; 309 my(@files, @filedata, @index, $file); 310 my($lcp1,$lcp2); 311 312 313 # get the list of .html files in this directory 314 opendir(DIR, $dir) || 315 die "$0: error opening directory $dir for reading: $!\n"; 316 @files = sort(grep(/\.html?$/, readdir(DIR))); 317 closedir(DIR); 318 319 open(HTML, ">$html") || 320 die "$0: error opening $html for output: $!\n"; 321 322 # for each .html file in the directory, extract the index 323 # embedded in the file and throw it into the big index. 324 print HTML "<DL COMPACT>\n"; 325 foreach $file (@files) { 326 $/ = ""; 327 328 open(IN, "<$dir/$file") || 329 die "$0: error opening $dir/$file for input: $!\n"; 330 @filedata = <IN>; 331 close(IN); 332 333 # pull out the NAME section 334 my $name; 335 ($name) = grep(/name="name"/i, @filedata); 336 ($lcp1,$lcp2) = ($name =~ m,/H1>\s(\S+)\s[\s-]*(.*?)\s*$,smi); 337 if (defined $lcp1 and $lcp1 =~ m,^<P>$,i) { # Uninteresting. Try again. 338 ($lcp1,$lcp2) = ($name =~ m,/H1>\s<P>\s*(\S+)\s[\s-]*(.*?)\s*$,smi); 339 } 340 my $url= "$pod/$file" ; 341 if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ) { 342 $url = Pod::Html::relativize_url( "$pod/$file", $html ) ; 343 } 344 345 if (defined $lcp1) { 346 print HTML qq(<DT><A HREF="$url">); 347 print HTML "$lcp1</A></DT><DD>$lcp2</DD>\n"; 348 } 349 350 next; 351 352 @index = grep(/<!-- INDEX BEGIN -->.*<!-- INDEX END -->/s, 353 @filedata); 354 for (@index) { 355 s/<!-- INDEX BEGIN -->(\s*<!--)(.*)(-->\s*)<!-- INDEX END -->/$lcp2/s; 356 s,#,$dir/$file#,g; 357 print HTML "$_\n<P><HR><P>\n"; 358 } 359 } 360 print HTML "</DL>\n"; 361 362 close(HTML); 363} 364 365 366sub split_on_head { 367 my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_; 368 my($pod, $dirname, $filename); 369 370 # split the files specified in @splithead on =head[1-6] pod directives 371 print "splitting files by head.\n" if $verbose && $#splithead >= 0; 372 foreach $pod (@splithead) { 373 # figure out the directory name and filename 374 $pod =~ s,^([^/]*)$,/$1,; 375 $pod =~ m,(.*)/(.*?)(\.pod)?$,; 376 $dirname = $1; 377 $filename = "$2.pod"; 378 379 # since we are splitting this file it shouldn't be converted. 380 push(@$ignore, "$podroot/$dirname/$filename"); 381 382 # split the pod 383 splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir, 384 $splitdirs); 385 } 386} 387 388 389sub split_on_item { 390 my($podroot, $splitdirs, $ignore, @splititem) = @_; 391 my($pwd, $dirname, $filename); 392 393 print "splitting files by item.\n" if $verbose && $#splititem >= 0; 394 $pwd = getcwd(); 395 my $splitter = File::Spec->rel2abs("$splitpod/splitpod", $pwd); 396 my $perl = File::Spec->rel2abs($^X, $pwd); 397 foreach my $pod (@splititem) { 398 # figure out the directory to split into 399 $pod =~ s,^([^/]*)$,/$1,; 400 $pod =~ m,(.*)/(.*?)(\.pod)?$,; 401 $dirname = "$1/$2"; 402 $filename = "$2.pod"; 403 404 # since we are splitting this file it shouldn't be converted. 405 push(@$ignore, "$podroot/$dirname.pod"); 406 407 # split the pod 408 push(@$splitdirs, "$podroot/$dirname"); 409 if (! -d "$podroot/$dirname") { 410 mkdir("$podroot/$dirname", 0755) || 411 die "$0: error creating directory $podroot/$dirname: $!\n"; 412 } 413 chdir("$podroot/$dirname") || 414 die "$0: error changing to directory $podroot/$dirname: $!\n"; 415 die "$splitter not found. Use '-splitpod dir' option.\n" 416 unless -f $splitter; 417 system($perl, $splitter, "../$filename") && 418 warn "$0: error running '$splitter ../$filename'" 419 ." from $podroot/$dirname"; 420 } 421 chdir($pwd); 422} 423 424 425# 426# splitpod - splits a .pod file into several smaller .pod files 427# where a new file is started each time a =head[1-6] pod directive 428# is encountered in the input file. 429# 430sub splitpod { 431 my($pod, $poddir, $htmldir, $splitdirs) = @_; 432 my(@poddata, @filedata, @heads); 433 my($file, $i, $j, $prevsec, $section, $nextsec); 434 435 print "splitting $pod\n" if $verbose; 436 437 # read the file in paragraphs 438 $/ = ""; 439 open(SPLITIN, "<$pod") || 440 die "$0: error opening $pod for input: $!\n"; 441 @filedata = <SPLITIN>; 442 close(SPLITIN) || 443 die "$0: error closing $pod: $!\n"; 444 445 # restore the file internally by =head[1-6] sections 446 @poddata = (); 447 for ($i = 0, $j = -1; $i <= $#filedata; $i++) { 448 $j++ if ($filedata[$i] =~ /^\s*=head[1-6]/); 449 if ($j >= 0) { 450 $poddata[$j] = "" unless defined $poddata[$j]; 451 $poddata[$j] .= "\n$filedata[$i]" if $j >= 0; 452 } 453 } 454 455 # create list of =head[1-6] sections so that we can rewrite 456 # L<> links as necessary. 457 my %heads = (); 458 foreach $i (0..$#poddata) { 459 $heads{anchorify($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/; 460 } 461 462 # create a directory of a similar name and store all the 463 # files in there 464 $pod =~ s,.*/(.*),$1,; # get the last part of the name 465 my $dir = $pod; 466 $dir =~ s/\.pod//g; 467 push(@$splitdirs, "$poddir/$dir"); 468 mkdir("$poddir/$dir", 0755) || 469 die "$0: could not create directory $poddir/$dir: $!\n" 470 unless -d "$poddir/$dir"; 471 472 $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/; 473 $section = ""; 474 $nextsec = $1; 475 476 # for each section of the file create a separate pod file 477 for ($i = 0; $i <= $#poddata; $i++) { 478 # determine the "prev" and "next" links 479 $prevsec = $section; 480 $section = $nextsec; 481 if ($i < $#poddata) { 482 $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/; 483 $nextsec = $1; 484 } else { 485 $nextsec = ""; 486 } 487 488 # determine an appropriate filename (this must correspond with 489 # what pod2html will try and guess) 490 # $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/; 491 $file = "$dir/" . anchorify($section) . ".pod"; 492 493 # create the new .pod file 494 print "\tcreating $poddir/$file\n" if $verbose; 495 open(SPLITOUT, ">$poddir/$file") || 496 die "$0: error opening $poddir/$file for output: $!\n"; 497 $poddata[$i] =~ s,L<([^<>]*)>, 498 defined $heads{anchorify($1)} ? "L<$dir/$1>" : "L<$1>" 499 ,ge; 500 print SPLITOUT $poddata[$i]."\n\n"; 501 print SPLITOUT "=over 4\n\n"; 502 print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec; 503 print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec; 504 print SPLITOUT "=item *\n\nUp to L<$dir>\n\n"; 505 print SPLITOUT "=back\n\n"; 506 close(SPLITOUT) || 507 die "$0: error closing $poddir/$file: $!\n"; 508 } 509} 510 511 512# 513# installdir - takes care of converting the .pod and .pm files in the 514# current directory to .html files and then installing those. 515# 516sub installdir { 517 my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_; 518 my(@dirlist, @podlist, @pmlist, $doindex); 519 520 @dirlist = (); # directories to recurse on 521 @podlist = (); # .pod files to install 522 @pmlist = (); # .pm files to install 523 524 # should files in this directory get an index? 525 $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1); 526 527 opendir(DIR, "$podroot/$dir") 528 || die "$0: error opening directory $podroot/$dir: $!\n"; 529 530 # find the directories to recurse on 531 @dirlist = map { if ($^O eq 'VMS') {/^(.*)\.dir$/i; "$dir/$1";} else {"$dir/$_";}} 532 grep(-d "$podroot/$dir/$_" && !/^\.{1,2}/, readdir(DIR)) if $recurse; 533 rewinddir(DIR); 534 535 # find all the .pod files within the directory 536 @podlist = map { /^(.*)\.pod$/; "$dir/$1" } 537 grep(! -d "$podroot/$dir/$_" && /\.pod$/, readdir(DIR)); 538 rewinddir(DIR); 539 540 # find all the .pm files within the directory 541 @pmlist = map { /^(.*)\.pm$/; "$dir/$1" } 542 grep(! -d "$podroot/$dir/$_" && /\.pm$/, readdir(DIR)); 543 544 closedir(DIR); 545 546 # recurse on all subdirectories we kept track of 547 foreach $dir (@dirlist) { 548 installdir($dir, $recurse, $podroot, $splitdirs, $ignore); 549 } 550 551 # install all the pods we found 552 foreach my $pod (@podlist) { 553 # check if we should ignore it. 554 next if $pod =~ m(/t/); # comes from a test file 555 next if grep($_ eq "$pod.pod", @$ignore); 556 557 # check if a .pm files exists too 558 if (grep($_ eq $pod, @pmlist)) { 559 print "$0: Warning both `$podroot/$pod.pod' and " 560 . "`$podroot/$pod.pm' exist, using pod\n"; 561 push(@ignore, "$pod.pm"); 562 } 563 runpod2html("$pod.pod", $doindex); 564 } 565 566 # install all the .pm files we found 567 foreach my $pm (@pmlist) { 568 # check if we should ignore it. 569 next if $pm =~ m(/t/); # comes from a test file 570 next if grep($_ eq "$pm.pm", @ignore); 571 572 runpod2html("$pm.pm", $doindex); 573 } 574} 575 576 577# 578# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html 579# file. 580# 581sub runpod2html { 582 my($pod, $doindex) = @_; 583 my($html, $i, $dir, @dirs); 584 585 $html = $pod; 586 $html =~ s/\.(pod|pm)$/.html/g; 587 588 # make sure the destination directories exist 589 @dirs = split("/", $html); 590 $dir = "$htmldir/"; 591 for ($i = 0; $i < $#dirs; $i++) { 592 if (! -d "$dir$dirs[$i]") { 593 mkdir("$dir$dirs[$i]", 0755) || 594 die "$0: error creating directory $dir$dirs[$i]: $!\n"; 595 } 596 $dir .= "$dirs[$i]/"; 597 } 598 599 # invoke pod2html 600 print "$podroot/$pod => $htmldir/$html\n" if $verbose; 601 Pod::Html::pod2html( 602 "--htmldir=$htmldir", 603 "--htmlroot=$htmlroot", 604 "--podpath=".join(":", @podpath), 605 "--podroot=$podroot", "--netscape", 606 "--header", 607 ($doindex ? "--index" : "--noindex"), 608 "--" . ($recurse ? "" : "no") . "recurse", 609 ($#libpods >= 0) ? "--libpods=" . join(":", @libpods) : "", 610 "--infile=$podroot/$pod", "--outfile=$htmldir/$html"); 611 die "$0: error running $pod2html: $!\n" if $?; 612} 613