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::Functions qw(rel2abs no_upwards); 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 [--ignore=<name>,...,<name>] [--verbose] 22 23=head1 DESCRIPTION 24 25I<installhtml> converts a collection of POD pages to a corresponding 26collection of HTML pages. This is primarily used to convert the pod 27pages found in the perl distribution. 28 29=head1 OPTIONS 30 31=over 4 32 33=item B<--help> help 34 35Displays the usage. 36 37=item B<--podroot> POD search path base directory 38 39The base directory to search for all .pod and .pm files to be converted. 40Default is current directory. 41 42=item B<--podpath> POD search path 43 44The list of directories to search for .pod and .pm files to be converted. 45Default is 'podroot/.'. 46 47=item B<--recurse> recurse on subdirectories 48 49Whether or not to convert all .pm and .pod files found in subdirectories 50too. Default is to not recurse. 51 52=item B<--htmldir> HTML destination directory 53 54The base directory which all HTML files will be written to. This should 55be a path relative to the filesystem, not the resulting URL. 56 57=item B<--htmlroot> URL base directory 58 59The base directory which all resulting HTML files will be visible at in 60a URL. The default is '/'. 61 62=item B<--splithead> POD files to split on =head directive 63 64Comma-separated list of pod files to split by the =head directive. The 65.pod suffix is optional. These files should have names specified 66relative to podroot. 67 68=item B<--splititem> POD files to split on =item directive 69 70Comma-separated list of all pod files to split by the =item directive. 71The .pod suffix is optional. I<installhtml> does not do the actual 72split, rather it invokes I<splitpod> to do the dirty work. As with 73--splithead, these files should have names specified relative to podroot. 74 75=item B<--splitpod> Directory containing the splitpod program 76 77The directory containing the splitpod program. The default is 'podroot/pod'. 78 79=item B<--ignore> files to be ignored 80 81Comma-separated of files that shouldn't be installed, given relative 82to podroot. 83 84=item B<--verbose> verbose output 85 86Self-explanatory. 87 88=back 89 90=head1 EXAMPLE 91 92The following command-line is an example of the one we use to convert 93perl documentation: 94 95 ./installhtml --podpath=lib:ext:pod:vms \ 96 --podroot=/usr/src/perl \ 97 --htmldir=/perl/nmanual \ 98 --htmlroot=/perl/nmanual \ 99 --splithead=pod/perlipc \ 100 --splititem=pod/perlfunc \ 101 --recurse \ 102 --verbose 103 104=head1 AUTHOR 105 106Chris Hall E<lt>hallc@cs.colorado.eduE<gt> 107 108=cut 109 110my $usage; 111 112$usage =<<END_OF_USAGE; 113Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name> 114 --htmldir=<name> --htmlroot=<name> --norecurse --recurse 115 --splithead=<name>,...,<name> --splititem=<name>,...,<name> 116 --ignore=<name>,...,<name> --verbose 117 118 --help - this message 119 --podpath - colon-separated list of directories containing .pod and 120 .pm files to be converted (. by default). 121 --podroot - filesystem base directory from which all relative paths in 122 podpath stem (default is .). 123 --htmldir - directory to store resulting html files in relative 124 to the filesystem (\$podroot/html by default). 125 --htmlroot - http-server base directory from which all relative paths 126 in podpath stem (default is /). 127 --norecurse - don't recurse on those subdirectories listed in podpath. 128 (default behavior). 129 --recurse - recurse on those subdirectories listed in podpath 130 --splithead - comma-separated list of .pod or .pm files to split. will 131 split each file into several smaller files at every occurrence 132 of a pod =head[1-6] directive. 133 --splititem - comma-separated list of .pod or .pm files to split using 134 splitpod. 135 --splitpod - directory where the program splitpod can be found 136 (\$podroot/pod by default). 137 --ignore - comma-separated list of files that shouldn't be installed. 138 --verbose - self-explanatory. 139 140END_OF_USAGE 141 142my (@podpath, $podroot, $htmldir, $htmlroot, $recurse, @splithead, 143 @splititem, $splitpod, $verbose, $pod2html, @ignore); 144 145@podpath = ( "." ); # colon-separated list of directories containing .pod 146 # and .pm files to be converted. 147$podroot = "."; # assume the pods we want are here 148$htmldir = ""; # nothing for now... 149$htmlroot = "/"; # default value 150$recurse = 0; # default behavior 151@splithead = (); # don't split any files by default 152@splititem = (); # don't split any files by default 153$splitpod = ""; # nothing for now. 154 155$verbose = 0; # whether or not to print debugging info 156 157$pod2html = "pod/pod2html"; 158 159usage("") unless @ARGV; 160 161# Overcome shell's p1,..,p8 limitation. 162# See vms/descrip_mms.template -> descrip.mms for invocation. 163if ( $^O eq 'VMS' ) { @ARGV = split(/\s+/,$ARGV[0]); } 164 165use vars qw( %Options ); 166 167# parse the command-line 168my $result = GetOptions( \%Options, qw( 169 help 170 podpath=s 171 podroot=s 172 htmldir=s 173 htmlroot=s 174 ignore=s 175 recurse! 176 splithead=s 177 splititem=s 178 splitpod=s 179 verbose 180)); 181usage("invalid parameters") unless $result; 182parse_command_line(); 183 184 185# set these variables to appropriate values if the user didn't specify 186# values for them. 187$htmldir = "$htmlroot/html" unless $htmldir; 188$splitpod = "$podroot/pod" unless $splitpod; 189 190 191# make sure that the destination directory exists 192(mkdir($htmldir, 0755) || 193 die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir; 194 195 196# the following array will eventually contain files that are to be 197# ignored in the conversion process. these are files that have been 198# process by splititem or splithead and should not be converted as a 199# result. 200my @splitdirs; 201 202# split pods. It's important to do this before convert ANY pods because 203# it may affect some of the links 204@splitdirs = (); # files in these directories won't get an index 205split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead); 206split_on_item($podroot, \@splitdirs, \@ignore, @splititem); 207 208 209# convert the pod pages found in @poddirs 210#warn "converting files\n" if $verbose; 211#warn "\@ignore\t= @ignore\n" if $verbose; 212foreach my $dir (@podpath) { 213 installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore); 214} 215 216 217# now go through and create master indices for each pod we split 218foreach my $dir (@splititem) { 219 print "creating index $htmldir/$dir.html\n" if $verbose; 220 create_index("$htmldir/$dir.html", "$htmldir/$dir"); 221} 222 223foreach my $dir (@splithead) { 224 (my $pod = $dir) =~ s,^.*/,,; 225 $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/; 226 # let pod2html create the file 227 runpod2html($dir, 1); 228 229 # now go through and truncate after the index 230 $dir =~ /^(.*?)(\.pod|\.pm)?$/sm; 231 my $file = "$htmldir/$1"; 232 print "creating index $file.html\n" if $verbose; 233 234 # read in everything until what would have been the first =head 235 # directive, patching the index as we go. 236 open(H, "<$file.html") || 237 die "$0: error opening $file.html for input: $!\n"; 238 $/ = ""; 239 my @data = (); 240 while (<H>) { 241 last if /name="name"/i; 242 $_ =~ s{href="#(.*)">}{ 243 my $url = "$pod/$1.html" ; 244 $url = Pod::Html::relativize_url( $url, "$file.html" ) 245 if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ); 246 "href=\"$url\">" ; 247 }egi; 248 push @data, $_; 249 } 250 close(H); 251 252 # now rewrite the file 253 open(H, ">$file.html") || 254 die "$0: error opening $file.html for output: $!\n"; 255 print H "@data", "\n"; 256 close(H); 257} 258 259############################################################################## 260 261 262sub usage { 263 warn "$0: @_\n" if @_; 264 die $usage; 265} 266 267 268sub parse_command_line { 269 usage() if defined $Options{help}; 270 $Options{help} = ""; # make -w shut up 271 272 # list of directories 273 @podpath = split(":", $Options{podpath}) if defined $Options{podpath}; 274 275 # lists of files 276 @splithead = split(",", $Options{splithead}) if defined $Options{splithead}; 277 @splititem = split(",", $Options{splititem}) if defined $Options{splititem}; 278 279 $htmldir = $Options{htmldir} if defined $Options{htmldir}; 280 $htmlroot = $Options{htmlroot} if defined $Options{htmlroot}; 281 $podroot = $Options{podroot} if defined $Options{podroot}; 282 $splitpod = $Options{splitpod} if defined $Options{splitpod}; 283 284 $recurse = $Options{recurse} if defined $Options{recurse}; 285 $verbose = $Options{verbose} if defined $Options{verbose}; 286 287 @ignore = map "$podroot/$_", split(",", $Options{ignore}) if defined $Options{ignore}; 288} 289 290 291sub create_index { 292 my($html, $dir) = @_; 293 (my $pod = $dir) =~ s,^.*/,,; 294 295 # get the list of .html files in this directory 296 opendir(DIR, $dir) || 297 die "$0: error opening directory $dir for reading: $!\n"; 298 my @files = sort(grep(/\.html?$/, readdir(DIR))); 299 closedir(DIR); 300 301 open(HTML, ">$html") || 302 die "$0: error opening $html for output: $!\n"; 303 304 # for each .html file in the directory, extract the index 305 # embedded in the file and throw it into the big index. 306 print HTML "<DL COMPACT>\n"; 307 foreach my $file (@files) { 308 309 my $filedata = do { 310 open(my $in, "<$dir/$file") || 311 die "$0: error opening $dir/$file for input: $!\n"; 312 local $/ = undef; 313 <$in>; 314 }; 315 316 # pull out the NAME section 317 my($lcp1, $lcp2) = 318 ($filedata =~ 319 m#<h1 id="NAME">NAME</h1>\s*<p>\s*(\S+)\s+-\s+(\S.*?\S)</p>#); 320 defined $lcp1 or die "$0: can't find NAME section in $dir/$file\n"; 321 322 my $url= "$pod/$file" ; 323 if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ) { 324 $url = Pod::Html::relativize_url( "$pod/$file", $html ) ; 325 } 326 327 print HTML qq(<DT><A HREF="$url">); 328 print HTML "$lcp1</A></DT><DD>$lcp2</DD>\n"; 329 } 330 print HTML "</DL>\n"; 331 332 close(HTML); 333} 334 335 336sub split_on_head { 337 my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_; 338 my($pod, $dirname, $filename); 339 340 # split the files specified in @splithead on =head[1-6] pod directives 341 print "splitting files by head.\n" if $verbose && $#splithead >= 0; 342 foreach $pod (@splithead) { 343 # figure out the directory name and filename 344 $pod =~ s,^([^/]*)$,/$1,; 345 $pod =~ m,(.*)/(.*?)(\.pod)?$,; 346 $dirname = $1; 347 $filename = "$2.pod"; 348 349 # since we are splitting this file it shouldn't be converted. 350 push(@$ignore, "$podroot/$dirname/$filename"); 351 352 # split the pod 353 splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir, 354 $splitdirs); 355 } 356} 357 358 359sub split_on_item { 360 my($podroot, $splitdirs, $ignore, @splititem) = @_; 361 my($pwd, $dirname, $filename); 362 363 print "splitting files by item.\n" if $verbose && $#splititem >= 0; 364 $pwd = getcwd(); 365 my $splitter = rel2abs("$splitpod/splitpod", $pwd); 366 my $perl = rel2abs($^X, $pwd); 367 foreach my $pod (@splititem) { 368 # figure out the directory to split into 369 $pod =~ s,^([^/]*)$,/$1,; 370 $pod =~ m,(.*)/(.*?)(\.pod)?$,; 371 $dirname = "$1/$2"; 372 $filename = "$2.pod"; 373 374 # since we are splitting this file it shouldn't be converted. 375 push(@$ignore, "$podroot/$dirname.pod"); 376 377 # split the pod 378 push(@$splitdirs, "$podroot/$dirname"); 379 if (! -d "$podroot/$dirname") { 380 mkdir("$podroot/$dirname", 0755) || 381 die "$0: error creating directory $podroot/$dirname: $!\n"; 382 } 383 chdir("$podroot/$dirname") || 384 die "$0: error changing to directory $podroot/$dirname: $!\n"; 385 die "$splitter not found. Use '-splitpod dir' option.\n" 386 unless -f $splitter; 387 system($perl, $splitter, "../$filename") && 388 warn "$0: error running '$splitter ../$filename'" 389 ." from $podroot/$dirname"; 390 } 391 chdir($pwd); 392} 393 394 395# 396# splitpod - splits a .pod file into several smaller .pod files 397# where a new file is started each time a =head[1-6] pod directive 398# is encountered in the input file. 399# 400sub splitpod { 401 my($pod, $poddir, $htmldir, $splitdirs) = @_; 402 my(@poddata, @filedata, @heads); 403 my($file, $i, $j, $prevsec, $section, $nextsec); 404 405 print "splitting $pod\n" if $verbose; 406 407 # read the file in paragraphs 408 $/ = ""; 409 open(SPLITIN, "<$pod") || 410 die "$0: error opening $pod for input: $!\n"; 411 @filedata = <SPLITIN>; 412 close(SPLITIN) || 413 die "$0: error closing $pod: $!\n"; 414 415 # restore the file internally by =head[1-6] sections 416 @poddata = (); 417 for ($i = 0, $j = -1; $i <= $#filedata; $i++) { 418 $j++ if ($filedata[$i] =~ /^\s*=head[1-6]/); 419 if ($j >= 0) { 420 $poddata[$j] = "" unless defined $poddata[$j]; 421 $poddata[$j] .= "\n$filedata[$i]" if $j >= 0; 422 } 423 } 424 425 # create list of =head[1-6] sections so that we can rewrite 426 # L<> links as necessary. 427 my %heads = (); 428 foreach $i (0..$#poddata) { 429 $heads{anchorify($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/; 430 } 431 432 # create a directory of a similar name and store all the 433 # files in there 434 $pod =~ s,.*/(.*),$1,; # get the last part of the name 435 my $dir = $pod; 436 $dir =~ s/\.pod//g; 437 push(@$splitdirs, "$poddir/$dir"); 438 mkdir("$poddir/$dir", 0755) || 439 die "$0: could not create directory $poddir/$dir: $!\n" 440 unless -d "$poddir/$dir"; 441 442 $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/; 443 $section = ""; 444 $nextsec = $1; 445 446 # for each section of the file create a separate pod file 447 for ($i = 0; $i <= $#poddata; $i++) { 448 # determine the "prev" and "next" links 449 $prevsec = $section; 450 $section = $nextsec; 451 if ($i < $#poddata) { 452 $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/; 453 $nextsec = $1; 454 } else { 455 $nextsec = ""; 456 } 457 458 # determine an appropriate filename (this must correspond with 459 # what pod2html will try and guess) 460 # $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/; 461 $file = "$dir/" . anchorify($section) . ".pod"; 462 463 # create the new .pod file 464 print "\tcreating $poddir/$file\n" if $verbose; 465 open(SPLITOUT, ">$poddir/$file") || 466 die "$0: error opening $poddir/$file for output: $!\n"; 467 $poddata[$i] =~ s,L<([^<>]*)>, 468 defined $heads{anchorify($1)} ? "L<$dir/$1>" : "L<$1>" 469 ,ge; 470 print SPLITOUT $poddata[$i]."\n\n"; 471 print SPLITOUT "=over 4\n\n"; 472 print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec; 473 print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec; 474 print SPLITOUT "=item *\n\nUp to L<$dir>\n\n"; 475 print SPLITOUT "=back\n\n"; 476 close(SPLITOUT) || 477 die "$0: error closing $poddir/$file: $!\n"; 478 } 479} 480 481 482# 483# installdir - takes care of converting the .pod and .pm files in the 484# current directory to .html files and then installing those. 485# 486sub installdir { 487 my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_; 488 489 my @dirlist; # directories to recurse on 490 my @podlist; # .pod files to install 491 my @pmlist; # .pm files to install 492 493 # should files in this directory get an index? 494 my $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1); 495 496 opendir(DIR, "$podroot/$dir") 497 || die "$0: error opening directory $podroot/$dir: $!\n"; 498 499 while(readdir DIR) { 500 no_upwards($_) or next; 501 my $is_dir = -d "$podroot/$dir/$_"; 502 next if $is_dir and not $recurse; 503 my $target = ( 504 $is_dir ? \@dirlist : 505 s/\.pod$// ? \@podlist : 506 s/\.pm$// ? \@pmlist : 507 undef 508 ); 509 push @$target, "$dir/$_" if $target; 510 } 511 512 closedir(DIR); 513 514 if ($^O eq 'VMS') { s/\.dir$//i for @dirlist } 515 516 # recurse on all subdirectories we kept track of 517 foreach $dir (@dirlist) { 518 installdir($dir, $recurse, $podroot, $splitdirs, $ignore); 519 } 520 521 # install all the pods we found 522 foreach my $pod (@podlist) { 523 # check if we should ignore it. 524 next if $pod =~ m(/t/); # comes from a test file 525 next if grep($_ eq "$pod.pod", @$ignore); 526 527 # check if a .pm files exists too 528 if (grep($_ eq $pod, @pmlist)) { 529 print "$0: Warning both '$podroot/$pod.pod' and " 530 . "'$podroot/$pod.pm' exist, using pod\n"; 531 push(@ignore, "$pod.pm"); 532 } 533 runpod2html("$pod.pod", $doindex); 534 } 535 536 # install all the .pm files we found 537 foreach my $pm (@pmlist) { 538 # check if we should ignore it. 539 next if $pm =~ m(/t/); # comes from a test file 540 next if grep($_ eq "$pm.pm", @ignore); 541 542 runpod2html("$pm.pm", $doindex); 543 } 544} 545 546 547# 548# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html 549# file. 550# 551sub runpod2html { 552 my($pod, $doindex) = @_; 553 my($html, $i, $dir, @dirs); 554 555 $html = $pod; 556 $html =~ s/\.(pod|pm)$/.html/g; 557 558 # make sure the destination directories exist 559 @dirs = split("/", $html); 560 $dir = "$htmldir/"; 561 for ($i = 0; $i < $#dirs; $i++) { 562 if (! -d "$dir$dirs[$i]") { 563 mkdir("$dir$dirs[$i]", 0755) || 564 die "$0: error creating directory $dir$dirs[$i]: $!\n"; 565 } 566 $dir .= "$dirs[$i]/"; 567 } 568 569 # invoke pod2html 570 print "$podroot/$pod => $htmldir/$html\n" if $verbose; 571 Pod::Html::pod2html( 572 "--htmldir=$htmldir", 573 "--htmlroot=$htmlroot", 574 "--podpath=".join(":", @podpath), 575 "--podroot=$podroot", 576 "--header", 577 ($doindex ? "--index" : "--noindex"), 578 "--" . ($recurse ? "" : "no") . "recurse", 579 "--infile=$podroot/$pod", "--outfile=$htmldir/$html"); 580 die "$0: error running $pod2html: $!\n" if $?; 581} 582