| 1 |
#!/usr/bin/perl -w |
| 2 |
|
| 3 |
=head1 NAME |
| 4 |
|
| 5 |
html2perlstream - convert an HTML document to Perl code for that document |
| 6 |
|
| 7 |
=head1 SYNOPSIS |
| 8 |
|
| 9 |
html2perlstream |
| 10 |
html2perlstream [options] - |
| 11 |
html2perlstream [options] file.html .. file.html |
| 12 |
|
| 13 |
=head1 DESCRIPTION |
| 14 |
|
| 15 |
Takes an HTML file, and produces from it a Perl script that will |
| 16 |
generate that HTML using the L<HTML::Stream> module. |
| 17 |
|
| 18 |
For usage, just say: |
| 19 |
|
| 20 |
html2perlstream |
| 21 |
|
| 22 |
The normal form is: |
| 23 |
|
| 24 |
html2perlstream [-options] [files] |
| 25 |
|
| 26 |
Any named HTML input file F<file.html> will cause output script |
| 27 |
F<file.html.pl> to be generated. To read from the standard input, |
| 28 |
specify '-', like this: |
| 29 |
|
| 30 |
html2perlstream - < test.html |
| 31 |
|
| 32 |
If reading from the standard input, the output Perl code goes to the |
| 33 |
standard output. So if you want to run it right away to see the output, |
| 34 |
just do this: |
| 35 |
|
| 36 |
html2perlstream - < test.html | perl |
| 37 |
|
| 38 |
=head1 OPTIONS |
| 39 |
|
| 40 |
=over 4 |
| 41 |
|
| 42 |
=item B<-d> |
| 43 |
|
| 44 |
Automatically load and run the output Perl file, then do a diff on |
| 45 |
that and the input file (you must input from a file and output to a |
| 46 |
file in order to use this). |
| 47 |
|
| 48 |
html2perlstream -w -d testin/test.html |
| 49 |
|
| 50 |
You'll get fewer differences if you build the code with C<-w>. |
| 51 |
|
| 52 |
=item B<-w> |
| 53 |
|
| 54 |
Try to generate code which will output all whitespace between tags verbatim. |
| 55 |
The default code only outputs whitespace if it is believed to be |
| 56 |
needed; e.g., if we are inside a C<PRE> environment. |
| 57 |
|
| 58 |
=back |
| 59 |
|
| 60 |
|
| 61 |
=head1 REQUIRES |
| 62 |
|
| 63 |
To run this, you need: |
| 64 |
|
| 65 |
HTML::Entities |
| 66 |
HTML::Parser |
| 67 |
|
| 68 |
|
| 69 |
=head1 VERSION |
| 70 |
|
| 71 |
$Id: html2perlstream,v 1.10 2001/08/17 00:50:00 eryq Exp $ |
| 72 |
|
| 73 |
|
| 74 |
=head1 AUTHOR |
| 75 |
|
| 76 |
Eryq, 11 Jan 1997, F<[email protected]> or thereabouts. |
| 77 |
|
| 78 |
Thanks (independently) to Tony Cebzanov and John Buckman for suggesting |
| 79 |
that I write a tool like this. |
| 80 |
|
| 81 |
=cut |
| 82 |
|
| 83 |
# Try this: |
| 84 |
# |
| 85 |
# perl -I. bin/html2perlstream -w testin/test.html; |
| 86 |
# perl -I. testin/test.html.pl > testin/test2.html |
| 87 |
# diff testin/test.html testin/test2.html |
| 88 |
|
| 89 |
use HTML::Parser; |
| 90 |
use HTML::Entities; |
| 91 |
use Getopt::Std; |
| 92 |
use FileHandle; |
| 93 |
|
| 94 |
use strict; |
| 95 |
use vars qw($VERSION |
| 96 |
$opt_d |
| 97 |
$opt_w |
| 98 |
); |
| 99 |
|
| 100 |
#------------------------------------------------------------ |
| 101 |
# |
| 102 |
# Globals... |
| 103 |
# |
| 104 |
#------------------------------------------------------------ |
| 105 |
|
| 106 |
# Version... |
| 107 |
( $VERSION ) = '$Revision: 1.10 $ ' =~ /\$Revision:\s+([^\s]+)/; |
| 108 |
|
| 109 |
# Escape-map: |
| 110 |
my %PerlUnescape = ( |
| 111 |
"\b"=>'b', |
| 112 |
"\f"=>'f', |
| 113 |
"\r"=>'r', |
| 114 |
"\t"=>'t', |
| 115 |
"\n"=>'n', |
| 116 |
); |
| 117 |
my $PerlEscapePat = join('',keys(%PerlUnescape)); |
| 118 |
|
| 119 |
#------------------------------------------------------------ |
| 120 |
# str2perl STRING,[QUOTECHAR] |
| 121 |
#------------------------------------------------------------ |
| 122 |
# Convert a string to a double-quoted Perl string. |
| 123 |
# Unsafe characters and non-printables are escaped. |
| 124 |
# |
| 125 |
# If QUOTECHAR is nonempty (default is '"'), it is escaped and also |
| 126 |
# wrapped around the string. |
| 127 |
|
| 128 |
sub str2perl { |
| 129 |
my ($str, $qq) = @_; |
| 130 |
defined($qq) or $qq = '"'; |
| 131 |
|
| 132 |
$str =~ s/[\\\$\@]/\\$&/g; # unsafe |
| 133 |
$str =~ s/$qq/\\$qq/g if $qq; # quote char |
| 134 |
$str =~ s/[$PerlEscapePat]/\\$PerlUnescape{$&}/og; # newlines, tabs... |
| 135 |
$str =~ s/[\x00-\x1F\x7F-\xFF]/sprintf("\\x%02X", ord($&))/eg; |
| 136 |
"$qq$str$qq"; |
| 137 |
} |
| 138 |
|
| 139 |
|
| 140 |
|
| 141 |
#============================================================ |
| 142 |
package Main::Parser; |
| 143 |
#============================================================ |
| 144 |
|
| 145 |
use HTML::Stream qw(:funcs); |
| 146 |
use vars qw(@ISA); |
| 147 |
|
| 148 |
@ISA = (qw(HTML::Parser)); |
| 149 |
|
| 150 |
|
| 151 |
# Are we inside a preformatter environment? |
| 152 |
my $PREcount = 0; |
| 153 |
|
| 154 |
#------------------------------------------------------------ |
| 155 |
# declaration |
| 156 |
#------------------------------------------------------------ |
| 157 |
# This method is called when a markup declaration has been |
| 158 |
# recognized. For typical HTML documents, the only declaration you are |
| 159 |
# likely to find is <!DOCTYPE ...>. The initial ``<!'' and ending ``>'' |
| 160 |
# is not part of the string passed as argument. Comments |
| 161 |
# are removed and entities have not been expanded yet. |
| 162 |
|
| 163 |
sub declaration { |
| 164 |
my ($self, $decl) = @_; |
| 165 |
|
| 166 |
print '$HTML->io->print(', ::str2perl("<!$decl>"), ");\n"; |
| 167 |
} |
| 168 |
|
| 169 |
#------------------------------------------------------------ |
| 170 |
# start |
| 171 |
#------------------------------------------------------------ |
| 172 |
# This method is called when a complete start tag has been |
| 173 |
# recognized. The first argument is the tag name (in lower case) |
| 174 |
# and the second argument is a reference to a hash that contain |
| 175 |
# all attributes found within the start tag. The attribute keys |
| 176 |
# are converted to lower case. Entities found in the attribute |
| 177 |
# values are already expanded. |
| 178 |
|
| 179 |
sub start { |
| 180 |
my ($self, $tag, $attr) = @_; |
| 181 |
|
| 182 |
# Bookkeeping: |
| 183 |
++$PREcount if ($tag eq 'pre'); |
| 184 |
|
| 185 |
# Output: |
| 186 |
if (keys %$attr) { |
| 187 |
|
| 188 |
# Output and remember first line, up to open paren: |
| 189 |
my $firstline = "\$HTML -> \U$tag\E("; |
| 190 |
print $firstline; |
| 191 |
|
| 192 |
# Determine nice indent: |
| 193 |
my $indent = ' ' x length($firstline); |
| 194 |
$indent =~ s/ {8}/\t/g; |
| 195 |
|
| 196 |
# Output tag params: |
| 197 |
my @keys = sort keys %$attr; |
| 198 |
my $i; |
| 199 |
for ($i = 0; $i < int(@keys); $i++) { |
| 200 |
print "\U$keys[$i]\E => "; |
| 201 |
print ::str2perl($attr->{$keys[$i]}); |
| 202 |
print ",\n$indent" unless ($i == (int(@keys)-1)); |
| 203 |
} |
| 204 |
print ");\n"; |
| 205 |
} |
| 206 |
else { |
| 207 |
print "\$HTML -> \U$tag\E;\n"; |
| 208 |
} |
| 209 |
} |
| 210 |
|
| 211 |
#------------------------------------------------------------ |
| 212 |
# end |
| 213 |
#------------------------------------------------------------ |
| 214 |
# This method is called when an end tag has been recognized. |
| 215 |
# The argument is the lower case tag name. |
| 216 |
|
| 217 |
sub end { |
| 218 |
my ($self, $tag) = @_; |
| 219 |
print "\$HTML -> _\U$tag\E;\n"; |
| 220 |
|
| 221 |
# Bookkeeping: |
| 222 |
--$PREcount if ($tag eq 'pre'); |
| 223 |
} |
| 224 |
|
| 225 |
|
| 226 |
#------------------------------------------------------------ |
| 227 |
# should_output_whitespace |
| 228 |
#------------------------------------------------------------ |
| 229 |
sub should_output_whitespace { |
| 230 |
$::opt_w || ($PREcount > 0); |
| 231 |
} |
| 232 |
|
| 233 |
#------------------------------------------------------------ |
| 234 |
# whitespace |
| 235 |
#------------------------------------------------------------ |
| 236 |
sub whitespace { |
| 237 |
my $ws = shift; |
| 238 |
|
| 239 |
return if (!$ws); |
| 240 |
if ($ws =~ /\A\n*\Z/) { |
| 241 |
print "\$HTML -> nl"; |
| 242 |
print( (length($ws) > 1) ? '('.length($ws).')' : ''); |
| 243 |
print ";\n"; |
| 244 |
} |
| 245 |
else { |
| 246 |
print "\$HTML -> t(", ::str2perl($ws), ");\n"; |
| 247 |
} |
| 248 |
} |
| 249 |
|
| 250 |
#------------------------------------------------------------ |
| 251 |
# text |
| 252 |
#------------------------------------------------------------ |
| 253 |
# This method is called as plain text in the document |
| 254 |
# is recognized. The text is passed on unmodified and might contain |
| 255 |
# multiple lines. Note that for efficiency reasons entities in the |
| 256 |
# text are not expanded. You should call |
| 257 |
# HTML::Entities::decode($text) before you process the text any |
| 258 |
# further. |
| 259 |
|
| 260 |
sub text { |
| 261 |
my ($self, $text) = @_; |
| 262 |
|
| 263 |
# Do nothing if empty string: |
| 264 |
return if (!defined($text) or ($text eq '')); |
| 265 |
|
| 266 |
# Unescape HTML: |
| 267 |
$text = HTML::Entities::decode($text); |
| 268 |
|
| 269 |
# Break into WHITE* NONWHITE* WHITE* |
| 270 |
my ($pre, $in, $dummy, $post) = ($text =~ /\A(\s*)((.|\n)*?)(\s*)\Z/); |
| 271 |
|
| 272 |
# Output leading whitespace: |
| 273 |
if ($pre && should_output_whitespace()) { |
| 274 |
whitespace($pre); |
| 275 |
} |
| 276 |
|
| 277 |
# Output main text: |
| 278 |
if (defined($in) && ($in ne '')) { |
| 279 |
if (length($in) < 70) { # output as t() |
| 280 |
print "\$HTML -> t(", ::str2perl($in), ");\n"; |
| 281 |
} |
| 282 |
else { # output as hereis() |
| 283 |
my $hereis = ::str2perl($in, ''); |
| 284 |
$hereis =~ s/\\n/\n/g; |
| 285 |
|
| 286 |
print "output \$HTML <<EOF;\n"; |
| 287 |
print $hereis; |
| 288 |
print "\nEOF\n"; |
| 289 |
} |
| 290 |
} |
| 291 |
|
| 292 |
# Output trailing whitespace: |
| 293 |
if ($post && should_output_whitespace()) { |
| 294 |
whitespace($post); |
| 295 |
} |
| 296 |
1; |
| 297 |
} |
| 298 |
|
| 299 |
#------------------------------------------------------------ |
| 300 |
# comment |
| 301 |
#------------------------------------------------------------ |
| 302 |
# This method is called as comments are recognized. |
| 303 |
# The leading and trailing ``--'' sequences has been stripped off |
| 304 |
# the comment text. |
| 305 |
|
| 306 |
sub comment { |
| 307 |
my ($self, $comment) = @_; |
| 308 |
|
| 309 |
# Remove lead/trail single space (we'll put it back, honest): |
| 310 |
$comment =~ s/^ //; |
| 311 |
$comment =~ s/ $//; |
| 312 |
|
| 313 |
print "\$HTML -> comment(", ::str2perl($comment), ");\n"; |
| 314 |
} |
| 315 |
|
| 316 |
|
| 317 |
#============================================================ |
| 318 |
package main; |
| 319 |
#============================================================ |
| 320 |
|
| 321 |
#------------------------------------------------------------ |
| 322 |
# print_header |
| 323 |
#------------------------------------------------------------ |
| 324 |
sub print_header { |
| 325 |
print <<EOF; |
| 326 |
#!/usr/bin/perl -Tw |
| 327 |
|
| 328 |
use HTML::Stream; |
| 329 |
\$HTML = new HTML::Stream \\*STDOUT; |
| 330 |
EOF |
| 331 |
|
| 332 |
if ($opt_w) { |
| 333 |
print <<EOF; |
| 334 |
|
| 335 |
# You generated this code with "html2perlstream -w", so we turn off |
| 336 |
# auto-formatting to make things look right: |
| 337 |
\$HTML->auto_format(0); |
| 338 |
EOF |
| 339 |
} |
| 340 |
else { |
| 341 |
print <<EOF; |
| 342 |
|
| 343 |
# You generated this code without "html2perlstream -w", so we keep |
| 344 |
# auto-formatting on to add some whitespace here and there: |
| 345 |
# \$HTML->auto_format(0); |
| 346 |
EOF |
| 347 |
} |
| 348 |
|
| 349 |
print <<EOF; |
| 350 |
|
| 351 |
# BEGIN GENERATED CODE |
| 352 |
EOF |
| 353 |
} |
| 354 |
|
| 355 |
#------------------------------------------------------------ |
| 356 |
# print_fooer |
| 357 |
#------------------------------------------------------------ |
| 358 |
sub print_footer { |
| 359 |
print <<EOF; |
| 360 |
# END GENERATED CODE |
| 361 |
1; |
| 362 |
EOF |
| 363 |
} |
| 364 |
|
| 365 |
#------------------------------------------------------------ |
| 366 |
# usage |
| 367 |
#------------------------------------------------------------ |
| 368 |
sub usage { |
| 369 |
print STDERR <<EOF; |
| 370 |
|
| 371 |
html2perlstream version $VERSION by Eryq, Jan 1997, eryq\@zeegee.com. |
| 372 |
|
| 373 |
Usage: html2perlstream [options] [files...] |
| 374 |
-d automatically run Perl on output, and diff with input |
| 375 |
-w output code to output whitespace verbatim |
| 376 |
|
| 377 |
EOF |
| 378 |
exit(-1); |
| 379 |
} |
| 380 |
|
| 381 |
#------------------------------------------------------------ |
| 382 |
# do_diff |
| 383 |
#------------------------------------------------------------ |
| 384 |
sub do_diff { |
| 385 |
my ($htmlfile, $perlfile) = @_; |
| 386 |
my $html2 = "$perlfile.out"; |
| 387 |
my $diff = "$perlfile.diff"; |
| 388 |
|
| 389 |
# Sync up output: |
| 390 |
STDOUT->flush; |
| 391 |
STDERR->flush; |
| 392 |
|
| 393 |
# Do it! |
| 394 |
my $cmd = join(' ', ($^X, (map {"-I$_"} @INC), $perlfile)); |
| 395 |
# print STDERR "diff command = $cmd\n"; |
| 396 |
print STDERR " Running code: output in $html2\n"; |
| 397 |
system "$cmd > $html2"; |
| 398 |
print STDERR " Doing diff: output in $diff\n"; |
| 399 |
system "diff $htmlfile $html2 > $diff"; |
| 400 |
} |
| 401 |
|
| 402 |
#------------------------------------------------------------ |
| 403 |
# main |
| 404 |
#------------------------------------------------------------ |
| 405 |
sub main { |
| 406 |
my $parser = new Main::Parser; |
| 407 |
|
| 408 |
# Usage if no args: |
| 409 |
@ARGV or usage(); |
| 410 |
|
| 411 |
# Get opts: |
| 412 |
getopts "wd"; |
| 413 |
|
| 414 |
# Process files: |
| 415 |
@ARGV or unshift @ARGV, '-'; |
| 416 |
my $infile; |
| 417 |
while ($infile = shift @ARGV) { |
| 418 |
my $outfile = (($infile eq '-') ? '-' : "$infile.pl"); |
| 419 |
|
| 420 |
# Trap for bad -d usage: |
| 421 |
if ($opt_d and (($infile eq '-') || ($outfile eq '-'))) { |
| 422 |
die "Cannot use -d if input is from STDIN\n"; |
| 423 |
} |
| 424 |
|
| 425 |
# Report: |
| 426 |
print STDERR "Converting ", ($infile eq '-' ? '<STDIN>' : $infile), |
| 427 |
" to " , ($outfile eq '-' ? '<STDOUT>' : $outfile), |
| 428 |
"\n"; |
| 429 |
|
| 430 |
# Open input and output: |
| 431 |
open INPUT, "<$infile" or die "$infile: $!"; |
| 432 |
open OUTPUT, ">$outfile" or die "$outfile: $!"; |
| 433 |
select OUTPUT; |
| 434 |
|
| 435 |
# Header: |
| 436 |
print_header(); |
| 437 |
$parser->parse_file(\*INPUT); |
| 438 |
|
| 439 |
# Footer: |
| 440 |
print_footer(); |
| 441 |
|
| 442 |
# Close: |
| 443 |
close OUTPUT; |
| 444 |
close INPUT; |
| 445 |
|
| 446 |
# Do a diff... |
| 447 |
do_diff($infile, $outfile) if ($opt_d); |
| 448 |
|
| 449 |
} |
| 450 |
print STDERR "Done.\n"; |
| 451 |
1; |
| 452 |
} |
| 453 |
exit (&main ? 0 : -1); |
| 454 |
|
| 455 |
#------------------------------------------------------------ |
| 456 |
1; |