3 # Test filter partially based on Ambrose's hideous subst.pl code
4 # The idea is that the .tmpl files are not valid HTML, and as a result
5 # HTML::Parse would be completely confused by these templates.
6 # This is just a simple scanner (not a parser) & should give better results.
8 # This script is meant to be a drop-in replacement of text-extract.pl
10 # A grander plan: Code could be written to detect template variables and
11 # construct gettext-c-format-string-like meta-strings (e.g., "Results %s
12 # through %s of %s records" that will be more likely to be translatable
13 # to languages where word order is very unlike English word order.
14 # --> This will be relatively major rework, and requires corresponding
15 # rework in tmpl_process.pl
20 use vars qw( $input );
21 use vars qw( $debug_dump_only_p );
22 use vars qw( $pedantic_p $pedantic_tag );
23 use vars qw( $pedantic_attribute_error_in_nonpedantic_mode_p );
24 use vars qw( $pedantic_tmpl_var_use_in_nonpedantic_mode_p );
25 use vars qw( $fatal_p );
27 ###############################################################################
30 use vars qw( $re_directive $re_tmpl_var $re_tmpl_var_escaped $re_tmpl_include );
31 use vars qw( $re_directive_control $re_tmpl_endif_endloop );
33 # $re_directive must not do any backreferences
34 $re_directive = q{<(?:(?i)(?:!--\s*)?\/?TMPL_(?:VAR|LOOP|INCLUDE|IF|ELSE|UNLESS)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
35 # TMPL_VAR or TMPL_INCLUDE
36 $re_tmpl_var = q{<(?:(?i)(?:!--\s*)?TMPL_(?:VAR)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
37 $re_tmpl_include = q{<(?:(?i)(?:!--\s*)?TMPL_(?:INCLUDE)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
38 # TMPL_VAR ESCAPE=1/HTML/URL
39 $re_tmpl_var_escaped = q{<(?:(?i)(?:!--\s*)?TMPL_(?:VAR|INCLUDE)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))\s+ESCAPE=(?:1|HTML|URL)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
40 # Any control flow directive
41 $re_directive_control = q{<(?:(?i)(?:!--\s*)?\/?TMPL_(?:LOOP|IF|ELSE|UNLESS)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
42 # /LOOP or /IF or /UNLESS
43 $re_tmpl_endif_endloop = q{<(?:(?i)(?:!--\s*)?\/TMPL_(?:LOOP|IF|UNLESS)(?:\s+(?:[a-zA-Z][-a-zA-Z0-9]*=)?(?:'[^']*'|"[^"]*"|[^\s<>]+))*\s*(?:--)?)>};
46 # Hideous stuff from subst.pl, slightly modified to use the above hideous stuff
47 # Note: The $re_tag's set $1 (<tag), $2 (>), and $3 (rest of string)
48 use vars qw( $re_comment $re_entity_name $re_end_entity $re_etag );
49 use vars qw( $re_tag_strict $re_tag_compat @re_tag );
52 my $etag = $compat? '>': '<>\/';
53 # This is no longer similar to the original regexp in subst.pl :-(
54 # Note that we don't want <> in compat mode; Mozilla knows about <
55 q{(<\/?(?:|(?:"(?:} . $re_directive . q{|[^"])*"|'(?:} . $re_directive . q{|[^'])*'|--(?:[^-]|-[^-])*--|(?:}
57 . q{|(?!--)[^"'<>} . $etag . q{]))+))([} . $etag . q{]|(?=<))(.*)};
60 $re_comment = '(?:--(?:[^-]|-[^-])*--)';
61 $re_entity_name = '(?:[^&%#;<>\s]+)'; # NOTE: not really correct SGML
62 $re_end_entity = '(?:;|$|(?=\s))'; # semicolon or before-whitespace
63 $re_etag = q{(?:<\/?(?:"[^"]*"|'[^']*'|[^"'>\/])*[>\/])}; # end-tag
64 @re_tag = ($re_tag_strict, $re_tag_compat) = (re_tag(0), re_tag(1));
67 # End of the hideous stuff
69 sub KIND_TEXT () { 'TEXT' }
70 sub KIND_CDATA () { 'CDATA' }
71 sub KIND_TAG () { 'TAG' }
72 sub KIND_DECL () { 'DECL' }
73 sub KIND_PI () { 'PI' }
74 sub KIND_DIRECTIVE () { 'HTML::Template' }
75 sub KIND_COMMENT () { 'COMMENT' } # empty DECL with exactly one SGML comment
76 sub KIND_UNKNOWN () { 'ERROR' }
78 use vars qw( $readahead $lc_0 $lc $syntaxerror_p );
79 use vars qw( $cdata_mode_p $cdata_close );
81 ###############################################################################
83 use vars qw( $appName $input_abbr );
85 sub construct_warn_prefix ($$) {
86 my($prefix, $lc) = @_;
87 # Construct some short but should-be-still-useful versions
88 # of this script's name and the input file's name
89 my $appName = $& if !defined $appName && $0 =~ /[^\/]+$/;
90 my $input_abbr = $& if !defined $input_abbr && $input =~ /[^\/]+$/;
91 # FIXME: The line number is not accurate, but should be "close enough"
92 # FIXME: This wording is worse than what was there, but it's wrong to
93 # FIXME: hard-code this thing in each warn statement. Need improvement.
94 return "$appName: $prefix: " . (defined $lc? "$input_abbr: line $lc: ": "$input_abbr: ");
97 sub warn_normal ($$) {
99 my $prefix = construct_warn_prefix('Warning', $lc);
100 $msg .= "\n" unless $msg =~ /\n$/s;
104 sub warn_pedantic ($$$) {
105 my($msg, $lc, $flag) = @_;
106 my $prefix = construct_warn_prefix("Warning$pedantic_tag", $lc);
107 $msg .= "\n" unless $msg =~ /\n$/s;
108 warn "$prefix$msg" if $pedantic_p || !$$flag;
110 $prefix = construct_warn_prefix("Warning$pedantic_tag", undef);
111 warn $prefix."Further similar negligible warnings will not be reported, use --pedantic for details\n" unless $$flag;
116 ###############################################################################
118 sub extract_attributes ($;$) {
121 $s = $1 if $s =~ /^<\S+(.*)\/\S$/s # XML-style self-closing tags
122 || $s =~ /^<\S+(.*)\S$/s; # SGML-style tags
124 for (my $i = 0; $s =~ /^(?:$re_directive_control)?\s+(?:$re_directive_control)?(?:([a-zA-Z][-a-zA-Z0-9]*)\s*=\s*)?('((?:$re_directive|[^'])*)'|"((?:$re_directive|[^"])*)"|((?:$re_directive|[^\s<>])+))/os;) {
125 my($key, $val, $val_orig, $rest)
126 = ($1, (defined $3? $3: defined $4? $4: $5), $2, $');
128 $attr{+lc($key)} = [$key, $val, $val_orig, $i];
130 if ($val =~ /$re_tmpl_include/os) {
131 warn_normal "TMPL_INCLUDE in attribute: $val_orig\n", $lc;
132 } elsif ($val =~ /$re_tmpl_var/os && $val !~ /$re_tmpl_var_escaped/os) {
133 # XXX: we probably should not warn if key is "onclick" etc
134 # XXX: there's just no reasonable thing to suggest
135 my $suggest = ($key =~ /^(?:action|archive|background|cite|classid|codebase|data|datasrc|for|href|longdesc|profile|src|usemap)$/i? 'URL': 'HTML');
136 undef $suggest if $key =~ /^(?:onblur|onchange|onclick|ondblclick|onfocus|onkeydown|onkeypress|onkeyup|onload|onmousedown|onmousemove|onmouseout|onmouseover|onmouseup|onreset|onselect|onsubmit|onunload)$/i;
138 "Suggest ESCAPE=$suggest for TMPL_VAR in attribute \"$key\""
140 $lc, \$pedantic_tmpl_var_use_in_nonpedantic_mode_p
141 if defined $suggest && ($pedantic_p || !$pedantic_tmpl_var_use_in_nonpedantic_mode_p);
142 } elsif ($val_orig !~ /^['"]/) {
143 my $t = $val; $t =~ s/$re_directive_control//os;
145 "Unquoted attribute contains character(s) that should be quoted"
147 $lc, \$pedantic_attribute_error_in_nonpedantic_mode_p
148 if $t =~ /[^-\.A-Za-z0-9]/s;
151 my $s2 = $s; $s2 =~ s/$re_tmpl_endif_endloop//g; # for the next check
152 if ($s2 =~ /\S/s) { # should never happen
153 if ($s =~ /^([^\n]*)\n/s) { # this is even worse
154 my $prefix = construct_warn_prefix('Error: ', $lc);
155 warn $prefix . "Completely confused while extracting attributes"
157 warn $prefix . (scalar split(/\n/, $s) - 1) . " more line(s) not shown.\n";
160 warn "Strange attribute syntax: $s\n", $lc;
166 sub next_token_internal (*) {
170 if (!defined $readahead || !length $readahead) {
171 my $next = scalar <$h>;
172 $eof_p = !defined $next;
178 $lc_0 = $lc; # remember line number of first line
179 if ($eof_p && !length $readahead) { # nothing left to do
181 } elsif ($readahead =~ /^\s+/s) { # whitespace
182 ($kind, $it, $readahead) = (KIND_TEXT, $&, $');
183 # FIXME the following (the [<\s] part) is an unreliable HACK :-(
184 } elsif ($readahead =~ /^(?:[^<]|<[<\s])+/s) { # non-space normal text
185 ($kind, $it, $readahead) = (KIND_TEXT, $&, $');
186 warn "Warning: Unescaped < $it\n", $lc_0
187 if !$cdata_mode_p && $it =~ /</s;
188 } else { # tag/declaration/processing instruction
192 if ($readahead =~ /^$cdata_close/) {
193 ($kind, $it, $readahead) = (KIND_TAG, $&, $');
196 ($kind, $it, $readahead) = (KIND_TEXT, $readahead, undef);
199 } elsif ($readahead =~ /^$re_tag_compat/os) {
200 ($kind, $it, $readahead) = (KIND_TAG, "$1>", $3);
202 warn "SGML \"closed start tag\" notation: $1<\n", $lc_0 if $2 eq '';
203 } elsif ($readahead =~ /^<!--(?:(?!-->).)*-->/s) {
204 ($kind, $it, $readahead) = (KIND_COMMENT, $&, $');
206 warn "Syntax error in comment: $&\n", $lc_0;
210 my $next = scalar <$h>;
211 $eof_p = !defined $next;
216 if ($kind ne KIND_TAG) {
218 } elsif ($it =~ /^<!/) {
220 $kind = KIND_COMMENT if $it =~ /^<!--(?:(?!-->).)*-->/;
221 } elsif ($it =~ /^<\?/) {
224 if ($it =~ /^$re_directive/ios && !$cdata_mode_p) {
225 $kind = KIND_DIRECTIVE;
227 if (!$ok_p && $eof_p) {
228 ($kind, $it, $readahead) = (KIND_UNKNOWN, $readahead, undef);
232 warn "Unrecognizable token found: $it\n", $lc_0
233 if $kind eq KIND_UNKNOWN;
234 return defined $it? (wantarray? ($kind, $it):
235 [$kind, $it]): undef;
241 if (!$cdata_mode_p) {
242 $it = next_token_internal($h);
243 if (defined $it && $it->[0] eq KIND_TAG) { # FIXME
244 ($cdata_mode_p, $cdata_close) = (1, "</$1\\s*>")
245 if $it->[1] =~ /^<(script|style|textarea)\b/i; #FIXME
246 push @$it, extract_attributes($it->[1], $lc_0); #FIXME
251 my $next = next_token_internal($h);
252 last if !defined $next;
253 if (defined $next && $next->[1] =~ /$cdata_close/i) { #FIXME
254 ($lc, $readahead) = ($lc_prev, $next->[1] . $readahead); #FIXME
257 last unless $cdata_mode_p;
258 $it .= $next->[1]; #FIXME
260 $it = [KIND_CDATA, $it]; #FIXME
261 $cdata_close = undef;
263 return defined $it? (wantarray? @$it: $it): undef;
266 ###############################################################################
268 sub debug_dump (*) { # for testing only
270 print "re_tag_compat is /$re_tag_compat/\n";
272 my $s = next_token $h;
273 last unless defined $s;
274 printf "%s\n", ('-' x 79);
275 my($kind, $t, $attr) = @$s; # FIXME
276 printf "%s:\n", $kind;
277 printf "%4dH%s\n", length($t),
278 join('', map {/[\0-\37]/? $_: "$_\b$_"} split(//, $t));
279 if ($kind eq KIND_TAG && %$attr) {
280 printf "Attributes:\n";
281 for my $a (keys %$attr) {
282 my($key, $val, $val_orig, $order) = @{$attr->{$a}};
283 printf "%s = %dH%s -- %s\n", $a, length $val,
284 join('', map {/[\0-\37]/? $_: "$_\b$_"} split(//, $val)),
291 ###############################################################################
295 $s =~ s/^(?:\s|\ $re_end_entity)+//os;
296 $s =~ s/(?:\s|\ $re_end_entity)+$//os;
300 ###############################################################################
302 sub text_extract (*) {
306 my $s = next_token $h;
307 last unless defined $s;
308 my($kind, $t, $attr) = @$s; # FIXME
309 if ($kind eq KIND_TEXT) {
311 $text{$t} = 1 if $t =~ /\S/s;
312 } elsif ($kind eq KIND_TAG && %$attr) {
313 # value [tag=input], meta
314 my $tag = lc($1) if $t =~ /^<(\S+)/s;
315 for my $a ('alt', 'content', 'title', 'value') {
317 next if $a eq 'content' && $tag ne 'meta';
318 next if $a eq 'value' && ($tag ne 'input'
319 || (ref $attr->{'type'} && $attr->{'type'}->[1] eq 'hidden')); # FIXME
320 my($key, $val, $val_orig, $order) = @{$attr->{$a}}; #FIXME
322 $text{$val} = 1 if $val =~ /\S/s;
327 # Emit all extracted strings.
328 # Don't emit pure whitespace, pure numbers, or TMPL_VAR's.
329 for my $t (keys %text) {
331 unless $t =~ /^(?:\s|\ $re_end_entity|$re_tmpl_var)*$/os || $t =~ /^\d+$/;
335 ###############################################################################
339 my $h = $exitcode? *STDERR: *STDOUT;
342 Extract strings from HTML file.
344 --debug-dump-only Do not extract strings; but display scanned tokens
345 -f, --file=FILE Extract from the specified FILE
346 --pedantic-warnings Issue warnings even for detected problems which
347 are likely to be harmless
348 --help Display this help and exit
353 ###############################################################################
355 sub usage_error (;$) {
356 print STDERR "$_[0]\n" if @_;
357 print STDERR "Try `$0 --help' for more information.\n";
361 ###############################################################################
364 'f|file=s' => \$input,
365 'debug-dump-only' => \$debug_dump_only_p,
366 'pedantic-warnings' => sub { $pedantic_p = 1 },
367 'help' => sub { usage(0) },
369 $pedantic_tag = $pedantic_p? '': ' (negligible)';
370 usage_error('Missing mandatory option -f') unless defined $input;
372 open(INPUT, "<$input") || die "$0: $input: $!\n";
373 if ($debug_dump_only_p) {
376 text_extract(*INPUT);
379 warn "This input will not work with Mozilla standards-compliant mode\n", undef
384 exit(-1) if $fatal_p;