1HTML::Parser(3) User Contributed Perl Documentation HTML::Parser(3)
2
6 HTML::Parser - HTML parser class
7
9 use strict;
10 use warnings;
11 use HTML::Parser ();
12 # Create parser object
14 my $p = HTML::Parser->new(
15 api_version => 3,
16 start_h => [\&start, "tagname, attr"],
17 end_h => [\&end, "tagname"],
18 marked_sections => 1,
19 );
20 # Parse document text chunk by chunk
22 $p->parse($chunk1);
23 $p->parse($chunk2);
24 # ...
26 # signal end of document
27 $p->eof;
28 # Parse directly from file
30 $p->parse_file("foo.html");
31 # or
33 open(my $fh, "<:utf8", "foo.html") || die;
34 $p->parse_file($fh);
35
37 Objects of the "HTML::Parser" class will recognize markup and separate
38 it from plain text (alias data content) in HTML documents. As
39 different kinds of markup and text are recognized, the corresponding
40 event handlers are invoked.
41 "HTML::Parser" is not a generic SGML parser. We have tried to make it
43 able to deal with the HTML that is actually "out there", and it
44 normally parses as closely as possible to the way the popular web
45 browsers do it instead of strictly following one of the many HTML
46 specifications from W3C. Where there is disagreement, there is often
47 an option that you can enable to get the official behaviour.
48 The document to be parsed may be supplied in arbitrary chunks. This
50 makes on-the-fly parsing as documents are received from the network
51 possible.
52 If event driven parsing does not feel right for your application, you
54 might want to use "HTML::PullParser". This is an "HTML::Parser"
55 subclass that allows a more conventional program structure.
56
58 The following method is used to construct a new "HTML::Parser" object:
59 $p = HTML::Parser->new( %options_and_handlers )
61 This class method creates a new "HTML::Parser" object and returns
62 it. Key/value argument pairs may be provided to assign event
63 handlers or initialize parser options. The handlers and parser
64 options can also be set or modified later by the method calls
65 described below.
66 If a top level key is in the form "<event>_h" (e.g., "text_h") then
68 it assigns a handler to that event, otherwise it initializes a
69 parser option. The event handler specification value must be an
70 array reference. Multiple handlers may also be assigned with the
71 'handlers => [%handlers]' option. See examples below.
72 If new() is called without any arguments, it will create a parser
74 that uses callback methods compatible with version 2 of
75 "HTML::Parser". See the section on "version 2 compatibility" below
76 for details.
77 The special constructor option 'api_version => 2' can be used to
79 initialize version 2 callbacks while still setting other options
80 and handlers. The 'api_version => 3' option can be used if you
81 don't want to set any options and don't want to fall back to v2
82 compatible mode.
83 Examples:
85 $p = HTML::Parser->new(
87 api_version => 3,
88 text_h => [ sub {...}, "dtext" ]
89 );
90 This creates a new parser object with a text event handler
92 subroutine that receives the original text with general entities
93 decoded.
94 $p = HTML::Parser->new(
96 api_version => 3,
97 start_h => [ 'my_start', "self,tokens" ]
98 );
99 This creates a new parser object with a start event handler method
101 that receives the $p and the tokens array.
102 $p = HTML::Parser->new(
104 api_version => 3,
105 handlers => {
106 text => [\@array, "event,text"],
107 comment => [\@array, "event,text"],
108 }
109 );
110 This creates a new parser object that stores the event type and the
112 original text in @array for text and comment events.
113 The following methods feed the HTML document to the "HTML::Parser"
115 object:
116 $p->parse( $string )
118 Parse $string as the next chunk of the HTML document. Handlers
119 invoked should not attempt to modify the $string in-place until
120 $p->parse returns.
121 If an invoked event handler aborts parsing by calling $p->eof, then
123 $p->parse() will return a FALSE value. Otherwise the return value
124 is a reference to the parser object ($p).
125 $p->parse( $code_ref )
127 If a code reference is passed as the argument to be parsed, then
128 the chunks to be parsed are obtained by invoking this function
129 repeatedly. Parsing continues until the function returns an empty
130 (or undefined) result. When this happens $p->eof is automatically
131 signaled.
132 Parsing will also abort if one of the event handlers calls $p->eof.
134 The effect of this is the same as:
136 while (1) {
138 my $chunk = &$code_ref();
139 if (!defined($chunk) || !length($chunk)) {
140 $p->eof;
141 return $p;
142 }
143 $p->parse($chunk) || return undef;
144 }
145 But it is more efficient as this loop runs internally in XS code.
147 $p->parse_file( $file )
149 Parse text directly from a file. The $file argument can be a
150 filename, an open file handle, or a reference to an open file
151 handle.
152 If $file contains a filename and the file can't be opened, then the
154 method returns an undefined value and $! tells why it failed.
155 Otherwise the return value is a reference to the parser object.
156 If a file handle is passed as the $file argument, then the file
158 will normally be read until EOF, but not closed.
159 If an invoked event handler aborts parsing by calling $p->eof, then
161 $p->parse_file() may not have read the entire file.
162 On systems with multi-byte line terminators, the values passed for
164 the offset and length argspecs may be too low if parse_file() is
165 called on a file handle that is not in binary mode.
166 If a filename is passed in, then parse_file() will open the file in
168 binary mode.
169 $p->eof
171 Signals the end of the HTML document. Calling the $p->eof method
172 outside a handler callback will flush any remaining buffered text
173 (which triggers the "text" event if there is any remaining text).
174 Calling $p->eof inside a handler will terminate parsing at that
176 point and cause $p->parse to return a FALSE value. This also
177 terminates parsing by $p->parse_file().
178 After $p->eof has been called, the parse() and parse_file() methods
180 can be invoked to feed new documents with the parser object.
181 The return value from eof() is a reference to the parser object.
183 Most parser options are controlled by boolean attributes. Each boolean
185 attribute is enabled by calling the corresponding method with a TRUE
186 argument and disabled with a FALSE argument. The attribute value is
187 left unchanged if no argument is given. The return value from each
188 method is the old attribute value.
189 Methods that can be used to get and/or set parser options are:
191 $p->attr_encoded
193 $p->attr_encoded( $bool )
194 By default, the "attr" and @attr argspecs will have general
195 entities for attribute values decoded. Enabling this attribute
196 leaves entities alone.
197 $p->backquote
199 $p->backquote( $bool )
200 By default, only ' and " are recognized as quote characters around
201 attribute values. MSIE also recognizes backquotes for some reason.
202 Enabling this attribute provides compatibility with this behaviour.
203 $p->boolean_attribute_value( $val )
205 This method sets the value reported for boolean attributes inside
206 HTML start tags. By default, the name of the attribute is also
207 used as its value. This affects the values reported for "tokens"
208 and "attr" argspecs.
209 $p->case_sensitive
211 $p->case_sensitive( $bool )
212 By default, tag names and attribute names are down-cased. Enabling
213 this attribute leaves them as found in the HTML source document.
214 $p->closing_plaintext
216 $p->closing_plaintext( $bool )
217 By default, "plaintext" element can never be closed. Everything up
218 to the end of the document is parsed in CDATA mode. This
219 historical behaviour is what at least MSIE does. Enabling this
220 attribute makes closing " </plaintext" > tag effective and the
221 parsing process will resume after seeing this tag. This emulates
222 early gecko-based browsers.
223 $p->empty_element_tags
225 $p->empty_element_tags( $bool )
226 By default, empty element tags are not recognized as such and the
227 "/" before ">" is just treated like a normal name character (unless
228 "strict_names" is enabled). Enabling this attribute make
229 "HTML::Parser" recognize these tags.
230 Empty element tags look like start tags, but end with the character
232 sequence "/>" instead of ">". When recognized by "HTML::Parser"
233 they cause an artificial end event in addition to the start event.
234 The "text" for the artificial end event will be empty and the
235 "tokenpos" array will be undefined even though the token array will
236 have one element containing the tag name.
237 $p->marked_sections
239 $p->marked_sections( $bool )
240 By default, section markings like <![CDATA[...]]> are treated like
241 ordinary text. When this attribute is enabled section markings are
242 honoured.
243 There are currently no events associated with the marked section
245 markup, but the text can be returned as "skipped_text".
246 $p->strict_comment
248 $p->strict_comment( $bool )
249 By default, comments are terminated by the first occurrence of
250 "-->". This is the behaviour of most popular browsers (like
251 Mozilla, Opera and MSIE), but it is not correct according to the
252 official HTML standard. Officially, you need an even number of
253 "--" tokens before the closing ">" is recognized and there may not
254 be anything but whitespace between an even and an odd "--".
255 The official behaviour is enabled by enabling this attribute.
257 Enabling of 'strict_comment' also disables recognizing these forms
259 as comments:
260 </ comment>
262 <! comment>
263 $p->strict_end
265 $p->strict_end( $bool )
266 By default, attributes and other junk are allowed to be present on
267 end tags in a manner that emulates MSIE's behaviour.
268 The official behaviour is enabled with this attribute. If enabled,
270 only whitespace is allowed between the tagname and the final ">".
271 $p->strict_names
273 $p->strict_names( $bool )
274 By default, almost anything is allowed in tag and attribute names.
275 This is the behaviour of most popular browsers and allows us to
276 parse some broken tags with invalid attribute values like:
277 <IMG SRC=newprevlstGr.gif ALT=[PREV LIST] BORDER=0>
279 By default, "LIST]" is parsed as a boolean attribute, not as part
281 of the ALT value as was clearly intended. This is also what
282 Mozilla sees.
283 The official behaviour is enabled by enabling this attribute. If
285 enabled, it will cause the tag above to be reported as text since
286 "LIST]" is not a legal attribute name.
287 $p->unbroken_text
289 $p->unbroken_text( $bool )
290 By default, blocks of text are given to the text handler as soon as
291 possible (but the parser takes care always to break text at a
292 boundary between whitespace and non-whitespace so single words and
293 entities can always be decoded safely). This might create breaks
294 that make it hard to do transformations on the text. When this
295 attribute is enabled, blocks of text are always reported in one
296 piece. This will delay the text event until the following (non-
297 text) event has been recognized by the parser.
298 Note that the "offset" argspec will give you the offset of the
300 first segment of text and "length" is the combined length of the
301 segments. Since there might be ignored tags in between, these
302 numbers can't be used to directly index in the original document
303 file.
304 $p->utf8_mode
306 $p->utf8_mode( $bool )
307 Enable this option when parsing raw undecoded UTF-8. This tells
308 the parser that the entities expanded for strings reported by
309 "attr", @attr and "dtext" should be expanded as decoded UTF-8 so
310 they end up compatible with the surrounding text.
311 If "utf8_mode" is enabled then it is an error to pass strings
313 containing characters with code above 255 to the parse() method,
314 and the parse() method will croak if you try.
315 Example: The Unicode character "\x{2665}" is "\xE2\x99\xA5" when
317 UTF-8 encoded. The character can also be represented by the entity
318 "♥" or "♥". If we feed the parser:
319 $p->parse("\xE2\x99\xA5♥");
321 then "dtext" will be reported as "\xE2\x99\xA5\x{2665}" without
323 "utf8_mode" enabled, but as "\xE2\x99\xA5\xE2\x99\xA5" when
324 enabled. The later string is what you want.
325 This option is only available with perl-5.8 or better.
327 $p->xml_mode
329 $p->xml_mode( $bool )
330 Enabling this attribute changes the parser to allow some XML
331 constructs. This enables the behaviour controlled by individually
332 by the "case_sensitive", "empty_element_tags", "strict_names" and
333 "xml_pic" attributes and also suppresses special treatment of
334 elements that are parsed as CDATA for HTML.
335 $p->xml_pic
337 $p->xml_pic( $bool )
338 By default, processing instructions are terminated by ">". When
339 this attribute is enabled, processing instructions are terminated
340 by "?>" instead.
341 As markup and text is recognized, handlers are invoked. The following
343 method is used to set up handlers for different events:
344 $p->handler( event => \&subroutine, $argspec )
346 $p->handler( event => $method_name, $argspec )
347 $p->handler( event => \@accum, $argspec )
348 $p->handler( event => "" );
349 $p->handler( event => undef );
350 $p->handler( event );
351 This method assigns a subroutine, method, or array to handle an
352 event.
353 Event is one of "text", "start", "end", "declaration", "comment",
355 "process", "start_document", "end_document" or "default".
356 The "\&subroutine" is a reference to a subroutine which is called
358 to handle the event.
359 The $method_name is the name of a method of $p which is called to
361 handle the event.
362 The @accum is an array that will hold the event information as sub-
364 arrays.
365 If the second argument is "", the event is ignored. If it is
367 undef, the default handler is invoked for the event.
368 The $argspec is a string that describes the information to be
370 reported for the event. Any requested information that does not
371 apply to a specific event is passed as "undef". If argspec is
372 omitted, then it is left unchanged.
373 The return value from $p->handler is the old callback routine or a
375 reference to the accumulator array.
376 Any return values from handler callback routines/methods are always
378 ignored. A handler callback can request parsing to be aborted by
379 invoking the $p->eof method. A handler callback is not allowed to
380 invoke the $p->parse() or $p->parse_file() method. An exception
381 will be raised if it tries.
382 Examples:
384 $p->handler(start => "start", 'self, attr, attrseq, text' );
386 This causes the "start" method of object $p to be called for
388 'start' events. The callback signature is "$p->start(\%attr,
389 \@attr_seq, $text)".
390 $p->handler(start => \&start, 'attr, attrseq, text' );
392 This causes subroutine start() to be called for 'start' events.
394 The callback signature is start(\%attr, \@attr_seq, $text).
395 $p->handler(start => \@accum, '"S", attr, attrseq, text' );
397 This causes 'start' event information to be saved in @accum. The
399 array elements will be ['S', \%attr, \@attr_seq, $text].
400 $p->handler(start => "");
402 This causes 'start' events to be ignored. It also suppresses
404 invocations of any default handler for start events. It is in most
405 cases equivalent to $p->handler(start => sub {}), but is more
406 efficient. It is different from the empty-sub-handler in that
407 "skipped_text" is not reset by it.
408 $p->handler(start => undef);
410 This causes no handler to be associated with start events. If
412 there is a default handler it will be invoked.
413 Filters based on tags can be set up to limit the number of events
415 reported. The main bottleneck during parsing is often the huge number
416 of callbacks made from the parser. Applying filters can improve
417 performance significantly.
418 The following methods control filters:
420 $p->ignore_elements( @tags )
422 Both the "start" event and the "end" event as well as any events
423 that would be reported in between are suppressed. The ignored
424 elements can contain nested occurrences of itself. Example:
425 $p->ignore_elements(qw(script style));
427 The "script" and "style" tags will always nest properly since their
429 content is parsed in CDATA mode. For most other tags
430 "ignore_elements" must be used with caution since HTML is often not
431 well formed.
432 $p->ignore_tags( @tags )
434 Any "start" and "end" events involving any of the tags given are
435 suppressed. To reset the filter (i.e. don't suppress any "start"
436 and "end" events), call "ignore_tags" without an argument.
437 $p->report_tags( @tags )
439 Any "start" and "end" events involving any of the tags not given
440 are suppressed. To reset the filter (i.e. report all "start" and
441 "end" events), call "report_tags" without an argument.
442 Internally, the system has two filter lists, one for "report_tags" and
444 one for "ignore_tags", and both filters are applied. This effectively
445 gives "ignore_tags" precedence over "report_tags".
446 Examples:
448 $p->ignore_tags(qw(style));
450 $p->report_tags(qw(script style));
451 results in only "script" events being reported.
453 Argspec
455 Argspec is a string containing a comma-separated list that describes
456 the information reported by the event. The following argspec
457 identifier names can be used:
458 "attr"
460 Attr causes a reference to a hash of attribute name/value pairs to
461 be passed.
462 Boolean attributes' values are either the value set by
464 $p->boolean_attribute_value, or the attribute name if no value has
465 been set by $p->boolean_attribute_value.
466 This passes undef except for "start" events.
468 Unless "xml_mode" or "case_sensitive" is enabled, the attribute
470 names are forced to lower case.
471 General entities are decoded in the attribute values and one layer
473 of matching quotes enclosing the attribute values is removed.
474 The Unicode character set is assumed for entity decoding.
476 @attr
478 Basically the same as "attr", but keys and values are passed as
479 individual arguments and the original sequence of the attributes is
480 kept. The parameters passed will be the same as the @attr
481 calculated here:
482 @attr = map { $_ => $attr->{$_} } @$attrseq;
484 assuming $attr and $attrseq here are the hash and array passed as
486 the result of "attr" and "attrseq" argspecs.
487 This passes no values for events besides "start".
489 "attrseq"
491 Attrseq causes a reference to an array of attribute names to be
492 passed. This can be useful if you want to walk the "attr" hash in
493 the original sequence.
494 This passes undef except for "start" events.
496 Unless "xml_mode" or "case_sensitive" is enabled, the attribute
498 names are forced to lower case.
499 "column"
501 Column causes the column number of the start of the event to be
502 passed. The first column on a line is 0.
503 "dtext"
505 Dtext causes the decoded text to be passed. General entities are
506 automatically decoded unless the event was inside a CDATA section
507 or was between literal start and end tags ("script", "style",
508 "xmp", "iframe", "title", "textarea" and "plaintext").
509 The Unicode character set is assumed for entity decoding. With
511 Perl version 5.6 or earlier only the Latin-1 range is supported,
512 and entities for characters outside the range 0..255 are left
513 unchanged.
514 This passes undef except for "text" events.
516 "event"
518 Event causes the event name to be passed.
519 The event name is one of "text", "start", "end", "declaration",
521 "comment", "process", "start_document" or "end_document".
522 "is_cdata"
524 Is_cdata causes a TRUE value to be passed if the event is inside a
525 CDATA section or between literal start and end tags ("script",
526 "style", "xmp", "iframe", "title", "textarea" and "plaintext").
527 if the flag is FALSE for a text event, then you should normally
529 either use "dtext" or decode the entities yourself before the text
530 is processed further.
531 "length"
533 Length causes the number of bytes of the source text of the event
534 to be passed.
535 "line"
537 Line causes the line number of the start of the event to be passed.
538 The first line in the document is 1. Line counting doesn't start
539 until at least one handler requests this value to be reported.
540 "offset"
542 Offset causes the byte position in the HTML document of the start
543 of the event to be passed. The first byte in the document has
544 offset 0.
545 "offset_end"
547 Offset_end causes the byte position in the HTML document of the end
548 of the event to be passed. This is the same as "offset" +
549 "length".
550 "self"
552 Self causes the current object to be passed to the handler. If the
553 handler is a method, this must be the first element in the argspec.
554 An alternative to passing self as an argspec is to register
556 closures that capture $self by themselves as handlers.
557 Unfortunately this creates circular references which prevent the
558 HTML::Parser object from being garbage collected. Using the "self"
559 argspec avoids this problem.
560 "skipped_text"
562 Skipped_text returns the concatenated text of all the events that
563 have been skipped since the last time an event was reported.
564 Events might be skipped because no handler is registered for them
565 or because some filter applies. Skipped text also includes marked
566 section markup, since there are no events that can catch it.
567 If an ""-handler is registered for an event, then the text for this
569 event is not included in "skipped_text". Skipped text both before
570 and after the ""-event is included in the next reported
571 "skipped_text".
572 "tag"
574 Same as "tagname", but prefixed with "/" if it belongs to an "end"
575 event and "!" for a declaration. The "tag" does not have any
576 prefix for "start" events, and is in this case identical to
577 "tagname".
578 "tagname"
580 This is the element name (or generic identifier in SGML jargon) for
581 start and end tags. Since HTML is case insensitive, this name is
582 forced to lower case to ease string matching.
583 Since XML is case sensitive, the tagname case is not changed when
585 "xml_mode" is enabled. The same happens if the "case_sensitive"
586 attribute is set.
587 The declaration type of declaration elements is also passed as a
589 tagname, even if that is a bit strange. In fact, in the current
590 implementation tagname is identical to "token0" except that the
591 name may be forced to lower case.
592 "token0"
594 Token0 causes the original text of the first token string to be
595 passed. This should always be the same as $tokens->[0].
596 For "declaration" events, this is the declaration type.
598 For "start" and "end" events, this is the tag name.
600 For "process" and non-strict "comment" events, this is everything
602 inside the tag.
603 This passes undef if there are no tokens in the event.
605 "tokenpos"
607 Tokenpos causes a reference to an array of token positions to be
608 passed. For each string that appears in "tokens", this array
609 contains two numbers. The first number is the offset of the start
610 of the token in the original "text" and the second number is the
611 length of the token.
612 Boolean attributes in a "start" event will have (0,0) for the
614 attribute value offset and length.
615 This passes undef if there are no tokens in the event (e.g.,
617 "text") and for artificial "end" events triggered by empty element
618 tags.
619 If you are using these offsets and lengths to modify "text", you
621 should either work from right to left, or be very careful to
622 calculate the changes to the offsets.
623 "tokens"
625 Tokens causes a reference to an array of token strings to be
626 passed. The strings are exactly as they were found in the original
627 text, no decoding or case changes are applied.
628 For "declaration" events, the array contains each word, comment,
630 and delimited string starting with the declaration type.
631 For "comment" events, this contains each sub-comment. If
633 $p->strict_comments is disabled, there will be only one sub-
634 comment.
635 For "start" events, this contains the original tag name followed by
637 the attribute name/value pairs. The values of boolean attributes
638 will be either the value set by $p->boolean_attribute_value, or the
639 attribute name if no value has been set by
640 $p->boolean_attribute_value.
641 For "end" events, this contains the original tag name (always one
643 token).
644 For "process" events, this contains the process instructions
646 (always one token).
647 This passes "undef" for "text" events.
649 "text"
651 Text causes the source text (including markup element delimiters)
652 to be passed.
653 "undef"
655 Pass an undefined value. Useful as padding where the same handler
656 routine is registered for multiple events.
657 '...'
659 A literal string of 0 to 255 characters enclosed in single (') or
660 double (") quotes is passed as entered.
661 The whole argspec string can be wrapped up in '@{...}' to signal that
663 the resulting event array should be flattened. This only makes a
664 difference if an array reference is used as the handler target.
665 Consider this example:
666 $p->handler(text => [], 'text');
668 $p->handler(text => [], '@{text}']);
669 With two text events; "foo", "bar"; then the first example will end up
671 with [["foo"], ["bar"]] and the second with ["foo", "bar"] in the
672 handler target array.
673 Events
675 Handlers for the following events can be registered:
676 "comment"
678 This event is triggered when a markup comment is recognized.
679 Example:
681 <!-- This is a comment -- -- So is this -->
683 "declaration"
685 This event is triggered when a markup declaration is recognized.
686 For typical HTML documents, the only declaration you are likely to
688 find is <!DOCTYPE ...>.
689 Example:
691 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
693 "http://www.w3.org/TR/html4/strict.dtd">
694 DTDs inside <!DOCTYPE ...> will confuse HTML::Parser.
696 "default"
698 This event is triggered for events that do not have a specific
699 handler. You can set up a handler for this event to catch stuff
700 you did not want to catch explicitly.
701 "end"
703 This event is triggered when an end tag is recognized.
704 Example:
706 </A>
708 "end_document"
710 This event is triggered when $p->eof is called and after any
711 remaining text is flushed. There is no document text associated
712 with this event.
713 "process"
715 This event is triggered when a processing instructions markup is
716 recognized.
717 The format and content of processing instructions are system and
719 application dependent.
720 Examples:
722 <? HTML processing instructions >
724 <? XML processing instructions ?>
725 "start"
727 This event is triggered when a start tag is recognized.
728 Example:
730 <A HREF="http://www.perl.com/">
732 "start_document"
734 This event is triggered before any other events for a new document.
735 A handler for it can be used to initialize stuff. There is no
736 document text associated with this event.
737 "text"
739 This event is triggered when plain text (characters) is recognized.
740 The text may contain multiple lines. A sequence of text may be
741 broken between several text events unless $p->unbroken_text is
742 enabled.
743 The parser will make sure that it does not break a word or a
745 sequence of whitespace between two text events.
746 Unicode
748 "HTML::Parser" can parse Unicode strings when running under perl-5.8 or
749 better. If Unicode is passed to $p->parse() then chunks of Unicode
750 will be reported to the handlers. The offset and length argspecs will
751 also report their position in terms of characters.
752 It is safe to parse raw undecoded UTF-8 if you either avoid decoding
754 entities and make sure to not use argspecs that do, or enable the
755 "utf8_mode" for the parser. Parsing of undecoded UTF-8 might be useful
756 when parsing from a file where you need the reported offsets and
757 lengths to match the byte offsets in the file.
758 If a filename is passed to $p->parse_file() then the file will be read
760 in binary mode. This will be fine if the file contains only ASCII or
761 Latin-1 characters. If the file contains UTF-8 encoded text then care
762 must be taken when decoding entities as described in the previous
763 paragraph, but better is to open the file with the UTF-8 layer so that
764 it is decoded properly:
765 open(my $fh, "<:utf8", "index.html") || die "...: $!";
767 $p->parse_file($fh);
768 If the file contains text encoded in a charset besides ASCII, Latin-1
770 or UTF-8 then decoding will always be needed.
771
773 When an "HTML::Parser" object is constructed with no arguments, a set
774 of handlers is automatically provided that is compatible with the old
775 HTML::Parser version 2 callback methods.
776 This is equivalent to the following method calls:
778 $p->handler(start => "start", "self, tagname, attr, attrseq, text");
780 $p->handler(end => "end", "self, tagname, text");
781 $p->handler(text => "text", "self, text, is_cdata");
782 $p->handler(process => "process", "self, token0, text");
783 $p->handler(
784 comment => sub {
785 my ($self, $tokens) = @_;
786 for (@$tokens) { $self->comment($_); }
787 },
788 "self, tokens"
789 );
790 $p->handler(
791 declaration => sub {
792 my $self = shift;
793 $self->declaration(substr($_[0], 2, -1));
794 },
795 "self, text"
796 );
797 Setting up these handlers can also be requested with the "api_version
799 => 2" constructor option.
800
802 The "HTML::Parser" class is able to be subclassed. Parser objects are
803 plain hashes and "HTML::Parser" reserves only hash keys that start with
804 "_hparser". The parser state can be set up by invoking the init()
805 method, which takes the same arguments as new().
806
808 The first simple example shows how you might strip out comments from an
809 HTML document. We achieve this by setting up a comment handler that
810 does nothing and a default handler that will print out anything else:
811 use HTML::Parser ();
813 HTML::Parser->new(
814 default_h => [sub { print shift }, 'text'],
815 comment_h => [""],
816 )->parse_file(shift || die)
817 || die $!;
818 An alternative implementation is:
820 use HTML::Parser ();
822 HTML::Parser->new(
823 end_document_h => [sub { print shift }, 'skipped_text'],
824 comment_h => [""],
825 )->parse_file(shift || die)
826 || die $!;
827 This will in most cases be much more efficient since only a single
829 callback will be made.
830 The next example prints out the text that is inside the <title> element
832 of an HTML document. Here we start by setting up a start handler.
833 When it sees the title start tag it enables a text handler that prints
834 any text found and an end handler that will terminate parsing as soon
835 as the title end tag is seen:
836 use HTML::Parser ();
838 sub start_handler {
840 return if shift ne "title";
841 my $self = shift;
842 $self->handler(text => sub { print shift }, "dtext");
843 $self->handler(
844 end => sub {
845 shift->eof if shift eq "title";
846 },
847 "tagname,self"
848 );
849 }
850 my $p = HTML::Parser->new(api_version => 3);
852 $p->handler(start => \&start_handler, "tagname,self");
853 $p->parse_file(shift || die) || die $!;
854 print "\n";
855 More examples are found in the eg/ directory of the "HTML-Parser"
857 distribution: the program "hrefsub" shows how you can edit all links
858 found in a document; the program "htextsub" shows how to edit the text
859 only; the program "hstrip" shows how you can strip out certain
860 tags/elements and/or attributes; and the program "htext" show how to
861 obtain the plain text, but not any script/style content.
862 You can browse the eg/ directory online from the [Browse] link on the
864 http://search.cpan.org/~gaas/HTML-Parser/ page.
865
867 The <style> and <script> sections do not end with the first "</", but
868 need the complete corresponding end tag. The standard behaviour is not
869 really practical.
870 When the strict_comment option is enabled, we still recognize comments
872 where there is something other than whitespace between even and odd
873 "--" markers.
874 Once $p->boolean_attribute_value has been set, there is no way to
876 restore the default behaviour.
877 There is currently no way to get both quote characters into the same
879 literal argspec.
880 Empty tags, e.g. "<>" and "</>", are not recognized. SGML allows them
882 to repeat the previous start tag or close the previous start tag
883 respectively.
884 NET tags, e.g. "code/.../" are not recognized. This is SGML shorthand
886 for "<code>...</code>".
887 Incomplete start or end tags, e.g. "<tt<b>...</b</tt>" are not
889 recognized.
890
892 The following messages may be produced by HTML::Parser. The notation
893 in this listing is the same as used in perldiag:
894 Not a reference to a hash
896 (F) The object blessed into or subclassed from HTML::Parser is not
897 a hash as required by the HTML::Parser methods.
898 Bad signature in parser state object at %p
900 (F) The _hparser_xs_state element does not refer to a valid state
901 structure. Something must have changed the internal value stored
902 in this hash element, or the memory has been overwritten.
903 _hparser_xs_state element is not a reference
905 (F) The _hparser_xs_state element has been destroyed.
906 Can't find '_hparser_xs_state' element in HTML::Parser hash
908 (F) The _hparser_xs_state element is missing from the parser hash.
909 It was either deleted, or not created when the object was created.
910 API version %s not supported by HTML::Parser %s
912 (F) The constructor option 'api_version' with an argument greater
913 than or equal to 4 is reserved for future extensions.
914 Bad constructor option '%s'
916 (F) An unknown constructor option key was passed to the new() or
917 init() methods.
918 Parse loop not allowed
920 (F) A handler invoked the parse() or parse_file() method. This is
921 not permitted.
922 marked sections not supported
924 (F) The $p->marked_sections() method was invoked in a HTML::Parser
925 module that was compiled without support for marked sections.
926 Unknown boolean attribute (%d)
928 (F) Something is wrong with the internal logic that set up aliases
929 for boolean attributes.
930 Only code or array references allowed as handler
932 (F) The second argument for $p->handler must be either a subroutine
933 reference, then name of a subroutine or method, or a reference to
934 an array.
935 No handler for %s events
937 (F) The first argument to $p->handler must be a valid event name;
938 i.e. one of "start", "end", "text", "process", "declaration" or
939 "comment".
940 Unrecognized identifier %s in argspec
942 (F) The identifier is not a known argspec name. Use one of the
943 names mentioned in the argspec section above.
944 Literal string is longer than 255 chars in argspec
946 (F) The current implementation limits the length of literals in an
947 argspec to 255 characters. Make the literal shorter.
948 Backslash reserved for literal string in argspec
950 (F) The backslash character "\" is not allowed in argspec literals.
951 It is reserved to permit quoting inside a literal in a later
952 version.
953 Unterminated literal string in argspec
955 (F) The terminating quote character for a literal was not found.
956 Bad argspec (%s)
958 (F) Only identifier names, literals, spaces and commas are allowed
959 in argspecs.
960 Missing comma separator in argspec
962 (F) Identifiers in an argspec must be separated with ",".
963 Parsing of undecoded UTF-8 will give garbage when decoding entities
965 (W) The first chunk parsed appears to contain undecoded UTF-8 and
966 one or more argspecs that decode entities are used for the callback
967 handlers.
968 The result of decoding will be a mix of encoded and decoded
970 characters for any entities that expand to characters with code
971 above 127. This is not a good thing.
972 The recommended solution is to apply Encode::decode_utf8() on the
974 data before feeding it to the $p->parse(). For $p->parse_file()
975 pass a file that has been opened in ":utf8" mode.
976 The alternative solution is to enable the "utf8_mode" and not
978 decode before passing strings to $p->parse(). The parser can
979 process raw undecoded UTF-8 sanely if the "utf8_mode" is enabled,
980 or if the "attr", @attr or "dtext" argspecs are avoided.
981 Parsing string decoded with wrong endian selection
983 (W) The first character in the document is U+FFFE. This is not a
984 legal Unicode character but a byte swapped "BOM". The result of
985 parsing will likely be garbage.
986 Parsing of undecoded UTF-32
988 (W) The parser found the Unicode UTF-32 "BOM" signature at the
989 start of the document. The result of parsing will likely be
990 garbage.
991 Parsing of undecoded UTF-16
993 (W) The parser found the Unicode UTF-16 "BOM" signature at the
994 start of the document. The result of parsing will likely be
995 garbage.
996
998 HTML::Entities, HTML::PullParser, HTML::TokeParser, HTML::HeadParser,
999 HTML::LinkExtor, HTML::Form
1000 HTML::TreeBuilder (part of the HTML-Tree distribution)
1002 <http://www.w3.org/TR/html4/>
1004 More information about marked sections and processing instructions may
1006 be found at <http://www.is-thought.co.uk/book/sgml-8.htm>.
1007
1009 Copyright 1996-2016 Gisle Aas. All rights reserved.
1010 Copyright 1999-2000 Michael A. Chase. All rights reserved.
1011 This library is free software; you can redistribute it and/or modify it
1013 under the same terms as Perl itself.
1014perl v5.34.1 2022-03-31 HTML::Parser(3)