1Parser(3) User Contributed Perl Documentation Parser(3)
2
6 HTML::Parser - HTML parser class
7
9 use HTML::Parser ();
10 # Create parser object
12 $p = HTML::Parser->new( api_version => 3,
13 start_h => [\&start, "tagname, attr"],
14 end_h => [\&end, "tagname"],
15 marked_sections => 1,
16 );
17 # Parse document text chunk by chunk
19 $p->parse($chunk1);
20 $p->parse($chunk2);
21 #...
22 $p->eof; # signal end of document
23 # Parse directly from file
25 $p->parse_file("foo.html");
26 # or
27 open(my $fh, "<:utf8", "foo.html") ⎪⎪ die;
28 $p->parse_file($fh);
29
31 Objects of the "HTML::Parser" class will recognize markup and separate
32 it from plain text (alias data content) in HTML documents. As differ‐
33 ent kinds of markup and text are recognized, the corresponding event
34 handlers are invoked.
35 "HTML::Parser" is not a generic SGML parser. We have tried to make it
37 able to deal with the HTML that is actually "out there", and it nor‐
38 mally parses as closely as possible to the way the popular web browsers
39 do it instead of strictly following one of the many HTML specifications
40 from W3C. Where there is disagreement, there is often an option that
41 you can enable to get the official behaviour.
42 The document to be parsed may be supplied in arbitrary chunks. This
44 makes on-the-fly parsing as documents are received from the network
45 possible.
46 If event driven parsing does not feel right for your application, you
48 might want to use "HTML::PullParser". This is an "HTML::Parser" sub‐
49 class that allows a more conventional program structure.
50
52 The following method is used to construct a new "HTML::Parser" object:
53 $p = HTML::Parser->new( %options_and_handlers )
55 This class method creates a new "HTML::Parser" object and returns
56 it. Key/value argument pairs may be provided to assign event han‐
57 dlers or initialize parser options. The handlers and parser
58 options can also be set or modified later by the method calls
59 described below.
60 If a top level key is in the form "<event>_h" (e.g., "text_h") then
62 it assigns a handler to that event, otherwise it initializes a
63 parser option. The event handler specification value must be an
64 array reference. Multiple handlers may also be assigned with the
65 'handlers => [%handlers]' option. See examples below.
66 If new() is called without any arguments, it will create a parser
68 that uses callback methods compatible with version 2 of
69 "HTML::Parser". See the section on "version 2 compatibility" below
70 for details.
71 The special constructor option 'api_version => 2' can be used to
73 initialize version 2 callbacks while still setting other options
74 and handlers. The 'api_version => 3' option can be used if you
75 don't want to set any options and don't want to fall back to v2
76 compatible mode.
77 Examples:
79 $p = HTML::Parser->new(api_version => 3,
81 text_h => [ sub {...}, "dtext" ]);
82 This creates a new parser object with a text event handler subrou‐
84 tine that receives the original text with general entities decoded.
85 $p = HTML::Parser->new(api_version => 3,
87 start_h => [ 'my_start', "self,tokens" ]);
88 This creates a new parser object with a start event handler method
90 that receives the $p and the tokens array.
91 $p = HTML::Parser->new(api_version => 3,
93 handlers => { text => [\@array, "event,text"],
94 comment => [\@array, "event,text"],
95 });
96 This creates a new parser object that stores the event type and the
98 original text in @array for text and comment events.
99 The following methods feed the HTML document to the "HTML::Parser"
101 object:
102 $p->parse( $string )
104 Parse $string as the next chunk of the HTML document. The return
105 value is normally a reference to the parser object (i.e. $p). Han‐
106 dlers invoked should not attempt to modify the $string in-place
107 until $p->parse returns.
108 If an invoked event handler aborts parsing by calling $p->eof, then
110 $p->parse() will return a FALSE value.
111 $p->parse( $code_ref )
113 If a code reference is passed as the argument to be parsed, then
114 the chunks to be parsed are obtained by invoking this function
115 repeatedly. Parsing continues until the function returns an empty
116 (or undefined) result. When this happens $p->eof is automatically
117 signaled.
118 Parsing will also abort if one of the event handlers calls $p->eof.
120 The effect of this is the same as:
122 while (1) {
124 my $chunk = &$code_ref();
125 if (!defined($chunk) ⎪⎪ !length($chunk)) {
126 $p->eof;
127 return $p;
128 }
129 $p->parse($chunk) ⎪⎪ return undef;
130 }
131 But it is more efficient as this loop runs internally in XS code.
133 $p->parse_file( $file )
135 Parse text directly from a file. The $file argument can be a file‐
136 name, an open file handle, or a reference to an open file handle.
137 If $file contains a filename and the file can't be opened, then the
139 method returns an undefined value and $! tells why it failed. Oth‐
140 erwise the return value is a reference to the parser object.
141 If a file handle is passed as the $file argument, then the file
143 will normally be read until EOF, but not closed.
144 If an invoked event handler aborts parsing by calling $p->eof, then
146 $p->parse_file() may not have read the entire file.
147 On systems with multi-byte line terminators, the values passed for
149 the offset and length argspecs may be too low if parse_file() is
150 called on a file handle that is not in binary mode.
151 If a filename is passed in, then parse_file() will open the file in
153 binary mode.
154 $p->eof
156 Signals the end of the HTML document. Calling the $p->eof method
157 outside a handler callback will flush any remaining buffered text
158 (which triggers the "text" event if there is any remaining text).
159 Calling $p->eof inside a handler will terminate parsing at that
161 point and cause $p->parse to return a FALSE value. This also ter‐
162 minates parsing by $p->parse_file().
163 After $p->eof has been called, the parse() and parse_file() methods
165 can be invoked to feed new documents with the parser object.
166 The return value from eof() is a reference to the parser object.
168 Most parser options are controlled by boolean attributes. Each boolean
170 attribute is enabled by calling the corresponding method with a TRUE
171 argument and disabled with a FALSE argument. The attribute value is
172 left unchanged if no argument is given. The return value from each
173 method is the old attribute value.
174 Methods that can be used to get and/or set parser options are:
176 $p->attr_encoded
178 $p->attr_encoded( $bool )
179 By default, the "attr" and @attr argspecs will have general enti‐
180 ties for attribute values decoded. Enabling this attribute leaves
181 entities alone.
182 $p->boolean_attribute_value( $val )
184 This method sets the value reported for boolean attributes inside
185 HTML start tags. By default, the name of the attribute is also
186 used as its value. This affects the values reported for "tokens"
187 and "attr" argspecs.
188 $p->case_sensitive
190 $p->case_sensitive( $bool )
191 By default, tagnames and attribute names are down-cased. Enabling
192 this attribute leaves them as found in the HTML source document.
193 $p->closing_plaintext
195 $p->closing_plaintext( $bool )
196 By default, "plaintext" element can never be closed. Everything up
197 to the end of the document is parsed in CDATA mode. This histori‐
198 cal behaviour is what at least MSIE does. Enabling this attribute
199 makes closing "</plaintext>" tag effective and the parsing process
200 will resume after seeing this tag. This emulates gecko-based
201 browsers.
202 $p->empty_element_tags
204 $p->empty_element_tags( $bool )
205 By default, empty element tags are not recognized as such and the
206 "/" before ">" is just treated like a normal name character (unless
207 "strict_names" is enabled). Enabling this attribute make
208 "HTML::Parser" recognize these tags.
209 Empty element tags look like start tags, but end with the character
211 sequence "/>" instead of ">". When recognized by "HTML::Parser"
212 they cause an artificial end event in addition to the start event.
213 The "text" for the artificial end event will be empty and the
214 "tokenpos" array will be undefined even though the the token array
215 will have one element containing the tag name.
216 $p->marked_sections
218 $p->marked_sections( $bool )
219 By default, section markings like <![CDATA[...]]> are treated like
220 ordinary text. When this attribute is enabled section markings are
221 honoured.
222 There are currently no events associated with the marked section
224 markup, but the text can be returned as "skipped_text".
225 $p->strict_comment
227 $p->strict_comment( $bool )
228 By default, comments are terminated by the first occurrence of
229 "-->". This is the behaviour of most popular browsers (like
230 Mozilla, Opera and MSIE), but it is not correct according to the
231 official HTML standard. Officially, you need an even number of
232 "--" tokens before the closing ">" is recognized and there may not
233 be anything but whitespace between an even and an odd "--".
234 The official behaviour is enabled by enabling this attribute.
236 Enabling of 'strict_comment' also disables recognizing these forms
238 as comments:
239 </ comment>
241 <! comment>
242 $p->strict_end
244 $p->strict_end( $bool )
245 By default, attributes and other junk are allowed to be present on
246 end tags in a manner that emulates MSIE's behaviour.
247 The official behaviour is enabled with this attribute. If enabled,
249 only whitespace is allowed between the tagname and the final ">".
250 $p->strict_names
252 $p->strict_names( $bool )
253 By default, almost anything is allowed in tag and attribute names.
254 This is the behaviour of most popular browsers and allows us to
255 parse some broken tags with invalid attribute values like:
256 <IMG SRC=newprevlstGr.gif ALT=[PREV LIST] BORDER=0>
258 By default, "LIST]" is parsed as a boolean attribute, not as part
260 of the ALT value as was clearly intended. This is also what
261 Mozilla sees.
262 The official behaviour is enabled by enabling this attribute. If
264 enabled, it will cause the tag above to be reported as text since
265 "LIST]" is not a legal attribute name.
266 $p->unbroken_text
268 $p->unbroken_text( $bool )
269 By default, blocks of text are given to the text handler as soon as
270 possible (but the parser takes care always to break text at a
271 boundary between whitespace and non-whitespace so single words and
272 entities can always be decoded safely). This might create breaks
273 that make it hard to do transformations on the text. When this
274 attribute is enabled, blocks of text are always reported in one
275 piece. This will delay the text event until the following
276 (non-text) event has been recognized by the parser.
277 Note that the "offset" argspec will give you the offset of the
279 first segment of text and "length" is the combined length of the
280 segments. Since there might be ignored tags in between, these num‐
281 bers can't be used to directly index in the original document file.
282 $p->utf8_mode
284 $p->utf8_mode( $bool )
285 Enable this option when parsing raw undecoded UTF-8. This tells
286 the parser that the entities expanded for strings reported by
287 "attr", @attr and "dtext" should be expanded as decoded UTF-8 so
288 they end up compatible with the surrounding text.
289 If "utf8_mode" is enabled then it is an error to pass strings con‐
291 taining characters with code above 255 to the parse() method, and
292 the parse() method will croak if you try.
293 Example: The Unicode character "\x{2665}" is "\xE2\x99\xA5" when
295 UTF-8 encoded. The character can also be represented by the entity
296 "♥" or "♥". If we feed the parser:
297 $p->parse("\xE2\x99\xA5♥");
299 then "dtext" will be reported as "\xE2\x99\xA5\x{2665}" without
301 "utf8_mode" enabled, but as "\xE2\x99\xA5\xE2\x99\xA5" when
302 enabled. The later string is what you want.
303 This option is only available with perl-5.8 or better.
305 $p->xml_mode
307 $p->xml_mode( $bool )
308 Enabling this attribute changes the parser to allow some XML con‐
309 structs. This enables the behaviour controlled by individually by
310 the "case_sensitive", "empty_element_tags", "strict_names" and
311 "xml_pic" attributes and also suppresses special treatment of ele‐
312 ments that are parsed as CDATA for HTML.
313 $p->xml_pic
315 $p->xml_pic( $bool )
316 By default, processing instructions are terminated by ">". When
317 this attribute is enabled, processing instructions are terminated
318 by "?>" instead.
319 As markup and text is recognized, handlers are invoked. The following
321 method is used to set up handlers for different events:
322 $p->handler( event => \&subroutine, $argspec )
324 $p->handler( event => $method_name, $argspec )
325 $p->handler( event => \@accum, $argspec )
326 $p->handler( event => "" );
327 $p->handler( event => undef );
328 $p->handler( event );
329 This method assigns a subroutine, method, or array to handle an
330 event.
331 Event is one of "text", "start", "end", "declaration", "comment",
333 "process", "start_document", "end_document" or "default".
334 The "\&subroutine" is a reference to a subroutine which is called
336 to handle the event.
337 The $method_name is the name of a method of $p which is called to
339 handle the event.
340 The @accum is an array that will hold the event information as
342 sub-arrays.
343 If the second argument is "", the event is ignored. If it is
345 undef, the default handler is invoked for the event.
346 The $argspec is a string that describes the information to be
348 reported for the event. Any requested information that does not
349 apply to a specific event is passed as "undef". If argspec is
350 omitted, then it is left unchanged.
351 The return value from $p->handler is the old callback routine or a
353 reference to the accumulator array.
354 Any return values from handler callback routines/methods are always
356 ignored. A handler callback can request parsing to be aborted by
357 invoking the $p->eof method. A handler callback is not allowed to
358 invoke the $p->parse() or $p->parse_file() method. An exception
359 will be raised if it tries.
360 Examples:
362 $p->handler(start => "start", 'self, attr, attrseq, text' );
364 This causes the "start" method of object $p to be called for
366 'start' events. The callback signature is $p->start(\%attr,
367 \@attr_seq, $text).
368 $p->handler(start => \&start, 'attr, attrseq, text' );
370 This causes subroutine start() to be called for 'start' events.
372 The callback signature is start(\%attr, \@attr_seq, $text).
373 $p->handler(start => \@accum, '"S", attr, attrseq, text' );
375 This causes 'start' event information to be saved in @accum. The
377 array elements will be ['S', \%attr, \@attr_seq, $text].
378 $p->handler(start => "");
380 This causes 'start' events to be ignored. It also suppresses invo‐
382 cations of any default handler for start events. It is in most
383 cases equivalent to $p->handler(start => sub {}), but is more effi‐
384 cient. It is different from the empty-sub-handler in that
385 "skipped_text" is not reset by it.
386 $p->handler(start => undef);
388 This causes no handler to be associated with start events. If
390 there is a default handler it will be invoked.
391 Filters based on tags can be set up to limit the number of events
393 reported. The main bottleneck during parsing is often the huge number
394 of callbacks made from the parser. Applying filters can improve per‐
395 formance significantly.
396 The following methods control filters:
398 $p->ignore_elements( @tags )
400 Both the "start" event and the "end" event as well as any events
401 that would be reported in between are suppressed. The ignored ele‐
402 ments can contain nested occurrences of itself. Example:
403 $p->ignore_elements(qw(script style));
405 The "script" and "style" tags will always nest properly since their
407 content is parsed in CDATA mode. For most other tags "ignore_ele‐
408 ments" must be used with caution since HTML is often not well
409 formed.
410 $p->ignore_tags( @tags )
412 Any "start" and "end" events involving any of the tags given are
413 suppressed. To reset the filter (i.e. don't suppress any "start"
414 and "end" events), call "ignore_tags" without an argument.
415 $p->report_tags( @tags )
417 Any "start" and "end" events involving any of the tags not given
418 are suppressed. To reset the filter (i.e. report all "start" and
419 "end" events), call "report_tags" without an argument.
420 Internally, the system has two filter lists, one for "report_tags" and
422 one for "ignore_tags", and both filters are applied. This effectively
423 gives "ignore_tags" precedence over "report_tags".
424 Examples:
426 $p->ignore_tags(qw(style));
428 $p->report_tags(qw(script style));
429 results in only "script" events being reported.
431 Argspec
433 Argspec is a string containing a comma-separated list that describes
435 the information reported by the event. The following argspec identi‐
436 fier names can be used:
437 "attr"
439 Attr causes a reference to a hash of attribute name/value pairs to
440 be passed.
441 Boolean attributes' values are either the value set by $p->bool‐
443 ean_attribute_value, or the attribute name if no value has been set
444 by $p->boolean_attribute_value.
445 This passes undef except for "start" events.
447 Unless "xml_mode" or "case_sensitive" is enabled, the attribute
449 names are forced to lower case.
450 General entities are decoded in the attribute values and one layer
452 of matching quotes enclosing the attribute values is removed.
453 The Unicode character set is assumed for entity decoding. With
455 Perl version 5.6 or earlier only the Latin-1 range is supported,
456 and entities for characters outside the range 0..255 are left
457 unchanged.
458 @attr
460 Basically the same as "attr", but keys and values are passed as
461 individual arguments and the original sequence of the attributes is
462 kept. The parameters passed will be the same as the @attr calcu‐
463 lated here:
464 @attr = map { $_ => $attr->{$_} } @$attrseq;
466 assuming $attr and $attrseq here are the hash and array passed as
468 the result of "attr" and "attrseq" argspecs.
469 This passes no values for events besides "start".
471 "attrseq"
473 Attrseq causes a reference to an array of attribute names to be
474 passed. This can be useful if you want to walk the "attr" hash in
475 the original sequence.
476 This passes undef except for "start" events.
478 Unless "xml_mode" or "case_sensitive" is enabled, the attribute
480 names are forced to lower case.
481 "column"
483 Column causes the column number of the start of the event to be
484 passed. The first column on a line is 0.
485 "dtext"
487 Dtext causes the decoded text to be passed. General entities are
488 automatically decoded unless the event was inside a CDATA section
489 or was between literal start and end tags ("script", "style",
490 "xmp", and "plaintext").
491 The Unicode character set is assumed for entity decoding. With
493 Perl version 5.6 or earlier only the Latin-1 range is supported,
494 and entities for characters outside the range 0..255 are left
495 unchanged.
496 This passes undef except for "text" events.
498 "event"
500 Event causes the event name to be passed.
501 The event name is one of "text", "start", "end", "declaration",
503 "comment", "process", "start_document" or "end_document".
504 "is_cdata"
506 Is_cdata causes a TRUE value to be passed if the event is inside a
507 CDATA section or between literal start and end tags ("script",
508 "style", "xmp", and "plaintext").
509 if the flag is FALSE for a text event, then you should normally
511 either use "dtext" or decode the entities yourself before the text
512 is processed further.
513 "length"
515 Length causes the number of bytes of the source text of the event
516 to be passed.
517 "line"
519 Line causes the line number of the start of the event to be passed.
520 The first line in the document is 1. Line counting doesn't start
521 until at least one handler requests this value to be reported.
522 "offset"
524 Offset causes the byte position in the HTML document of the start
525 of the event to be passed. The first byte in the document has off‐
526 set 0.
527 "offset_end"
529 Offset_end causes the byte position in the HTML document of the end
530 of the event to be passed. This is the same as "offset" +
531 "length".
532 "self"
534 Self causes the current object to be passed to the handler. If the
535 handler is a method, this must be the first element in the argspec.
536 An alternative to passing self as an argspec is to register clo‐
538 sures that capture $self by themselves as handlers. Unfortunately
539 this creates circular references which prevent the HTML::Parser
540 object from being garbage collected. Using the "self" argspec
541 avoids this problem.
542 "skipped_text"
544 Skipped_text returns the concatenated text of all the events that
545 have been skipped since the last time an event was reported.
546 Events might be skipped because no handler is registered for them
547 or because some filter applies. Skipped text also includes marked
548 section markup, since there are no events that can catch it.
549 If an ""-handler is registered for an event, then the text for this
551 event is not included in "skipped_text". Skipped text both before
552 and after the ""-event is included in the next reported
553 "skipped_text".
554 "tag"
556 Same as "tagname", but prefixed with "/" if it belongs to an "end"
557 event and "!" for a declaration. The "tag" does not have any pre‐
558 fix for "start" events, and is in this case identical to "tagname".
559 "tagname"
561 This is the element name (or generic identifier in SGML jargon) for
562 start and end tags. Since HTML is case insensitive, this name is
563 forced to lower case to ease string matching.
564 Since XML is case sensitive, the tagname case is not changed when
566 "xml_mode" is enabled. The same happens if the "case_sensitive"
567 attribute is set.
568 The declaration type of declaration elements is also passed as a
570 tagname, even if that is a bit strange. In fact, in the current
571 implementation tagname is identical to "token0" except that the
572 name may be forced to lower case.
573 "token0"
575 Token0 causes the original text of the first token string to be
576 passed. This should always be the same as $tokens->[0].
577 For "declaration" events, this is the declaration type.
579 For "start" and "end" events, this is the tag name.
581 For "process" and non-strict "comment" events, this is everything
583 inside the tag.
584 This passes undef if there are no tokens in the event.
586 "tokenpos"
588 Tokenpos causes a reference to an array of token positions to be
589 passed. For each string that appears in "tokens", this array con‐
590 tains two numbers. The first number is the offset of the start of
591 the token in the original "text" and the second number is the
592 length of the token.
593 Boolean attributes in a "start" event will have (0,0) for the
595 attribute value offset and length.
596 This passes undef if there are no tokens in the event (e.g.,
598 "text") and for artificial "end" events triggered by empty element
599 tags.
600 If you are using these offsets and lengths to modify "text", you
602 should either work from right to left, or be very careful to calcu‐
603 late the changes to the offsets.
604 "tokens"
606 Tokens causes a reference to an array of token strings to be
607 passed. The strings are exactly as they were found in the original
608 text, no decoding or case changes are applied.
609 For "declaration" events, the array contains each word, comment,
611 and delimited string starting with the declaration type.
612 For "comment" events, this contains each sub-comment. If
614 $p->strict_comments is disabled, there will be only one sub-com‐
615 ment.
616 For "start" events, this contains the original tag name followed by
618 the attribute name/value pairs. The values of boolean attributes
619 will be either the value set by $p->boolean_attribute_value, or the
620 attribute name if no value has been set by $p->bool‐
621 ean_attribute_value.
622 For "end" events, this contains the original tag name (always one
624 token).
625 For "process" events, this contains the process instructions
627 (always one token).
628 This passes "undef" for "text" events.
630 "text"
632 Text causes the source text (including markup element delimiters)
633 to be passed.
634 "undef"
636 Pass an undefined value. Useful as padding where the same handler
637 routine is registered for multiple events.
638 '...'
640 A literal string of 0 to 255 characters enclosed in single (') or
641 double (") quotes is passed as entered.
642 The whole argspec string can be wrapped up in '@{...}' to signal that
644 the resulting event array should be flattened. This only makes a dif‐
645 ference if an array reference is used as the handler target. Consider
646 this example:
647 $p->handler(text => [], 'text');
649 $p->handler(text => [], '@{text}']);
650 With two text events; "foo", "bar"; then the first example will end up
652 with [["foo"], ["bar"]] and the second with ["foo", "bar"] in the han‐
653 dler target array.
654 Events
656 Handlers for the following events can be registered:
658 "comment"
660 This event is triggered when a markup comment is recognized.
661 Example:
663 <!-- This is a comment -- -- So is this -->
665 "declaration"
667 This event is triggered when a markup declaration is recognized.
668 For typical HTML documents, the only declaration you are likely to
670 find is <!DOCTYPE ...>.
671 Example:
673 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
675 "http://www.w3.org/TR/html40/strict.dtd">
676 DTDs inside <!DOCTYPE ...> will confuse HTML::Parser.
678 "default"
680 This event is triggered for events that do not have a specific han‐
681 dler. You can set up a handler for this event to catch stuff you
682 did not want to catch explicitly.
683 "end"
685 This event is triggered when an end tag is recognized.
686 Example:
688 </A>
690 "end_document"
692 This event is triggered when $p->eof is called and after any
693 remaining text is flushed. There is no document text associated
694 with this event.
695 "process"
697 This event is triggered when a processing instructions markup is
698 recognized.
699 The format and content of processing instructions are system and
701 application dependent.
702 Examples:
704 <? HTML processing instructions >
706 <? XML processing instructions ?>
707 "start"
709 This event is triggered when a start tag is recognized.
710 Example:
712 <A HREF="http://www.perl.com/">
714 "start_document"
716 This event is triggered before any other events for a new document.
717 A handler for it can be used to initialize stuff. There is no doc‐
718 ument text associated with this event.
719 "text"
721 This event is triggered when plain text (characters) is recognized.
722 The text may contain multiple lines. A sequence of text may be
723 broken between several text events unless $p->unbroken_text is
724 enabled.
725 The parser will make sure that it does not break a word or a
727 sequence of whitespace between two text events.
728 Unicode
730 The "HTML::Parser" can parse Unicode strings when running under
732 perl-5.8 or better. If Unicode is passed to $p->parse() then chunks of
733 Unicode will be reported to the handlers. The offset and length
734 argspecs will also report their position in terms of characters.
735 It is safe to parse raw undecoded UTF-8 if you either avoid decoding
737 entities and make sure to not use argspecs that do, or enable the
738 "utf8_mode" for the parser. Parsing of undecoded UTF-8 might be useful
739 when parsing from a file where you need the reported offsets and
740 lengths to match the byte offsets in the file.
741 If a filename is passed to $p->parse_file() then the file will be read
743 in binary mode. This will be fine if the file contains only ASCII or
744 Latin-1 characters. If the file contains UTF-8 encoded text then care
745 must be taken when decoding entities as described in the previous para‐
746 graph, but better is to open the file with the UTF-8 layer so that it
747 is decoded properly:
748 open(my $fh, "<:utf8", "index.html") ⎪⎪ die "...: $!";
750 $p->parse_file($fh);
751 If the file contains text encoded in a charset besides ASCII, Latin-1
753 or UTF-8 then decoding will always be needed.
754
756 When an "HTML::Parser" object is constructed with no arguments, a set
757 of handlers is automatically provided that is compatible with the old
758 HTML::Parser version 2 callback methods.
759 This is equivalent to the following method calls:
761 $p->handler(start => "start", "self, tagname, attr, attrseq, text");
763 $p->handler(end => "end", "self, tagname, text");
764 $p->handler(text => "text", "self, text, is_cdata");
765 $p->handler(process => "process", "self, token0, text");
766 $p->handler(comment =>
767 sub {
768 my($self, $tokens) = @_;
769 for (@$tokens) {$self->comment($_);}},
770 "self, tokens");
771 $p->handler(declaration =>
772 sub {
773 my $self = shift;
774 $self->declaration(substr($_[0], 2, -1));},
775 "self, text");
776 Setting up these handlers can also be requested with the "api_version
778 => 2" constructor option.
779
781 The "HTML::Parser" class is subclassable. Parser objects are plain
782 hashes and "HTML::Parser" reserves only hash keys that start with
783 "_hparser". The parser state can be set up by invoking the init()
784 method, which takes the same arguments as new().
785
787 The first simple example shows how you might strip out comments from an
788 HTML document. We achieve this by setting up a comment handler that
789 does nothing and a default handler that will print out anything else:
790 use HTML::Parser;
792 HTML::Parser->new(default_h => [sub { print shift }, 'text'],
793 comment_h => [""],
794 )->parse_file(shift ⎪⎪ die) ⎪⎪ die $!;
795 An alternative implementation is:
797 use HTML::Parser;
799 HTML::Parser->new(end_document_h => [sub { print shift },
800 'skipped_text'],
801 comment_h => [""],
802 )->parse_file(shift ⎪⎪ die) ⎪⎪ die $!;
803 This will in most cases be much more efficient since only a single
805 callback will be made.
806 The next example prints out the text that is inside the <title> element
808 of an HTML document. Here we start by setting up a start handler.
809 When it sees the title start tag it enables a text handler that prints
810 any text found and an end handler that will terminate parsing as soon
811 as the title end tag is seen:
812 use HTML::Parser ();
814 sub start_handler
816 {
817 return if shift ne "title";
818 my $self = shift;
819 $self->handler(text => sub { print shift }, "dtext");
820 $self->handler(end => sub { shift->eof if shift eq "title"; },
821 "tagname,self");
822 }
823 my $p = HTML::Parser->new(api_version => 3);
825 $p->handler( start => \&start_handler, "tagname,self");
826 $p->parse_file(shift ⎪⎪ die) ⎪⎪ die $!;
827 print "\n";
828 More examples are found in the eg/ directory of the "HTML-Parser" dis‐
830 tribution: the program "hrefsub" shows how you can edit all links found
831 in a document; the program "htextsub" shows how to edit the text only;
832 the program "hstrip" shows how you can strip out certain tags/elements
833 and/or attributes; and the program "htext" show how to obtain the plain
834 text, but not any script/style content.
835 You can browse the eg/ directory online from the [Browse] link on the
837 http://search.cpan.org/~gaas/HTML-Parser/ page.
838
840 The <style> and <script> sections do not end with the first "</", but
841 need the complete corresponding end tag. The standard behaviour is not
842 really practical.
843 When the strict_comment option is enabled, we still recognize comments
845 where there is something other than whitespace between even and odd
846 "--" markers.
847 Once $p->boolean_attribute_value has been set, there is no way to
849 restore the default behaviour.
850 There is currently no way to get both quote characters into the same
852 literal argspec.
853 Empty tags, e.g. "<>" and "</>", are not recognized. SGML allows them
855 to repeat the previous start tag or close the previous start tag
856 respectively.
857 NET tags, e.g. "code/.../" are not recognized. This is SGML shorthand
859 for "<code>...</code>".
860 Unclosed start or end tags, e.g. "<tt<b>...</b</tt>" are not recog‐
862 nized.
863
865 The following messages may be produced by HTML::Parser. The notation
866 in this listing is the same as used in perldiag:
867 Not a reference to a hash
869 (F) The object blessed into or subclassed from HTML::Parser is not
870 a hash as required by the HTML::Parser methods.
871 Bad signature in parser state object at %p
873 (F) The _hparser_xs_state element does not refer to a valid state
874 structure. Something must have changed the internal value stored
875 in this hash element, or the memory has been overwritten.
876 _hparser_xs_state element is not a reference
878 (F) The _hparser_xs_state element has been destroyed.
879 Can't find '_hparser_xs_state' element in HTML::Parser hash
881 (F) The _hparser_xs_state element is missing from the parser hash.
882 It was either deleted, or not created when the object was created.
883 API version %s not supported by HTML::Parser %s
885 (F) The constructor option 'api_version' with an argument greater
886 than or equal to 4 is reserved for future extensions.
887 Bad constructor option '%s'
889 (F) An unknown constructor option key was passed to the new() or
890 init() methods.
891 Parse loop not allowed
893 (F) A handler invoked the parse() or parse_file() method. This is
894 not permitted.
895 marked sections not supported
897 (F) The $p->marked_sections() method was invoked in a HTML::Parser
898 module that was compiled without support for marked sections.
899 Unknown boolean attribute (%d)
901 (F) Something is wrong with the internal logic that set up aliases
902 for boolean attributes.
903 Only code or array references allowed as handler
905 (F) The second argument for $p->handler must be either a subroutine
906 reference, then name of a subroutine or method, or a reference to
907 an array.
908 No handler for %s events
910 (F) The first argument to $p->handler must be a valid event name;
911 i.e. one of "start", "end", "text", "process", "declaration" or
912 "comment".
913 Unrecognized identifier %s in argspec
915 (F) The identifier is not a known argspec name. Use one of the
916 names mentioned in the argspec section above.
917 Literal string is longer than 255 chars in argspec
919 (F) The current implementation limits the length of literals in an
920 argspec to 255 characters. Make the literal shorter.
921 Backslash reserved for literal string in argspec
923 (F) The backslash character "\" is not allowed in argspec literals.
924 It is reserved to permit quoting inside a literal in a later ver‐
925 sion.
926 Unterminated literal string in argspec
928 (F) The terminating quote character for a literal was not found.
929 Bad argspec (%s)
931 (F) Only identifier names, literals, spaces and commas are allowed
932 in argspecs.
933 Missing comma separator in argspec
935 (F) Identifiers in an argspec must be separated with ",".
936 Parsing of undecoded UTF-8 will give garbage when decoding entities
938 (W) The first chunk parsed appears to contain undecoded UTF-8 and
939 one or more argspecs that decode entities are used for the callback
940 handlers.
941 The result of decoding will be a mix of encoded and decoded charac‐
943 ters for any entities that expand to characters with code above
944 127. This is not a good thing.
945 The solution is to use the Encode::encode_utf8() on the data before
947 feeding it to the $p->parse(). For $p->parse_file() pass a file
948 that has been opened in ":utf8" mode.
949 The parser can process raw undecoded UTF-8 sanely if the
951 "utf8_mode" is enabled or if the "attr", "@attr" or "dtext"
952 argspecs is avoided.
953 Parsing string decoded with wrong endianess
955 (W) The first character in the document is U+FFFE. This is not a
956 legal Unicode character but a byte swapped BOM. The result of
957 parsing will likely be garbage.
958 Parsing of undecoded UTF-32
960 (W) The parser found the Unicode UTF-32 BOM signature at the start
961 of the document. The result of parsing will likely be garbage.
962 Parsing of undecoded UTF-16
964 (W) The parser found the Unicode UTF-16 BOM signature at the start
965 of the document. The result of parsing will likely be garbage.
966
968 HTML::Entities, HTML::PullParser, HTML::TokeParser, HTML::HeadParser,
969 HTML::LinkExtor, HTML::Form
970 HTML::TreeBuilder (part of the HTML-Tree distribution)
972 http://www.w3.org/TR/html4
974 More information about marked sections and processing instructions may
976 be found at "http://www.sgml.u-net.com/book/sgml-8.htm".
977
979 Copyright 1996-2007 Gisle Aas. All rights reserved.
980 Copyright 1999-2000 Michael A. Chase. All rights reserved.
981 This library is free software; you can redistribute it and/or modify it
983 under the same terms as Perl itself.
984perl v5.8.8 2006-04-26 Parser(3)