use strict;
use warnings;
use HTML::Parser ();
# Create parser object
my $p = HTML::Parser->new(
api_version => 3,
start_h => [\&start, "tagname, attr"],
end_h => [\&end, "tagname"],
marked_sections => 1,
);
# Parse document text chunk by chunk
$p->parse($chunk1);
$p->parse($chunk2);
# ...
# signal end of document
$p->eof;
# Parse directly from file
$p->parse_file("foo.html");
# or
open(my $fh, "<:utf8", "foo.html") || die;
$p->parse_file($fh);
"HTML::Parser" is not a generic SGML parser. We have tried to make it able to deal with the HTML that is actually ``out there'', and it normally parses as closely as possible to the way the popular web browsers do it instead of strictly following one of the many HTML specifications from W3C. Where there is disagreement, there is often an option that you can enable to get the official behaviour.
The document to be parsed may be supplied in arbitrary chunks. This makes on-the-fly parsing as documents are received from the network possible.
If event driven parsing does not feel right for your application, you might want to use "HTML::PullParser". This is an "HTML::Parser" subclass that allows a more conventional program structure.
If a top level key is in the form ``<event>_h'' (e.g., ``text_h'') then it assigns a handler to that event, otherwise it initializes a parser option. The event handler specification value must be an array reference. Multiple handlers may also be assigned with the 'handlers => [%handlers]' option. See examples below.
If new() is called without any arguments, it will create a parser that uses callback methods compatible with version 2 of "HTML::Parser". See the section on ``version 2 compatibility'' below for details.
The special constructor option 'api_version => 2' can be used to initialize version 2 callbacks while still setting other options and handlers. The 'api_version => 3' option can be used if you don't want to set any options and don't want to fall back to v2 compatible mode.
Examples:
$p = HTML::Parser->new(
api_version => 3,
text_h => [ sub {...}, "dtext" ]
);
This creates a new parser object with a text event handler subroutine that receives the original text with general entities decoded.
$p = HTML::Parser->new( api_version => 3, start_h => [ 'my_start', "self,tokens" ] );
This creates a new parser object with a start event handler method that receives the $p and the tokens array.
$p = HTML::Parser->new(
api_version => 3,
handlers => {
text => [\@array, "event,text"],
comment => [\@array, "event,text"],
}
);
This creates a new parser object that stores the event type and the original text in @array for text and comment events.
The following methods feed the HTML document to the "HTML::Parser" object:
If an invoked event handler aborts parsing by calling $p->eof, then $p->parse() will return a FALSE value. Otherwise the return value is a reference to the parser object ($p).
Parsing will also abort if one of the event handlers calls $p->eof.
The effect of this is the same as:
while (1) {
my $chunk = &$code_ref();
if (!defined($chunk) || !length($chunk)) {
$p->eof;
return $p;
}
$p->parse($chunk) || return undef;
}
But it is more efficient as this loop runs internally in XS code.
If $file contains a filename and the file can't be opened, then the method returns an undefined value and $! tells why it failed. Otherwise the return value is a reference to the parser object.
If a file handle is passed as the $file argument, then the file will normally be read until EOF, but not closed.
If an invoked event handler aborts parsing by calling $p->eof, then $p->parse_file() may not have read the entire file.
On systems with multi-byte line terminators, the values passed for the offset and length argspecs may be too low if parse_file() is called on a file handle that is not in binary mode.
If a filename is passed in, then parse_file() will open the file in binary mode.
Calling $p->eof inside a handler will terminate parsing at that point and cause $p->parse to return a FALSE value. This also terminates parsing by $p->parse_file().
After $p->eof has been called, the parse() and parse_file() methods can be invoked to feed new documents with the parser object.
The return value from eof() is a reference to the parser object.
Most parser options are controlled by boolean attributes. Each boolean attribute is enabled by calling the corresponding method with a TRUE argument and disabled with a FALSE argument. The attribute value is left unchanged if no argument is given. The return value from each method is the old attribute value.
Methods that can be used to get and/or set parser options are:
Empty element tags look like start tags, but end with the character sequence ``/>'' instead of ``>''. When recognized by "HTML::Parser" they cause an artificial end event in addition to the start event. The "text" for the artificial end event will be empty and the "tokenpos" array will be undefined even though the token array will have one element containing the tag name.
There are currently no events associated with the marked section markup, but the text can be returned as "skipped_text".
The official behaviour is enabled by enabling this attribute.
Enabling of 'strict_comment' also disables recognizing these forms as comments:
</ comment> <! comment>
The official behaviour is enabled with this attribute. If enabled, only whitespace is allowed between the tagname and the final ``>''.
<IMG SRC=newprevlstGr.gif ALT=[PREV LIST] BORDER=0>
By default, ``LIST]'' is parsed as a boolean attribute, not as part of the ALT value as was clearly intended. This is also what Mozilla sees.
The official behaviour is enabled by enabling this attribute. If enabled, it will cause the tag above to be reported as text since ``LIST]'' is not a legal attribute name.
Note that the "offset" argspec will give you the offset of the first segment of text and "length" is the combined length of the segments. Since there might be ignored tags in between, these numbers can't be used to directly index in the original document file.
If "utf8_mode" is enabled then it is an error to pass strings containing characters with code above 255 to the parse() method, and the parse() method will croak if you try.
Example: The Unicode character ``\x{2665}'' is ``\xE2\x99\xA5'' when UTF-8 encoded. The character can also be represented by the entity ``♥'' or ``♥''. If we feed the parser:
$p->parse("\xE2\x99\xA5♥");
then "dtext" will be reported as ``\xE2\x99\xA5\x{2665}'' without "utf8_mode" enabled, but as ``\xE2\x99\xA5\xE2\x99\xA5'' when enabled. The later string is what you want.
This option is only available with perl-5.8 or better.
As markup and text is recognized, handlers are invoked. The following method is used to set up handlers for different events:
Event is one of "text", "start", "end", "declaration", "comment", "process", "start_document", "end_document" or "default".
The "\&subroutine" is a reference to a subroutine which is called to handle the event.
The $method_name is the name of a method of $p which is called to handle the event.
The @accum is an array that will hold the event information as sub-arrays.
If the second argument is "", the event is ignored. If it is undef, the default handler is invoked for the event.
The $argspec is a string that describes the information to be reported for the event. Any requested information that does not apply to a specific event is passed as "undef". If argspec is omitted, then it is left unchanged.
The return value from $p->handler is the old callback routine or a reference to the accumulator array.
Any return values from handler callback routines/methods are always ignored. A handler callback can request parsing to be aborted by invoking the $p->eof method. A handler callback is not allowed to invoke the $p->parse() or $p->parse_file() method. An exception will be raised if it tries.
Examples:
$p->handler(start => "start", 'self, attr, attrseq, text' );
This causes the ``start'' method of object $p to be called for 'start' events. The callback signature is "$p->start(\%attr, \@attr_seq, $text)".
$p->handler(start => \&start, 'attr, attrseq, text' );
This causes subroutine start() to be called for 'start' events. The callback signature is start(\%attr, \@attr_seq, $text).
$p->handler(start => \@accum, '"S", attr, attrseq, text' );
This causes 'start' event information to be saved in @accum. The array elements will be ['S', \%attr, \@attr_seq, $text].
$p->handler(start => "");
This causes 'start' events to be ignored. It also suppresses invocations of any default handler for start events. It is in most cases equivalent to $p->handler(start => sub {}), but is more efficient. It is different from the empty-sub-handler in that "skipped_text" is not reset by it.
$p->handler(start => undef);
This causes no handler to be associated with start events. If there is a default handler it will be invoked.
Filters based on tags can be set up to limit the number of events reported. The main bottleneck during parsing is often the huge number of callbacks made from the parser. Applying filters can improve performance significantly.
The following methods control filters:
$p->ignore_elements(qw(script style));
The "script" and "style" tags will always nest properly since their content is parsed in CDATA mode. For most other tags "ignore_elements" must be used with caution since HTML is often not well formed.
Internally, the system has two filter lists, one for "report_tags" and one for "ignore_tags", and both filters are applied. This effectively gives "ignore_tags" precedence over "report_tags".
Examples:
$p->ignore_tags(qw(style)); $p->report_tags(qw(script style));
results in only "script" events being reported.
Boolean attributes' values are either the value set by $p->boolean_attribute_value, or the attribute name if no value has been set by $p->boolean_attribute_value.
This passes undef except for "start" events.
Unless "xml_mode" or "case_sensitive" is enabled, the attribute names are forced to lower case.
General entities are decoded in the attribute values and one layer of matching quotes enclosing the attribute values is removed.
The Unicode character set is assumed for entity decoding.
@attr = map { $_ => $attr->{$_} } @$attrseq;
assuming $attr and $attrseq here are the hash and array passed as the result of "attr" and "attrseq" argspecs.
This passes no values for events besides "start".
This passes undef except for "start" events.
Unless "xml_mode" or "case_sensitive" is enabled, the attribute names are forced to lower case.
The Unicode character set is assumed for entity decoding. With Perl version 5.6 or earlier only the Latin-1 range is supported, and entities for characters outside the range 0..255 are left unchanged.
This passes undef except for "text" events.
The event name is one of "text", "start", "end", "declaration", "comment", "process", "start_document" or "end_document".
if the flag is FALSE for a text event, then you should normally either use "dtext" or decode the entities yourself before the text is processed further.
An alternative to passing self as an argspec is to register closures that capture $self by themselves as handlers. Unfortunately this creates circular references which prevent the HTML::Parser object from being garbage collected. Using the "self" argspec avoids this problem.
If an ""-handler is registered for an event, then the text for this event is not included in "skipped_text". Skipped text both before and after the ""-event is included in the next reported "skipped_text".
Since XML is case sensitive, the tagname case is not changed when "xml_mode" is enabled. The same happens if the "case_sensitive" attribute is set.
The declaration type of declaration elements is also passed as a tagname, even if that is a bit strange. In fact, in the current implementation tagname is identical to "token0" except that the name may be forced to lower case.
For "declaration" events, this is the declaration type.
For "start" and "end" events, this is the tag name.
For "process" and non-strict "comment" events, this is everything inside the tag.
This passes undef if there are no tokens in the event.
Boolean attributes in a "start" event will have (0,0) for the attribute value offset and length.
This passes undef if there are no tokens in the event (e.g., "text") and for artificial "end" events triggered by empty element tags.
If you are using these offsets and lengths to modify "text", you should either work from right to left, or be very careful to calculate the changes to the offsets.
For "declaration" events, the array contains each word, comment, and delimited string starting with the declaration type.
For "comment" events, this contains each sub-comment. If $p->strict_comments is disabled, there will be only one sub-comment.
For "start" events, this contains the original tag name followed by the attribute name/value pairs. The values of boolean attributes will be either the value set by $p->boolean_attribute_value, or the attribute name if no value has been set by $p->boolean_attribute_value.
For "end" events, this contains the original tag name (always one token).
For "process" events, this contains the process instructions (always one token).
This passes "undef" for "text" events.
The whole argspec string can be wrapped up in '@{...}' to signal that the resulting event array should be flattened. This only makes a difference if an array reference is used as the handler target. Consider this example:
$p->handler(text => [], 'text');
$p->handler(text => [], '@{text}']);
With two text events; "foo", "bar"; then the first example will end up with [[``foo''], [``bar'']] and the second with [``foo'', ``bar''] in the handler target array.
Example:
<!-- This is a comment -- -- So is this -->
For typical HTML documents, the only declaration you are likely to find is <!DOCTYPE ...>.
Example:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
DTDs inside <!DOCTYPE ...> will confuse HTML::Parser.
Example:
</A>
The format and content of processing instructions are system and application dependent.
Examples:
<? HTML processing instructions > <? XML processing instructions ?>
Example:
<A HREF="http://www.perl.com/">
The parser will make sure that it does not break a word or a sequence of whitespace between two text events.
It is safe to parse raw undecoded UTF-8 if you either avoid decoding entities and make sure to not use argspecs that do, or enable the "utf8_mode" for the parser. Parsing of undecoded UTF-8 might be useful when parsing from a file where you need the reported offsets and lengths to match the byte offsets in the file.
If a filename is passed to $p->parse_file() then the file will be read in binary mode. This will be fine if the file contains only ASCII or Latin-1 characters. If the file contains UTF-8 encoded text then care must be taken when decoding entities as described in the previous paragraph, but better is to open the file with the UTF-8 layer so that it is decoded properly:
open(my $fh, "<:utf8", "index.html") || die "...: $!"; $p->parse_file($fh);
If the file contains text encoded in a charset besides ASCII, Latin-1 or UTF-8 then decoding will always be needed.
This is equivalent to the following method calls:
$p->handler(start => "start", "self, tagname, attr, attrseq, text");
$p->handler(end => "end", "self, tagname, text");
$p->handler(text => "text", "self, text, is_cdata");
$p->handler(process => "process", "self, token0, text");
$p->handler(
comment => sub {
my($self, $tokens) = @_;
for (@$tokens) {$self->comment($_);}
},
"self, tokens"
);
$p->handler(
declaration => sub {
my $self = shift;
$self->declaration(substr($_[0], 2, -1));
},
"self, text"
);
Setting up these handlers can also be requested with the ``api_version => 2'' constructor option.
use HTML::Parser;
HTML::Parser->new(
default_h => [sub { print shift }, 'text'],
comment_h => [""],
)->parse_file(shift || die) || die $!;
An alternative implementation is:
use HTML::Parser;
HTML::Parser->new(
end_document_h => [sub { print shift }, 'skipped_text'],
comment_h => [""],
)->parse_file(shift || die) || die $!;
This will in most cases be much more efficient since only a single callback will be made.
The next example prints out the text that is inside the <title> element of an HTML document. Here we start by setting up a start handler. When it sees the title start tag it enables a text handler that prints any text found and an end handler that will terminate parsing as soon as the title end tag is seen:
use HTML::Parser ();
sub start_handler {
return if shift ne "title";
my $self = shift;
$self->handler(text => sub { print shift }, "dtext");
$self->handler(
end => sub {
shift->eof if shift eq "title";
},
"tagname,self"
);
}
my $p = HTML::Parser->new(api_version => 3);
$p->handler(start => \&start_handler, "tagname,self");
$p->parse_file(shift || die) || die $!;
print "\n";
More examples are found in the eg/ directory of the "HTML-Parser" distribution: the program "hrefsub" shows how you can edit all links found in a document; the program "htextsub" shows how to edit the text only; the program "hstrip" shows how you can strip out certain tags/elements and/or attributes; and the program "htext" show how to obtain the plain text, but not any script/style content.
You can browse the eg/ directory online from the [Browse] link on the http://search.cpan.org/~gaas/HTML-Parser/ page.
When the strict_comment option is enabled, we still recognize comments where there is something other than whitespace between even and odd ``--'' markers.
Once $p->boolean_attribute_value has been set, there is no way to restore the default behaviour.
There is currently no way to get both quote characters into the same literal argspec.
Empty tags, e.g. ``<>'' and ``</>'', are not recognized. SGML allows them to repeat the previous start tag or close the previous start tag respectively.
NET tags, e.g. ``code/.../'' are not recognized. This is SGML shorthand for ``<code>...</code>''.
Incomplete start or end tags, e.g. ``<tt<b>...</b</tt>'' are not recognized.
The result of decoding will be a mix of encoded and decoded characters for any entities that expand to characters with code above 127. This is not a good thing.
The recommended solution is to apply Encode::decode_utf8() on the data before feeding it to the $p->parse(). For $p->parse_file() pass a file that has been opened in ``:utf8'' mode.
The alternative solution is to enable the "utf8_mode" and not decode before passing strings to $p->parse(). The parser can process raw undecoded UTF-8 sanely if the "utf8_mode" is enabled, or if the "attr", @attr or "dtext" argspecs are avoided.
HTML::TreeBuilder (part of the HTML-Tree distribution)
More information about marked sections and processing instructions may be found at <http://www.is-thought.co.uk/book/sgml-8.htm>.
Copyright 1996-2016 Gisle Aas. All rights reserved. Copyright 1999-2000 Michael A. Chase. All rights reserved.
This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.