22.1 The sgmllib Module

The name of the sgmllib module is misleading: sgmllib parses only a tiny subset of SGML, but it is still a good way to get information from HTML files. sgmllib supplies one class, SGMLParser, which you subclass to override and add methods. The most frequently used methods of an instance s of your subclass X of SGMLParser are as follows.

s.close(  )

Tells the parser that there is no more input data. When X overrides close, x.close must call SGMLParser.close to ensure that buffered data get processed.

s.do_tag(attributes)

X supplies a method with such a name for each tag, with no corresponding end tag, that X wants to process. tag must be in lowercase in the method name, but can be in any mix of cases in the parsed text. SGMLParser's handle_tag method calls do_tag as appropriate. attributes is a list of pairs (name,value), where name is each attribute's name, lowercased, and value is the value, processed to resolve entity references and character references and to remove surrounding quotes.

s.end_tag(  )

X supplies a method with such a name for each tag whose end tag X wants to process. tag must be in lowercase in the method name, but can be in any mix of cases in the parsed text. X must also supply a method named start_tag, otherwise end_tag is ignored. SGMLParser's handle_endtag method calls end_tag as appropriate.

s.feed(data)

Passes to the parser some of the text being parsed. The parser may process some prefix of the text, holding the rest in a buffer until the next call to s.feed or s.close.

s.handle_charref(ref)

Called to process a character reference '&#ref;'. SGMLParser's implementation of handle_charref handles decimal numbers in range(0,256), like:

def handle_charref(self, ref):
    try: 
        c = chr(int(ref))
    except (TypeError, ValueError): 
        self.unknown_charref(ref)
    else: self.handle_data(c)

Your subclass X may override handle_charref or unknown_charref in order to support other forms of character references '&#...;'.

s.handle_comment(comment)

Called to handle comments. comment is the string within '<!--...-->', without the delimiters. SGMLParser's implementation of handle_comment does nothing.

s.handle_data(data)

Called to process each arbitrary string data. Your subclass X normally overrides handle_data. SGMLParser's implementation of handle_data does nothing.

s.handle_endtag(tag,method)

Called to handle termination tags for which X supplies methods named start_tag and end_tag. tag is the tag string, lowercased. method is the bound method for end_tag. SGMLParser's implementation of handle_endtag calls method( ).

s.handle_entityref(ref)

Called to process an entity reference '&ref;'. SGMLParser's implementation of handle_entityref looks ref up in s.entitydefs, like:

def handle_entityref(self, ref):

    try: t = self.entitydefs[ref]
    except KeyError: self.unknown_entityref(ref)
    else: self.handle_data(t)

Your subclass X may override handle_entityref or unknown_entityref in order to support entity references '&...;' in different ways. SGMLParser's attribute entitydefs includes keys 'amp', 'apos', 'gt', 'lt', and 'quot'.

s.handle_starttag(tag, method, attributes)

Called to handle tags for which X supplies a method start_tag or do_tag. tag is the tag string, lowercased. method is the bound method for start_tag or do_tag. attributes is a list of pairs (name,value), where name is each attribute's name, lowercased, and value is the value, processed to resolve entity references and character references and to remove surrounding quotes. When X supplies both start_tag and do_tag methods, start_tag has precedence and do_tag is ignored. SGMLParser's implementation of handle_starttag calls method(attributes).

s.report_unbalanced(tag)

Called when tags terminate without being open. tag is the tag string, lowercased. SGMLParser's implementation of report_unbalanced does nothing.

s.start_tag(attributes)

X supplies a method thus named for each tag, with an end tag, that X wants to process. tag must be in lowercase in the method name, but can be in any mix of cases in the parsed text. SGMLParser's handle_tag method calls start_tag as appropriate. attributes is a list of pairs (name,value), where name is each attribute's name, lowercased, and value is the value, processed to resolve entity references and character references and to remove surrounding quotes.

s.unknown_charref(ref)

Called to process invalid or unrecognized character references. SGMLParser's implementation of unknown_charref does nothing.

s.unknown_endtag(tag)

Called to process termination tags for which X supplies no specific method. SGMLParser's implementation of unknown_endtag does nothing.

s.unknown_entityref(ref)

Called to process unknown entity references. SGMLParser's implementation of unknown_entityref does nothing.

s.unknown_starttag(tag, attributes)

Called to process tags for which X supplies no specific method. tag is the tag string, lowercased. attributes is a list of pairs (name,value), where name is each attribute's name, lowercased, and value is the value, processed to resolve entity references and character references and to remove surrounding quotes. SGMLParser's implementation of unknown_starttag does nothing.

The following example uses sgmllib for a typical HTML-related task: fetching a page from the Web with urllib, parsing it, and outputting the hyperlinks. The example uses urlparse to check the page's links, and outputs only links whose URLs have an explicit scheme of 'http'.

import sgmllib, urllib, urlparse

class LinksParser(sgmllib.SGMLParser):
    def __init_  _(self):
        sgmllib.SGMLParser.__init_  _(self)
        self.seen = {}
    def do_a(self, attributes):
        for name, value in attributes:
            if name == 'href' and value not in self.seen:
                self.seen[value] = True
                pieces = urlparse.urlparse(value)
                if pieces[0] != 'http': return
                print urlparse.urlunparse(pieces)
                return

p = LinksParser(  )
f = urllib.urlopen('http://www.python.org/index.html')
BUFSIZE = 8192
while True:
    data = f.read(BUFSIZE)
    if not data: break
    p.feed(data)
p.close(  )

Class LinksParser only needs to define method do_a. The superclass calls back to this method for all <a> tags, and the method loops on the attributes, looking for one named 'href', then works with the corresponding value (i.e., the relevant URL).