root/BeautifulSoup.py

"""Beautiful Soup
Elixir and Tonic
"The Screen-Scraper's Friend"
http://www.crummy.com/software/BeautifulSoup/

Beautiful Soup parses a (possibly invalid) XML or HTML document into a
tree representation. It provides methods and Pythonic idioms that make
it easy to navigate, search, and modify the tree.

A well-structured XML/HTML document yields a well-behaved data
structure. An ill-structured XML/HTML document yields a
correspondingly ill-behaved data structure. If your document is only
locally well-structured, you can use this library to find and process
the well-structured part of it.

Beautiful Soup works with Python 2.2 and up. It has no external
dependencies, but you'll have more success at converting data to UTF-8
if you also install these three packages:

* chardet, for auto-detecting character encodings
  http://chardet.feedparser.org/
* cjkcodecs and iconv_codec, which add more encodings to the ones supported
  by stock Python.
  http://cjkpython.i18n.org/

Beautiful Soup defines classes for two main parsing strategies:
    
 * BeautifulStoneSoup, for parsing XML, SGML, or your domain-specific
   language that kind of looks like XML.

 * BeautifulSoup, for parsing run-of-the-mill HTML code, be it valid
   or invalid. This class has web browser-like heuristics for
   obtaining a sensible parse tree in the face of common HTML errors.

Beautiful Soup also defines a class (UnicodeDammit) for autodetecting
the encoding of an HTML or XML document, and converting it to
Unicode. Much of this code is taken from Mark Pilgrim's Universal Feed
Parser.

For more than you ever wanted to know about Beautiful Soup, see the
documentation:
http://www.crummy.com/software/BeautifulSoup/documentation.html
"""
from __future__ import generators

__author__ = "Leonard Richardson (crummy.com)"
__contributors__ = ["Sam Ruby (intertwingly.net)",
                    "the unwitting Mark Pilgrim (diveintomark.org)",
                    "http://www.crummy.com/software/BeautifulSoup/AUTHORS.html"]
__version__ = "3.0.3"
__copyright__ = "Copyright (c) 2004-2006 Leonard Richardson"
__license__ = "PSF"

from sgmllib import SGMLParser, SGMLParseError
import codecs
import types
import re
import sgmllib
from htmlentitydefs import name2codepoint

# This RE makes Beautiful Soup able to parse XML with namespaces.
sgmllib.tagfind = re.compile('[a-zA-Z][-_.:a-zA-Z0-9]*')

# This RE makes Beautiful Soup capable of recognizing numeric character
# references that use hexadecimal.
sgmllib.charref = re.compile('&#(\d+|x[0-9a-fA-F]+);')

DEFAULT_OUTPUT_ENCODING = "utf-8"

# First, the classes that represent markup elements.

class PageElement:
    """Contains the navigational information for some part of the page
    (either a tag or a piece of text)"""

    def setup(self, parent=None, previous=None):
        """Sets up the initial relations between this element and
        other elements."""        
        self.parent = parent
        self.previous = previous
        self.next = None
        self.previousSibling = None
        self.nextSibling = None
        if self.parent and self.parent.contents:
            self.previousSibling = self.parent.contents[-1]
            self.previousSibling.nextSibling = self

    def replaceWith(self, replaceWith):        
        oldParent = self.parent
        myIndex = self.parent.contents.index(self)
        if hasattr(replaceWith, 'parent') and replaceWith.parent == self.parent:
            # We're replacing this element with one of its siblings.
            index = self.parent.contents.index(replaceWith)
            if index and index < myIndex:
                # Furthermore, it comes before this element. That
                # means that when we extract it, the index of this
                # element will change.
                myIndex = myIndex - 1
        self.extract()        
        oldParent.insert(myIndex, replaceWith)
        
    def extract(self):
        """Destructively rips this element out of the tree."""        
        if self.parent:
            try:
                self.parent.contents.remove(self)
            except ValueError:
                pass

        #Find the two elements that would be next to each other if
        #this element (and any children) hadn't been parsed. Connect
        #the two.        
        lastChild = self._lastRecursiveChild()
        nextElement = lastChild.next

        if self.previous:
            self.previous.next = nextElement
        if nextElement:
            nextElement.previous = self.previous
        self.previous = None
        lastChild.next = None

        self.parent = None        
        if self.previousSibling:
            self.previousSibling.nextSibling = self.nextSibling
        if self.nextSibling:
            self.nextSibling.previousSibling = self.previousSibling
        self.previousSibling = self.nextSibling = None       

    def _lastRecursiveChild(self):
        "Finds the last element beneath this object to be parsed."
        lastChild = self
        while hasattr(lastChild, 'contents') and lastChild.contents:
            lastChild = lastChild.contents[-1]
        return lastChild

    def insert(self, position, newChild):
        if (isinstance(newChild, basestring)
            or isinstance(newChild, unicode)) \
            and not isinstance(newChild, NavigableString):
            newChild = NavigableString(newChild)        

        position =  min(position, len(self.contents))
        if hasattr(newChild, 'parent') and newChild.parent != None:
            # We're 'inserting' an element that's already one
            # of this object's children. 
            if newChild.parent == self:
                index = self.find(newChild)
                if index and index < position:
                    # Furthermore we're moving it further down the
                    # list of this object's children. That means that
                    # when we extract this element, our target index
                    # will jump down one.
                    position = position - 1
            newChild.extract()
            
        newChild.parent = self
        previousChild = None
        if position == 0:
            newChild.previousSibling = None
            newChild.previous = self
        else:
            previousChild = self.contents[position-1]
            newChild.previousSibling = previousChild
            newChild.previousSibling.nextSibling = newChild
            newChild.previous = previousChild._lastRecursiveChild()
        if newChild.previous:
            newChild.previous.next = newChild        

        newChildsLastElement = newChild._lastRecursiveChild()

        if position >= len(self.contents):
            newChild.nextSibling = None
            
            parent = self
            parentsNextSibling = None
            while not parentsNextSibling:
                parentsNextSibling = parent.nextSibling
                parent = parent.parent
                if not parent: # This is the last element in the document.
                    break
            if parentsNextSibling:
                newChildsLastElement.next = parentsNextSibling
            else:
                newChildsLastElement.next = None
        else:
            nextChild = self.contents[position]            
            newChild.nextSibling = nextChild            
            if newChild.nextSibling:
                newChild.nextSibling.previousSibling = newChild
            newChildsLastElement.next = nextChild

        if newChildsLastElement.next:
            newChildsLastElement.next.previous = newChildsLastElement
        self.contents.insert(position, newChild)

    def findNext(self, name=None, attrs={}, text=None, **kwargs):
        """Returns the first item that matches the given criteria and
        appears after this Tag in the document."""
        return self._findOne(self.findAllNext, name, attrs, text, **kwargs)

    def findAllNext(self, name=None, attrs={}, text=None, limit=None,
                    **kwargs):
        """Returns all items that match the given criteria and appear
        before after Tag in the document."""
        return self._findAll(name, attrs, text, limit, self.nextGenerator)

    def findNextSibling(self, name=None, attrs={}, text=None, **kwargs):
        """Returns the closest sibling to this Tag that matches the
        given criteria and appears after this Tag in the document."""
        return self._findOne(self.findNextSiblings, name, attrs, text,
                             **kwargs)

    def findNextSiblings(self, name=None, attrs={}, text=None, limit=None,
                         **kwargs):
        """Returns the siblings of this Tag that match the given
        criteria and appear after this Tag in the document."""
        return self._findAll(name, attrs, text, limit,
                             self.nextSiblingGenerator, **kwargs)
    fetchNextSiblings = findNextSiblings # Compatibility with pre-3.x

    def findPrevious(self, name=None, attrs={}, text=None, **kwargs):
        """Returns the first item that matches the given criteria and
        appears before this Tag in the document."""
        return self._findOne(self.findAllPrevious, name, attrs, text, **kwargs)

    def findAllPrevious(self, name=None, attrs={}, text=None, limit=None,
                        **kwargs):
        """Returns all items that match the given criteria and appear
        before this Tag in the document."""
        return self._findAll(name, attrs, text, limit, self.previousGenerator,
                           **kwargs)
    fetchPrevious = findAllPrevious # Compatibility with pre-3.x

    def findPreviousSibling(self, name=None, attrs={}, text=None, **kwargs):
        """Returns the closest sibling to this Tag that matches the
        given criteria and appears before this Tag in the document."""
        return self._findOne(self.findPreviousSiblings, name, attrs, text,
                             **kwargs)

    def findPreviousSiblings(self, name=None, attrs={}, text=None,
                             limit=None, **kwargs):
        """Returns the siblings of this Tag that match the given
        criteria and appear before this Tag in the document."""
        return self._findAll(name, attrs, text, limit,
                             self.previousSiblingGenerator, **kwargs)
    fetchPreviousSiblings = findPreviousSiblings # Compatibility with pre-3.x

    def findParent(self, name=None, attrs={}, **kwargs):
        """Returns the closest parent of this Tag that matches the given
        criteria."""
        # NOTE: We can't use _findOne because findParents takes a different
        # set of arguments.
        r = None
        l = self.findParents(name, attrs, 1)
        if l:
            r = l[0]
        return r

    def findParents(self, name=None, attrs={}, limit=None, **kwargs):
        """Returns the parents of this Tag that match the given
        criteria."""

        return self._findAll(name, attrs, None, limit, self.parentGenerator,
                             **kwargs)
    fetchParents = findParents # Compatibility with pre-3.x

    #These methods do the real heavy lifting.

    def _findOne(self, method, name, attrs, text, **kwargs):
        r = None
        l = method(name, attrs, text, 1, **kwargs)
        if l:
            r = l[0]
        return r
    
    def _findAll(self, name, attrs, text, limit, generator, **kwargs):
        "Iterates over a generator looking for things that match."

        if isinstance(name, SoupStrainer):
            strainer = name
        else:
            # Build a SoupStrainer
            strainer = SoupStrainer(name, attrs, text, **kwargs)
        results = ResultSet(strainer)
        g = generator()
        while True:
            try:
                i = g.next()
            except StopIteration:
                break
            if i:
                found = strainer.search(i)
                if found:
                    results.append(found)
                    if limit and len(results) >= limit:
                        break
        return results

    #These Generators can be used to navigate starting from both
    #NavigableStrings and Tags.                
    def nextGenerator(self):
        i = self
        while i:
            i = i.next
            yield i

    def nextSiblingGenerator(self):
        i = self
        while i:
            i = i.nextSibling
            yield i

    def previousGenerator(self):
        i = self
        while i:
            i = i.previous
            yield i

    def previousSiblingGenerator(self):
        i = self
        while i:
            i = i.previousSibling
            yield i

    def parentGenerator(self):
        i = self
        while i:
            i = i.parent
            yield i

    # Utility methods
    def substituteEncoding(self, str, encoding=None):
        encoding = encoding or "utf-8"
        return str.replace("%SOUP-ENCODING%", encoding)    

    def toEncoding(self, s, encoding=None):
        """Encodes an object to a string in some encoding, or to Unicode.
        ."""
        if isinstance(s, unicode):
            if encoding:
                s = s.encode(encoding)
        elif isinstance(s, str):
            if encoding:
                s = s.encode(encoding)
            else:
                s = unicode(s)
        else:
            if encoding:
                s  = self.toEncoding(str(s), encoding)
            else:
                s = unicode(s)
        return s

class NavigableString(unicode, PageElement):

    def __getattr__(self, attr):
        """text.string gives you text. This is for backwards
        compatibility for Navigable*String, but for CData* it lets you
        get the string without the CData wrapper."""
        if attr == 'string':
            return self
        else:
            raise AttributeError, "'%s' object has no attribute '%s'" % (self.__class__.__name__, attr)

    def __unicode__(self):
        return __str__(self, None)

    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        if encoding:
            return self.encode(encoding)
        else:
            return self
        
class CData(NavigableString):

    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        return "<![CDATA[%s]]>" % NavigableString.__str__(self, encoding)

class ProcessingInstruction(NavigableString):
    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        output = self
        if "%SOUP-ENCODING%" in output:
            output = self.substituteEncoding(output, encoding)
        return "<?%s?>" % self.toEncoding(output, encoding)

class Comment(NavigableString):
    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        return "<!--%s-->" % NavigableString.__str__(self, encoding)    

class Declaration(NavigableString):
    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        return "<!%s>" % NavigableString.__str__(self, encoding)        

class Tag(PageElement):
    """Represents a found HTML tag with its attributes and contents."""

    XML_ENTITIES_TO_CHARS = { 'apos' : "'",
                              "quot" : '"',
                              "amp" : "&",
                              "lt" : "<",
                              "gt" : ">"
                              }
    # An RE for finding ampersands that aren't the start of of a
    # numeric entity.
    BARE_AMPERSAND = re.compile("&(?!#\d+;|#x[0-9a-fA-F]+;|\w+;)")

    def __init__(self, parser, name, attrs=None, parent=None,
                 previous=None):
        "Basic constructor."

        # We don't actually store the parser object: that lets extracted
        # chunks be garbage-collected
        self.parserClass = parser.__class__
        self.isSelfClosing = parser.isSelfClosingTag(name)
        self.convertHTMLEntities = parser.convertHTMLEntities
        self.name = name
        if attrs == None:
            attrs = []
        self.attrs = attrs
        self.contents = []
        self.setup(parent, previous)
        self.hidden = False
        self.containsSubstitutions = False

    def get(self, key, default=None):
        """Returns the value of the 'key' attribute for the tag, or
        the value given for 'default' if it doesn't have that
        attribute."""
        return self._getAttrMap().get(key, default)    

    def has_key(self, key):
        return self._getAttrMap().has_key(key)

    def __getitem__(self, key):
        """tag[key] returns the value of the 'key' attribute for the tag,
        and throws an exception if it's not there."""
        return self._getAttrMap()[key]

    def __iter__(self):
        "Iterating over a tag iterates over its contents."
        return iter(self.contents)

    def __len__(self):
        "The length of a tag is the length of its list of contents."
        return len(self.contents)

    def __contains__(self, x):
        return x in self.contents

    def __nonzero__(self):
        "A tag is non-None even if it has no contents."
        return True

    def __setitem__(self, key, value):        
        """Setting tag[key] sets the value of the 'key' attribute for the
        tag."""
        self._getAttrMap()
        self.attrMap[key] = value
        found = False
        for i in range(0, len(self.attrs)):
            if self.attrs[i][0] == key:
                self.attrs[i] = (key, value)
                found = True
        if not found:
            self.attrs.append((key, value))
        self._getAttrMap()[key] = value

    def __delitem__(self, key):
        "Deleting tag[key] deletes all 'key' attributes for the tag."
        for item in self.attrs:
            if item[0] == key:
                self.attrs.remove(item)
                #We don't break because bad HTML can define the same
                #attribute multiple times.
            self._getAttrMap()
            if self.attrMap.has_key(key):
                del self.attrMap[key]

    def __call__(self, *args, **kwargs):
        """Calling a tag like a function is the same as calling its
        findAll() method. Eg. tag('a') returns a list of all the A tags
        found within this tag."""
        return apply(self.findAll, args, kwargs)

    def __getattr__(self, tag):
        #print "Getattr %s.%s" % (self.__class__, tag)
        if len(tag) > 3 and tag.rfind('Tag') == len(tag)-3:
            return self.find(tag[:-3])
        elif tag.find('__') != 0:
            return self.find(tag)

    def __eq__(self, other):
        """Returns true iff this tag has the same name, the same attributes,
        and the same contents (recursively) as the given tag.

        NOTE: right now this will return false if two tags have the
        same attributes in a different order. Should this be fixed?"""
        if not hasattr(other, 'name') or not hasattr(other, 'attrs') or not hasattr(other, 'contents') or self.name != other.name or self.attrs != other.attrs or len(self) != len(other):
            return False
        for i in range(0, len(self.contents)):
            if self.contents[i] != other.contents[i]:
                return False
        return True

    def __ne__(self, other):
        """Returns true iff this tag is not identical to the other tag,
        as defined in __eq__."""
        return not self == other

    def __repr__(self, encoding=DEFAULT_OUTPUT_ENCODING):
        """Renders this tag as a string."""
        return self.__str__(encoding)

    def __unicode__(self):
        return self.__str__(None)

    def _convertEntities(self, match):
        x = match.group(1)
        if x in name2codepoint:
            return unichr(name2codepoint[x])            
        elif "&" + x + ";" in self.XML_ENTITIES_TO_CHARS:
            return '&%s;' % x
        else:
            return '&amp;%s;' % x

    def __str__(self, encoding=DEFAULT_OUTPUT_ENCODING,
                prettyPrint=False, indentLevel=0):
        """Returns a string or Unicode representation of this tag and
        its contents. To get Unicode, pass None for encoding.

        NOTE: since Python's HTML parser consumes whitespace, this
        method is not certain to reproduce the whitespace present in
        the original string."""

        encodedName = self.toEncoding(self.name, encoding)

        attrs = []
        if self.attrs:
            for key, val in self.attrs:
                fmt = '%s="%s"'
                if isString(val):                    
                    if self.containsSubstitutions and '%SOUP-ENCODING%' in val:
                        val = self.substituteEncoding(val, encoding)

                    # The attribute value either:
                    #
                    # * Contains no embedded double quotes or single quotes.
                    #   No problem: we enclose it in double quotes.
                    # * Contains embedded single quotes. No problem:
                    #   double quotes work here too.
                    # * Contains embedded double quotes. No problem:
                    #   we enclose it in single quotes.
                    # * Embeds both single _and_ double quotes. This
                    #   can't happen naturally, but it can happen if
                    #   you modify an attribute value after parsing
                    #   the document. Now we have a bit of a
                    #   problem. We solve it by enclosing the
                    #   attribute in single quotes, and escaping any
                    #   embedded single quotes to XML entities.
                    if '"' in val:
                        # This can't happen naturally, but it can happen
                        # if you modify an attribute value after parsing.
                        if "'" in val:
                            val = val.replace('"', "&quot;")
                        else:
                            fmt = "%s='%s'"

                    # Optionally convert any HTML entities
                    if self.convertHTMLEntities:
                        val = re.sub("&(\w+);", self._convertEntities, val)

                    # Now we're okay w/r/t quotes. But the attribute
                    # value might also contain angle brackets, or
                    # ampersands that aren't part of entities. We need
                    # to escape those to XML entities too.
                    val = val.replace("<", "&lt;").replace(">", "&gt;")
                    val = self.BARE_AMPERSAND.sub("&amp;", val)

                                      
                attrs.append(fmt % (self.toEncoding(key, encoding),
                                    self.toEncoding(val, encoding)))
        close = ''
        closeTag = ''
        if self.isSelfClosing:
            close = ' /'
        else:
            closeTag = '</%s>' % encodedName

        indentTag, indentContents = 0, 0
        if prettyPrint:
            indentTag = indentLevel
            space = (' ' * (indentTag-1))
            indentContents = indentTag + 1
        contents = self.renderContents(encoding, prettyPrint, indentContents)
        if self.hidden:
            s = contents
        else:
            s = []
            attributeString = ''
            if attrs:
                attributeString = ' ' + ' '.join(attrs)            
            if prettyPrint:
                s.append(space)
            s.append('<%s%s%s>' % (encodedName, attributeString, close))
            if prettyPrint:
                s.append("\n")
            s.append(contents)
            if prettyPrint and contents and contents[-1] != "\n":
                s.append("\n")
            if prettyPrint and closeTag:
                s.append(space)
            s.append(closeTag)
            if prettyPrint and closeTag and self.nextSibling:
                s.append("\n")
            s = ''.join(s)
        return s

    def prettify(self, encoding=DEFAULT_OUTPUT_ENCODING):
        return self.__str__(encoding, True)

    def renderContents(self, encoding=DEFAULT_OUTPUT_ENCODING,
                       prettyPrint=False, indentLevel=0):
        """Renders the contents of this tag as a string in the given
        encoding. If encoding is None, returns a Unicode string.."""
        s=[]
        for c in self:
            text = None
            if isinstance(c, NavigableString):
                text = c.__str__(encoding)
            elif isinstance(c, Tag):
                s.append(c.__str__(encoding, prettyPrint, indentLevel))
            if text and prettyPrint:
                text = text.strip()              
            if text:
                if prettyPrint:
                    s.append(" " * (indentLevel-1))
                s.append(text)
                if prettyPrint:
                    s.append("\n")
        return ''.join(s)    

    #Soup methods

    def find(self, name=None, attrs={}, recursive=True, text=None,
             **kwargs):
        """Return only the first child of this Tag matching the given
        criteria."""
        r = None
        l = self.findAll(name, attrs, recursive, text, 1, **kwargs)
        if l:
            r = l[0]
        return r
    findChild = find

    def findAll(self, name=None, attrs={}, recursive=True, text=None,
                limit=None, **kwargs):
        """Extracts a list of Tag objects that match the given
        criteria.  You can specify the name of the Tag and any
        attributes you want the Tag to have.

        The value of a key-value pair in the 'attrs' map can be a
        string, a list of strings, a regular expression object, or a
        callable that takes a string and returns whether or not the
        string matches for some custom definition of 'matches'. The
        same is true of the tag name."""
        generator = self.recursiveChildGenerator
        if not recursive:
            generator = self.childGenerator
        return self._findAll(name, attrs, text, limit, generator, **kwargs)
    findChildren = findAll

    # Pre-3.x compatibility methods
    first = find
    fetch = findAll
    
    def fetchText(self, text=None, recursive=True, limit=None):
        return self.findAll(text=text, recursive=recursive, limit=limit)

    def firstText(self, text=None, recursive=True):
        return self.find(text=text, recursive=recursive)
    
    #Utility methods

    def append(self, tag):
        """Appends the given tag to the contents of this tag."""
        self.contents.append(tag)

    #Private methods

    def _getAttrMap(self):
        """Initializes a map representation of this tag's attributes,
        if not already initialized."""
        if not getattr(self, 'attrMap'):
            self.attrMap = {}
            for (key, value) in self.attrs:
                self.attrMap[key] = value 
        return self.attrMap

    #Generator methods
    def childGenerator(self):
        for i in range(0, len(self.contents)):
            yield self.contents[i]
        raise StopIteration
    
    def recursiveChildGenerator(self):
        stack = [(self, 0)]
        while stack:
            tag, start = stack.pop()
            if isinstance(tag, Tag):            
                for i in range(start, len(tag.contents)):
                    a = tag.contents[i]
                    yield a
                    if isinstance(a, Tag) and tag.contents:
                        if i < len(tag.contents) - 1:
                            stack.append((tag, i+1))
                        stack.append((a, 0))
                        break
        raise StopIteration

# Next, a couple classes to represent queries and their results.
class SoupStrainer:
    """Encapsulates a number of ways of matching a markup element (tag or
    text)."""

    def __init__(self, name=None, attrs={}, text=None, **kwargs):
        self.name = name
        if isString(attrs):
            kwargs['class'] = attrs
            attrs = None
        if kwargs:
            if attrs:
                attrs = attrs.copy()
                attrs.update(kwargs)
            else:
                attrs = kwargs
        self.attrs = attrs
        self.text = text

    def __str__(self):
        if self.text:
            return self.text
        else:
            return "%s|%s" % (self.name, self.attrs)
    
    def searchTag(self, markupName=None, markupAttrs={}):
        found = None
        markup = None
        if isinstance(markupName, Tag):
            markup = markupName
            markupAttrs = markup
        callFunctionWithTagData = callable(self.name) \
                                and not isinstance(markupName, Tag)

        if (not self.name) \
               or callFunctionWithTagData \
               or (markup and self._matches(markup, self.name)) \
               or (not markup and self._matches(markupName, self.name)):
            if callFunctionWithTagData:
                match = self.name(markupName, markupAttrs)
            else:
                match = True            
                markupAttrMap = None
                for attr, matchAgainst in self.attrs.items():
                    if not markupAttrMap:
                         if hasattr(markupAttrs, 'get'):
                            markupAttrMap = markupAttrs
                         else:
                            markupAttrMap = {}
                            for k,v in markupAttrs:
                                markupAttrMap[k] = v
                    attrValue = markupAttrMap.get(attr)
                    if not self._matches(attrValue, matchAgainst):
                        match = False
                        break
            if match:
                if markup:
                    found = markup
                else:
                    found = markupName
        return found

    def search(self, markup):
        #print 'looking for %s in %s' % (self, markup)
        found = None
        # If given a list of items, scan it for a text element that
        # matches.        
        if isList(markup) and not isinstance(markup, Tag):
            for element in markup:
                if isinstance(element, NavigableString) \
                       and self.search(element):
                    found = element
                    break
        # If it's a Tag, make sure its name or attributes match.
        # Don't bother with Tags if we're searching for text.
        elif isinstance(markup, Tag):
            if not self.text:
                found = self.searchTag(markup)
        # If it's text, make sure the text matches.
        elif isinstance(markup, NavigableString) or \
                 isString(markup):
            if self._matches(markup, self.text):
                found = markup
        else:
            raise Exception, "I don't know how to match against a %s" \
                  % markup.__class__
        return found
        
    def _matches(self, markup, matchAgainst):    
        #print "Matching %s against %s" % (markup, matchAgainst)
        result = False
        if matchAgainst == True and type(matchAgainst) == types.BooleanType:
            result = markup != None
        elif callable(matchAgainst):
            result = matchAgainst(markup)
        else:
            #Custom match methods take the tag as an argument, but all
            #other ways of matching match the tag name as a string.
            if isinstance(markup, Tag):
                markup = markup.name
            if markup and not isString(markup):
                markup = unicode(markup)
            #Now we know that chunk is either a string, or None.
            if hasattr(matchAgainst, 'match'):
                # It's a regexp object.
                result = markup and matchAgainst.search(markup)
            elif isList(matchAgainst):
                result = markup in matchAgainst
            elif hasattr(matchAgainst, 'items'):
                result = markup.has_key(matchAgainst)
            elif matchAgainst and isString(markup):
                if isinstance(markup, unicode):
                    matchAgainst = unicode(matchAgainst)
                else:
                    matchAgainst = str(matchAgainst)

            if not result:
                result = matchAgainst == markup
        return result

class ResultSet(list):
    """A ResultSet is just a list that keeps track of the SoupStrainer
    that created it."""
    def __init__(self, source):
        list.__init__([])
        self.source = source

# Now, some helper functions.

def isList(l):
    """Convenience method that works with all 2.x versions of Python
    to determine whether or not something is listlike."""
    return hasattr(l, '__iter__') \
           or (type(l) in (types.ListType, types.TupleType))

def isString(s):
    """Convenience method that works with all 2.x versions of Python
    to determine whether or not something is stringlike."""
    try:
        return isinstance(s, unicode) or isintance(s, basestring) 
    except NameError:
        return isinstance(s, str)

def buildTagMap(default, *args):
    """Turns a list of maps, lists, or scalars into a single map.
    Used to build the SELF_CLOSING_TAGS, NESTABLE_TAGS, and
    NESTING_RESET_TAGS maps out of lists and partial maps."""
    built = {}
    for portion in args:
        if hasattr(portion, 'items'):
            #It's a map. Merge it.
            for k,v in portion.items():
                built[k] = v
        elif isList(portion):
            #It's a list. Map each item to the default.
            for k in portion:
                built[k] = default
        else:
            #It's a scalar. Map it to the default.
            built[portion] = default
    return built

# Now, the parser classes.

class BeautifulStoneSoup(Tag, SGMLParser):

    """This class contains the basic parser and search code. It defines
    a parser that knows nothing about tag behavior except for the
    following:
   
      You can't close a tag without closing all the tags it encloses.
      That is, "<foo><bar></foo>" actually means
      "<foo><bar></bar></foo>".

    [Another possible explanation is "<foo><bar /></foo>", but since
    this class defines no SELF_CLOSING_TAGS, it will never use that
    explanation.]

    This class is useful for parsing XML or made-up markup languages,
    or when BeautifulSoup makes an assumption counter to what you were
    expecting."""

    SELF_CLOSING_TAGS = {}
    NESTABLE_TAGS = {}
    RESET_NESTING_TAGS = {}
    QUOTE_TAGS = {}

    MARKUP_MASSAGE = [(re.compile('(<[^<>]*)/>'),
                       lambda x: x.group(1) + ' />'),
                      (re.compile('<!\s+([^<>]*)>'),
                       lambda x: '<!' + x.group(1) + '>')
                      ]

    ROOT_TAG_NAME = u'[document]'

    HTML_ENTITIES = "html"
    XML_ENTITIES = "xml"
    ALL_ENTITIES = [HTML_ENTITIES, XML_ENTITIES]

    def __init__(self, markup="", parseOnlyThese=None, fromEncoding=None,
                 markupMassage=True, smartQuotesTo=XML_ENTITIES,
                 convertEntities=None, selfClosingTags=None):
        """The Soup object is initialized as the 'root tag', and the
        provided markup (which can be a string or a file-like object)
        is fed into the underlying parser. 

        sgmllib will process most bad HTML, and the BeautifulSoup
        class has some tricks for dealing with some HTML that kills
        sgmllib, but Beautiful Soup can nonetheless choke or lose data
        if your data uses self-closing tags or declarations
        incorrectly.

        By default, Beautiful Soup uses regexes to sanitize input,
        avoiding the vast majority of these problems. If the problems
        don't apply to you, pass in False for markupMassage, and
        you'll get better performance.

        The default parser massage techniques fix the two most common
        instances of invalid HTML that choke sgmllib:

         <br/> (No space between name of closing tag and tag close)
         <! --Comment--> (Extraneous whitespace in declaration)

        You can pass in a custom list of (RE object, replace method)
        tuples to get Beautiful Soup to scrub your input the way you
        want."""

        self.parseOnlyThese = parseOnlyThese
        self.fromEncoding = fromEncoding
        self.smartQuotesTo = smartQuotesTo

        if convertEntities:
            # It doesn't make sense to convert encoded characters to
            # entities even while you're converting entities to Unicode.
            # Just convert it all to Unicode.
            self.smartQuotesTo = None

        if isList(convertEntities):
            self.convertHTMLEntities = self.HTML_ENTITIES in convertEntities
            self.convertXMLEntities = self.XML_ENTITIES in convertEntities
        else:
            self.convertHTMLEntities = self.HTML_ENTITIES == convertEntities
            self.convertXMLEntities = self.XML_ENTITIES == convertEntities

        self.instanceSelfClosingTags = buildTagMap(None, selfClosingTags)
        SGMLParser.__init__(self)
            
        if hasattr(markup, 'read'):        # It's a file-type object.
            markup = markup.read()
        self.markup = markup
        self.markupMassage = markupMassage
        try:
            self._feed()
        except StopParsing:
            pass
        self.markup = None                 # The markup can now be GCed

    def _feed(self, inDocumentEncoding=None):
        # Convert the document to Unicode.
        markup = self.markup
        if isinstance(markup, unicode):
            if not hasattr(self, 'originalEncoding'):
                self.originalEncoding = None
        else:
            dammit = UnicodeDammit\
                     (markup, [self.fromEncoding, inDocumentEncoding],
                      smartQuotesTo=self.smartQuotesTo)
            markup = dammit.unicode
            self.originalEncoding = dammit.originalEncoding
        if markup:
            if self.markupMassage:
                if not isList(self.markupMassage):
                    self.markupMassage = self.MARKUP_MASSAGE            
                for fix, m in self.markupMassage:
                    markup = fix.sub(m, markup)
        self.reset()

        SGMLParser.feed(self, markup or "")
        SGMLParser.close(self)
        # Close out any unfinished strings and close all the open tags.
        self.endData()
        while self.currentTag.name != self.ROOT_TAG_NAME:
            self.popTag()

    def __getattr__(self, methodName):
        """This method routes method call requests to either the SGMLParser
        superclass or the Tag superclass, depending on the method name."""
        #print "__getattr__ called on %s.%s" % (self.__class__, methodName)

        if methodName.find('start_') == 0 or methodName.find('end_') == 0 \
               or methodName.find('do_') == 0:
            return SGMLParser.__getattr__(self, methodName)
        elif methodName.find('__') != 0:
            return Tag.__getattr__(self, methodName)
        else:
            raise AttributeError

    def isSelfClosingTag(self, name):
        """Returns true iff the given string is the name of a
        self-closing tag according to this parser."""
        return self.SELF_CLOSING_TAGS.has_key(name) \
               or self.instanceSelfClosingTags.has_key(name)
            
    def reset(self):
        Tag.__init__(self, self, self.ROOT_TAG_NAME)
        self.hidden = 1
        SGMLParser.reset(self)
        self.currentData = []
        self.currentTag = None
        self.tagStack = []
        self.quoteStack = []
        self.pushTag(self)
    
    def popTag(self):
        tag = self.tagStack.pop()
        # Tags with just one string-owning child get the child as a
        # 'string' property, so that soup.tag.string is shorthand for
        # soup.tag.contents[0]
        if len(self.currentTag.contents) == 1 and \
           isinstance(self.currentTag.contents[0], NavigableString):
            self.currentTag.string = self.currentTag.contents[0]

        #print "Pop", tag.name
        if self.tagStack:
            self.currentTag = self.tagStack[-1]
        return self.currentTag

    def pushTag(self, tag):
        #print "Push", tag.name
        if self.currentTag:
            self.currentTag.append(tag)
        self.tagStack.append(tag)
        self.currentTag = self.tagStack[-1]

    def endData(self, containerClass=NavigableString):
        if self.currentData:
            currentData = ''.join(self.currentData)
            if currentData.endswith('<') and self.convertHTMLEntities:
                currentData = currentData[:-1] + '&lt;'
            if not currentData.strip():
                if '\n' in currentData:
                    currentData = '\n'
                else:
                    currentData = ' '
            self.currentData = []
            if self.parseOnlyThese and len(self.tagStack) <= 1 and \
                   (not self.parseOnlyThese.text or \
                    not self.parseOnlyThese.search(currentData)):
                return
            o = containerClass(currentData)
            o.setup(self.currentTag, self.previous)
            if self.previous:
                self.previous.next = o
            self.previous = o
            self.currentTag.contents.append(o)


    def _popToTag(self, name, inclusivePop=True):
        """Pops the tag stack up to and including the most recent
        instance of the given tag. If inclusivePop is false, pops the tag
        stack up to but *not* including the most recent instqance of
        the given tag."""
        #print "Popping to %s" % name
        if name == self.ROOT_TAG_NAME:
            return            

        numPops = 0
        mostRecentTag = None
        for i in range(len(self.tagStack)-1, 0, -1):
            if name == self.tagStack[i].name:
                numPops = len(self.tagStack)-i
                break
        if not inclusivePop:
            numPops = numPops - 1

        for i in range(0, numPops):
            mostRecentTag = self.popTag()
        return mostRecentTag    

    def _smartPop(self, name):

        """We need to pop up to the previous tag of this type, unless
        one of this tag's nesting reset triggers comes between this
        tag and the previous tag of this type, OR unless this tag is a
        generic nesting trigger and another generic nesting trigger
        comes between this tag and the previous tag of this type.

        Examples:
         <p>Foo<b>Bar<p> should pop to 'p', not 'b'.
         <p>Foo<table>Bar<p> should pop to 'table', not 'p'.
         <p>Foo<table><tr>Bar<p> should pop to 'tr', not 'p'.
         <p>Foo<b>Bar<p> should pop to 'p', not 'b'.

         <li><ul><li> *<li>* should pop to 'ul', not the first 'li'.
         <tr><table><tr> *<tr>* should pop to 'table', not the first 'tr'
         <td><tr><td> *<td>* should pop to 'tr', not the first 'td'
        """

        nestingResetTriggers = self.NESTABLE_TAGS.get(name)
        isNestable = nestingResetTriggers != None
        isResetNesting = self.RESET_NESTING_TAGS.has_key(name)
        popTo = None
        inclusive = True
        for i in range(len(self.tagStack)-1, 0, -1):
            p = self.tagStack[i]
            if (not p or p.name == name) and not isNestable:
                #Non-nestable tags get popped to the top or to their
                #last occurance.
                popTo = name
                break
            if (nestingResetTriggers != None
                and p.name in nestingResetTriggers) \
                or (nestingResetTriggers == None and isResetNesting
                    and self.RESET_NESTING_TAGS.has_key(p.name)):
                
                #If we encounter one of the nesting reset triggers
                #peculiar to this tag, or we encounter another tag
                #that causes nesting to reset, pop up to but not
                #including that tag.
                popTo = p.name
                inclusive = False
                break
            p = p.parent
        if popTo:
            self._popToTag(popTo, inclusive)

    def unknown_starttag(self, name, attrs, selfClosing=0):
        #print "Start tag %s: %s" % (name, attrs)
        if self.quoteStack:
            #This is not a real tag.
            #print "<%s> is not real!" % name
            attrs = ''.join(map(lambda(x, y): ' %s="%s"' % (x, y), attrs))
            self.currentData.append('<%s%s>' % (name, attrs))
            return        
        self.endData()

        if not self.isSelfClosingTag(name) and not selfClosing:
            self._smartPop(name)

        if self.parseOnlyThese and len(self.tagStack) <= 1 \
               and (self.parseOnlyThese.text or not self.parseOnlyThese.searchTag(name, attrs)):
            return

        tag = Tag(self, name, attrs, self.currentTag, self.previous)
        if self.previous:
            self.previous.next = tag
        self.previous = tag
        self.pushTag(tag)
        if selfClosing or self.isSelfClosingTag(name):
            self.popTag()                
        if name in self.QUOTE_TAGS:
            #print "Beginning quote (%s)" % name
            self.quoteStack.append(name)
            self.literal = 1
        return tag

    def unknown_endtag(self, name):
        #print "End tag %s" % name
        if self.quoteStack and self.quoteStack[-1] != name:
            #This is not a real end tag.
            #print "</%s> is not real!" % name
            self.currentData.append('</%s>' % name)
            return
        self.endData()
        self._popToTag(name)
        if self.quoteStack and self.quoteStack[-1] == name:
            self.quoteStack.pop()
            self.literal = (len(self.quoteStack) > 0)

    def handle_data(self, data):
        if self.convertHTMLEntities:
            if data[0] == '&':
                data = self.BARE_AMPERSAND.sub("&amp;",data)
            else:
                data = data.replace('&','&amp;') \
                           .replace('<','&lt;') \
                           .replace('>','&gt;')
        self.currentData.append(data)

    def _toStringSubclass(self, text, subclass):
        """Adds a certain piece of text to the tree as a NavigableString
        subclass."""
        self.endData()
        self.handle_data(text)
        self.endData(subclass)

    def handle_pi(self, text):
        """Handle a processing instruction as a ProcessingInstruction
        object, possibly one with a %SOUP-ENCODING% slot into which an
        encoding will be plugged later."""
        if text[:3] == "xml":
            text = "xml version='1.0' encoding='%SOUP-ENCODING%'"
        self._toStringSubclass(text, ProcessingInstruction)

    def handle_comment(self, text):
        "Handle comments as Comment objects."
        self._toStringSubclass(text, Comment)

    def handle_charref(self, ref):
        "Handle character references as data."
    Â