diff options
| -rw-r--r-- | beautifulsoup/__init__.py | 8 | ||||
| -rw-r--r-- | beautifulsoup/builder/__init__.py | 31 | ||||
| -rw-r--r-- | beautifulsoup/builder/lxml_builder.py | 6 | ||||
| -rw-r--r-- | beautifulsoup/element.py | 23 | ||||
| -rw-r--r-- | tests/test_lxml.py | 13 |
5 files changed, 56 insertions, 25 deletions
diff --git a/beautifulsoup/__init__.py b/beautifulsoup/__init__.py index 922005c..ca32589 100644 --- a/beautifulsoup/__init__.py +++ b/beautifulsoup/__init__.py @@ -106,11 +106,9 @@ class BeautifulSoup(Tag): able to build a tree using 'start tag' events, 'end tag' events, 'data' events, and "done with data" events. - If you encounter a self-closing tag, call handle_starttag and then - handle_endtag, but note that the tag will not be displayed as a - self-closing tag unless you also have your builder's - isSelfClosingTag() implementation return True when passed the tag - name. + If you encounter an empty-element tag (aka a self-closing tag, + like HTML's <br> tag), call handle_starttag and then + handle_endtag. """ ROOT_TAG_NAME = u'[document]' diff --git a/beautifulsoup/builder/__init__.py b/beautifulsoup/builder/__init__.py index 5c275d7..37e9c8a 100644 --- a/beautifulsoup/builder/__init__.py +++ b/beautifulsoup/builder/__init__.py @@ -12,16 +12,37 @@ class TreeBuilder(Entities): """Turn a document into a Beautiful Soup object tree.""" assume_html = False + preserve_whitespace_tags = set() + empty_element_tags = None # A tag will be considered an empty-element + # tag when and only when it has no contents. def __init__(self): self.soup = None - def isSelfClosingTag(self, name): - return name in self.self_closing_tags - def reset(self): pass + def can_be_empty_element(self, tag_name): + """Might a tag with this name be an empty-element tag? + + The final markup may or may not actually present this tag as + self-closing. + + For instance: an HTML builder does not consider a <p> tag to + be an empty-element tag (it's not in empty_element_tags). This + means an empty <p> tag will be presented as "<p></p>", not + "<p />". + + The default builder has no opinion about which tags are + empty-element tags, so a tag will be presented as an + empty-element tag if and only if it has no contents. + "<foo></foo>" will become "<foo />", and "<foo>bar</foo>" will + be left alone. + """ + if self.empty_element_tags is None: + return True + return tag_name in self.empty_element_tags + def feed(self, markup): raise NotImplementedError() @@ -101,8 +122,8 @@ class HTMLTreeBuilder(TreeBuilder): assume_html = True preserve_whitespace_tags = set(['pre', 'textarea']) - self_closing_tags = set(['br' , 'hr', 'input', 'img', 'meta', - 'spacer', 'link', 'frame', 'base']) + empty_element_tags = set(['br' , 'hr', 'input', 'img', 'meta', + 'spacer', 'link', 'frame', 'base']) # Used by set_up_substitutions to detect the charset in a META tag CHARSET_RE = re.compile("((^|;)\s*charset=)([^;]*)", re.M) diff --git a/beautifulsoup/builder/lxml_builder.py b/beautifulsoup/builder/lxml_builder.py index 5e053b1..0cc9e51 100644 --- a/beautifulsoup/builder/lxml_builder.py +++ b/beautifulsoup/builder/lxml_builder.py @@ -7,9 +7,6 @@ import types class LXMLTreeBuilderForXML(TreeBuilder): DEFAULT_PARSER_CLASS = etree.XMLParser - preserve_whitespace_tags = set() - self_closing_tags = set() - @property def default_parser(self): # This can either return a parser object or a class, which @@ -55,9 +52,6 @@ class LXMLTreeBuilderForXML(TreeBuilder): def end(self, name): self.soup.endData() completed_tag = self.soup.tagStack[-1] - if len(completed_tag.contents) == 0: - completed_tag.isSelfClosing = True - self.soup.handle_endtag(name) def pi(self, target, data): diff --git a/beautifulsoup/element.py b/beautifulsoup/element.py index 5793d59..c0c28d4 100644 --- a/beautifulsoup/element.py +++ b/beautifulsoup/element.py @@ -429,7 +429,6 @@ class Tag(PageElement, Entities): # chunks be garbage-collected. self.parserClass = parser.__class__ self.name = name - self.isSelfClosing = builder.isSelfClosingTag(name) if attrs == None: attrs = [] if isinstance(attrs, types.DictType): @@ -447,6 +446,26 @@ class Tag(PageElement, Entities): # Set up any substitutions, such as the charset in a META tag. self.contains_substitutions = builder.set_up_substitutions(self) + self.can_be_empty_element = (builder.can_be_empty_element(name)) + + @property + def is_empty_element(self): + """Is this tag an empty-element tag? (aka a self-closing tag) + + A tag that has contents is never an empty-element tag. + + A tag that has no contents may or may not be an empty-element + tag. It depends on the builder used to create the tag. If the + builder has a designated list of empty-element tags, then only + a tag whose name shows up in that list is considered an + empty-element tag. + + If the builder has no designated list of empty-element tags, + then any tag with no contents is an empty-element tag. + """ + return len(self.contents) == 0 and self.can_be_empty_element + isSelfClosing = is_empty_element # BS3 + @property def string(self): @@ -624,7 +643,7 @@ class Tag(PageElement, Entities): attrs.append(decoded) close = '' closeTag = '' - if self.isSelfClosing: + if self.is_empty_element: close = ' /' else: closeTag = '</%s>' % self.name diff --git a/tests/test_lxml.py b/tests/test_lxml.py index dd1c363..602fe05 100644 --- a/tests/test_lxml.py +++ b/tests/test_lxml.py @@ -492,15 +492,14 @@ class TestLXMLXMLBuilder(SoupTest): def default_builder(self): return LXMLTreeBuilderForXML() - def test_self_closing_tag(self): + def test_empty_element_tag(self): soup = self.soup("<p><iamselfclosing /></p>") - self.assertTrue(soup.iamselfclosing.isSelfClosing) + self.assertTrue(soup.iamselfclosing.is_empty_element) - def test_self_empty_tag_treated_as_self_closing(self): + def test_self_empty_tag_treated_as_empty_element(self): soup = self.soup("<p><iamclosed></iamclosed></p>") - self.assertTrue(soup.iamclosed.isSelfClosing) + self.assertTrue(soup.iamclosed.is_empty_element) - def test_self_nonempty_tag_is_not_self_closing(self): + def test_self_nonempty_tag_is_not_empty_element(self): soup = self.soup("<p><ihavecontents>contents</ihavecontents></p>") - self.assertFalse(soup.ihavecontents.isSelfClosing) - + self.assertFalse(soup.ihavecontents.is_empty_element) |