def _postAnalyzeImpl(self):
self.trace('Analysis complete. Beginning processing...')
#-------------------------------------------------------------------------------------------
# CLEANUP HTML WHITESPACE
# For html already in the page, either explicit or created by the preprocessor remove
# the newlines between tags to prevent them from being viewed as line breaks during
# Markup rendering.
self._result = self._removeHTMLTagWhitespace(self._raw.replace(u'\r', u' '))
self.trace('CLEANED -> preliminary HTML whitespace', showResult=True)
tags = self._createTags()
#-------------------------------------------------------------------------------------------
# REMOVE UNWANTED LINE BREAKS
# Newlines that occur between close and open Markup tags should be removed.
res = MarkupProcessor._STRIP_WHITESPACE_PATTERN.finditer(self._result)
if res:
for r in res:
start = r.start()
end = r.end()
length = end - start
self._result = self._result[:start] + u' '*length + self._result[end:]
res = MarkupProcessor._CONTINUE_LINE_PATTERN.finditer(self._result)
if res:
for r in res:
start = r.start()
end = r.end()
length = end - start
replace = u' '*length
self._result = self._result[:start] + replace + self._result[end:]
res = MarkupProcessor._UNWANTED_NEWLINE_PATTERN.finditer(self._result)
if res:
for r in res:
start = r.start()
end = r.end()
# Skip line breaks inside of leaf tags.
skip = False
startTag = None
endTag = None
for t in tags:
if not startTag:
c = t.findContainingBlock(start)
if c and c.isLeafTag:
skip = True
break
startTag = c
if not endTag:
endTag = t.findContainingBlock(end)
if startTag and endTag:
break
if skip:
continue
# Skip tags that are not display:block or voids
skip = (startTag and not (startTag.isBlockDisplay or startTag.isVoidTag)) or \
(endTag and not (endTag.isBlockDisplay or endTag.isVoidTag))
if skip:
continue
replace = r.group('close') + r.group('space').replace(u'\n', u' ') + r.group('open')
self._result = self._result[:start] + replace + self._result[end:]
self.trace('Line breaks cleaned up; Starting Tag rendering...')
#-------------------------------------------------------------------------------------------
# RENDER RESULT
for b in self._blocks:
if b.blockType == BlockSyntaxEnum.COMMENT:
start = b.start
while start - 1 > 0 and self._result[start-1] in [u'\r', u'\n', u'\t']:
start -= 1
end = b.end
while end < len(self._result) and self._result[end] in [u'\r', u'\n', u'\t']:
end += 1
length = end - start
if length > 0:
self._result = self._result[:start] + u' '*length + self._result[end:]
#-------------------------------------------------------------------------------------------
# EXECUTE THE MULTI-STAGE RENDERING PROCESS
self._executeTagActions(tags, 'redefinitionTraversal', 'Redefinition Traversal Failure')
self._executeTagActions(tags, 'redefinitionReversal', 'Redefinition Reversal Failure')
self._executeTagActions(tags, 'connectivityTraversal', 'Connectivity Traversal Failure')
self._executeTagActions(tags, 'connectivityReversal', 'Connectivity Reversal Failure')
self.trace(self.traceHierarchy('Global pre-render complete:', tags))
self._executeTagActions(tags, 'render', 'Render Failure')
self._executeTagActions(tags, 'cleanupTraversal', 'Cleanup Traversal Failure')
self._executeTagActions(tags, 'cleanupReversal', 'Cleanup Render Failure')
self.trace('RENDERED RESULT -> Markup converted to HTML', showResult=True)
#-------------------------------------------------------------------------------------------
# POST-PROCESSING HTML WHITESPACE CLEANUP
self._result = self._removeHTMLTagWhitespace(self._result.replace(u'\r', u' '))
self.trace('CLEANED -> post HTML whitespace', showResult=True)
#-------------------------------------------------------------------------------------------
# STRIP MARKUP
res = MarkupProcessor._STRIP_PATTERN.finditer(self._result)
offset = 0
for r in res:
s = r.start() + offset
e = r.end() + offset
offset += self.insertCharacters(s, e, u'')
self.trace('STRIPPED -> markup', showResult=True)
#-------------------------------------------------------------------------------------------
# HANDLE NEWLINES NOT IN TAGS
offset = 0
tagIndex = 0
preserveBlocks = DomUtils.getPreserveBlocks(self._result)
while True:
index = self._result.find('\n', offset)
if index == -1:
break
# Ignore line breaks inside preserved tags
skip = False
for b in preserveBlocks:
if b.start <= index <= b.end:
offset = b.end + 1
skip = True
break
if skip:
continue
while tagIndex < len(tags):
if index > tags[tagIndex].end():
tagIndex += 1
continue
break
if tagIndex < len(tags) and tags[tagIndex].contains(index):
offset = tags[tagIndex].end()
tagIndex += 1
continue
offset = index + self.insertCharacters(index, index+1, u'<br />')
self.trace('ADDED breaks to newlines not in tags', showResult=True)
#-------------------------------------------------------------------------------------------
# REMOVE BR DIV COMBOS
res = MarkupProcessor._DIV_LINE_BREAK_PATTERN.finditer(self._result)
offset = 0
for r in res:
s = r.start() + offset
e = r.end() + offset
offset += self.insertCharacters(s, e, u'')
self._result = self._result.replace(u'[*', u'[').replace(u'*]', u']')
self._tags = tags
#-------------------------------------------------------------------------------------------
# FINAL MODIFICATIONS TO THE RENDERED DOM
self._result = self._modifyResult(self._result)
#-------------------------------------------------------------------------------------------
# CLEANUP WHITE SPACE
self._result = DomUtils.minifyDom(self._result)
self.trace('Post processing complete.')
def _removeHTMLTagWhitespace(self, source):
preserveBlocks = DomUtils.getPreserveBlocks(source)
res = MarkupProcessor._HTML_INSIDE_TAG_PATTERN.finditer(source)
if res:
for r in res:
start = r.start()
end = r.end()
replace = r.group('tag').replace(u'\n', u' ')
source = source[:start] + replace + source[end:]
res = MarkupProcessor._HTML_PRE_TAG_WHITESPACE_PATTERN.finditer(source)
if res:
for r in res:
start = r.start()
end = r.end()
tag = r.group('tag')
preSource = source[:start]
if StringUtils.begins(tag, (u'<span', u'<a')):
strippedPreSource = preSource.strip()
# Preserve lines between span tags
if StringUtils.ends(strippedPreSource, (u'</span>', u'</a>')):
continue
# Preserve lines between span tags and non-html entities like text
if not strippedPreSource.endswith(u'>'):
continue
skip = False
for b in preserveBlocks:
if b.start <= start <= b.end or b.start <= end <= b.end:
skip = True
break
if skip:
continue
length = len(r.group('whitespace'))
replace = u' '*length + tag
source = preSource + replace + source[end:]
res = MarkupProcessor._HTML_POST_TAG_WHITESPACE_PATTERN.finditer(source)
if res:
for r in res:
start = r.start()
end = r.end()
tag = r.group('tag')
postSource = source[end:]
if tag in (u'</span>', u'</a>'):
strippedPostSource = postSource.strip()
# Preserve lines between span tags
if StringUtils.begins(strippedPostSource, (u'<span', u'<a')):
continue
# Preserve lines between span tags and non-html entities like text
if not strippedPostSource.startswith(u'<'):
continue
skip = False
for b in preserveBlocks:
if b.start <= start <= b.end or b.start <= end <= b.end:
skip = True
break
if skip:
continue
length = len(r.group('whitespace'))
replace = tag + u' '*length
source = source[:start] + replace + postSource
return source