Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commitd2ba0c0

Browse files
committed
Implement generic (R)CDATA parsing algorithm
--HG--extra : convert_revision : svn%3Aacbfec75-9323-0410-a652-858a13e371e0/trunk%401146
1 parent55be4f1 commitd2ba0c0

File tree

1 file changed

+37
-50
lines changed

1 file changed

+37
-50
lines changed

‎src/html5lib/html5parser.py‎

Lines changed: 37 additions & 50 deletions
Original file line numberDiff line numberDiff line change
@@ -298,6 +298,27 @@ def startTagHtml(self, name, attributes):
298298
defprocessEndTag(self,name):
299299
self.endTagHandler[name](name)
300300

301+
defparseRCDataCData(self,name,attributes,contentType):
302+
"""Generic (R)CDATA Parsing algorithm
303+
contentType - RCDATA or CDATA
304+
"""
305+
assertcontentTypein ("CDATA","RCDATA")
306+
element=self.tree.insertElement(name,attributes)
307+
self.parser.tokenizer.contentModelFlag=contentModelFlags[contentType]
308+
fortokeninself.parser.tokenizer:
309+
iftoken["type"]in ("Characters","SpaceCharacters"):
310+
self.tree.insertText(token["data"])
311+
eliftoken["type"]=="ParseError":
312+
self.parser.parseError(token["data"],token.get("datavars", {}))
313+
else:
314+
assertself.parser.tokenizer.contentModelFlag==contentModelFlags["PCDATA"]
315+
asserttoken["type"]=="EndTag"andtoken["name"]==name,repr(token)
316+
assertself.tree.openElements.pop()==element
317+
return
318+
#Otherwise we hit EOF
319+
assertself.tree.openElements.pop()==element
320+
self.parser.parseError("expected-closing-tag-but-got-eof")
321+
301322

302323
classInitialPhase(Phase):
303324
# This phase deals with error handling as well which is currently not
@@ -549,10 +570,6 @@ def appendToHead(self, element):
549570

550571
# the real thing
551572
defprocessEOF (self):
552-
ifself.tree.openElements[-1].namein ("title","style","script","noscript"):
553-
self.parser.parseError("expected-named-closing-tag-but-got-eof",
554-
{"name":self.tree.openElements[-1].name})
555-
self.tree.openElements.pop()
556573
self.anythingElse()
557574
self.parser.phase.processEOF()
558575

@@ -568,44 +585,18 @@ def startTagHead(self, name, attributes):
568585
self.parser.parseError("two-heads-are-not-better-than-one")
569586

570587
defstartTagTitle(self,name,attributes):
571-
ifself.tree.headPointerisnotNoneandself.parser.phase==self.parser.phases["inHead"]:
572-
element=self.tree.createElement(name,attributes)
573-
self.appendToHead(element)
574-
self.tree.openElements.append(element)
575-
else:
576-
self.tree.insertElement(name,attributes)
577-
self.parser.tokenizer.contentModelFlag=contentModelFlags["RCDATA"]
588+
self.parseRCDataCData(name,attributes,"RCDATA")
578589

579590
defstartTagStyle(self,name,attributes):
580-
ifself.tree.headPointerisnotNoneandself.parser.phase==self.parser.phases["inHead"]:
581-
element=self.tree.createElement(name,attributes)
582-
self.appendToHead(element)
583-
self.tree.openElements.append(element)
584-
else:
585-
self.tree.insertElement(name,attributes)
586-
self.parser.tokenizer.contentModelFlag=contentModelFlags["CDATA"]
591+
self.parseRCDataCData(name,attributes,"CDATA")
587592

588593
defstartTagNoScript(self,name,attributes):
589-
# XXX Need to decide whether to implement the scripting disabled case.
590-
ifself.tree.headPointerisnotNoneandself.parser.phase==self.parser.phases["inHead"]:
591-
element=self.tree.createElement(name,attributes)
592-
self.appendToHead(element)
593-
self.tree.openElements.append(element)
594-
else:
595-
self.tree.insertElement(name,attributes)
596-
self.parser.tokenizer.contentModelFlag=contentModelFlags["CDATA"]
594+
#Need to decide whether to implement the scripting-disabled case
595+
self.parseRCDataCData(name,attributes,"CDATA")
597596

598597
defstartTagScript(self,name,attributes):
599-
#XXX Inner HTML case may be wrong
600-
element=None
601-
if (self.tree.headPointerisnotNoneandself.parser.phase==self.parser.phases["inHead"]):
602-
element=self.tree.createElement(name,attributes)
603-
self.appendToHead(element)
604-
self.tree.openElements.append(element)
605-
else:
606-
element=self.tree.insertElement(name,attributes)
607-
element._flags.append("parser-inserted")
608-
self.parser.tokenizer.contentModelFlag=contentModelFlags["CDATA"]
598+
#I think this is equivalent to the CDATA stuff since we don't execute script
599+
self.parseRCDataCData(name,attributes,"CDATA")
609600

610601
defstartTagBaseLinkMeta(self,name,attributes):
611602
if (self.tree.headPointerisnotNoneandself.parser.phase==self.parser.phases["inHead"]):
@@ -620,10 +611,8 @@ def startTagOther(self, name, attributes):
620611
self.parser.phase.processStartTag(name,attributes)
621612

622613
defendTagHead(self,name):
623-
ifself.tree.openElements[-1].name=="head":
624-
self.tree.openElements.pop()
625-
else:
626-
self.parser.parseError("unexpected-end-tag", {"name":"head"})
614+
assertself.tree.openElements[-1].name=="head"
615+
self.tree.openElements.pop()
627616
self.parser.phase=self.parser.phases["afterHead"]
628617

629618
defendTagImplyAfterHead(self,name):
@@ -640,10 +629,8 @@ def endTagOther(self, name):
640629
self.parser.parseError("unexpected-end-tag", {"name":name})
641630

642631
defanythingElse(self):
643-
ifself.tree.openElements[-1].name=="head":
644-
self.endTagHead("head")
645-
else:
646-
self.parser.phase=self.parser.phases["afterHead"]
632+
self.endTagHead("head")
633+
647634

648635
# XXX If we implement a parser for which scripting is disabled we need to
649636
# implement this phase.
@@ -682,8 +669,10 @@ def startTagFrameset(self, name, attributes):
682669
defstartTagFromHead(self,name,attributes):
683670
self.parser.parseError("unexpected-start-tag-out-of-my-head",
684671
{"name":name})
685-
self.parser.phase=self.parser.phases["inHead"]
686-
self.parser.phase.processStartTag(name,attributes)
672+
self.tree.openElements.append(self.tree.headPointer)
673+
self.parser.phases["inHead"].processStartTag(name,attributes)
674+
node=self.tree.openElements.pop()
675+
assertnodeisself.tree.headPointer,"Node name is %s, expected head"%node.name
687676

688677
defstartTagOther(self,name,attributes):
689678
self.anythingElse()
@@ -932,8 +921,7 @@ def startTagAppletMarqueeObject(self, name, attributes):
932921

933922
defstartTagXmp(self,name,attributes):
934923
self.tree.reconstructActiveFormattingElements()
935-
self.tree.insertElement(name,attributes)
936-
self.parser.tokenizer.contentModelFlag=contentModelFlags["CDATA"]
924+
self.parseRCDataCData(name,attributes,"CDATA")
937925

938926
defstartTagTable(self,name,attributes):
939927
ifself.tree.elementInScope("p"):
@@ -993,8 +981,7 @@ def startTagTextarea(self, name, attributes):
993981

994982
defstartTagCdata(self,name,attributes):
995983
"""iframe, noembed noframes, noscript(if scripting enabled)"""
996-
self.tree.insertElement(name,attributes)
997-
self.parser.tokenizer.contentModelFlag=contentModelFlags["CDATA"]
984+
self.parseRCDataCData(name,attributes,"CDATA")
998985

999986
defstartTagSelect(self,name,attributes):
1000987
self.tree.reconstructActiveFormattingElements()

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp