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

Commit0c8d7bc

Browse files
committed
Fixhtml5lib#6: dom2sax crash by replacing dom2sax with a generic to_sax
This moves the functionality to a new treeadapters module (wherelater the adapters from test_treewalker.py will get moved) andremoves the previous dom2sax function.
1 parentf61e328 commit0c8d7bc

File tree

6 files changed

+69
-78
lines changed

6 files changed

+69
-78
lines changed

‎CHANGES.rst‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -6,6 +6,10 @@ Change Log
66

77
Released on XXX, 2013
88

9+
* Removed ``dom2sax`` from DOM treebuilders. It has been replaced by
10+
``treeadapters.sax.to_sax`` which is generic and supports any
11+
treewalker; it also resolves all known bugs with ``dom2sax``.
12+
913

1014
1.0b1
1115
~~~~~

‎html5lib/constants.py‎

Lines changed: 18 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -433,6 +433,24 @@
433433
(namespaces["mathml"],"mtext")
434434
))
435435

436+
adjustForeignAttributes= {
437+
"xlink:actuate": ("xlink","actuate",namespaces["xlink"]),
438+
"xlink:arcrole": ("xlink","arcrole",namespaces["xlink"]),
439+
"xlink:href": ("xlink","href",namespaces["xlink"]),
440+
"xlink:role": ("xlink","role",namespaces["xlink"]),
441+
"xlink:show": ("xlink","show",namespaces["xlink"]),
442+
"xlink:title": ("xlink","title",namespaces["xlink"]),
443+
"xlink:type": ("xlink","type",namespaces["xlink"]),
444+
"xml:base": ("xml","base",namespaces["xml"]),
445+
"xml:lang": ("xml","lang",namespaces["xml"]),
446+
"xml:space": ("xml","space",namespaces["xml"]),
447+
"xmlns": (None,"xmlns",namespaces["xmlns"]),
448+
"xmlns:xlink": ("xmlns","xlink",namespaces["xmlns"])
449+
}
450+
451+
unadjustForeignAttributes=dict([((ns,local),qname)forqname, (prefix,local,ns)in
452+
adjustForeignAttributes.items()])
453+
436454
spaceCharacters=frozenset((
437455
"\t",
438456
"\n",

‎html5lib/html5parser.py‎

Lines changed: 2 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -17,6 +17,7 @@
1717
from .constantsimportcdataElements,rcdataElements
1818
from .constantsimporttokenTypes,ReparseException,namespaces
1919
from .constantsimporthtmlIntegrationPointElements,mathmlTextIntegrationPointElements
20+
from .constantsimportadjustForeignAttributesasadjustForeignAttributesMap
2021

2122

2223
defparse(doc,treebuilder="etree",encoding=None,
@@ -333,20 +334,7 @@ def adjustSVGAttributes(self, token):
333334
deltoken["data"][originalName]
334335

335336
defadjustForeignAttributes(self,token):
336-
replacements= {
337-
"xlink:actuate": ("xlink","actuate",namespaces["xlink"]),
338-
"xlink:arcrole": ("xlink","arcrole",namespaces["xlink"]),
339-
"xlink:href": ("xlink","href",namespaces["xlink"]),
340-
"xlink:role": ("xlink","role",namespaces["xlink"]),
341-
"xlink:show": ("xlink","show",namespaces["xlink"]),
342-
"xlink:title": ("xlink","title",namespaces["xlink"]),
343-
"xlink:type": ("xlink","type",namespaces["xlink"]),
344-
"xml:base": ("xml","base",namespaces["xml"]),
345-
"xml:lang": ("xml","lang",namespaces["xml"]),
346-
"xml:space": ("xml","space",namespaces["xml"]),
347-
"xmlns": (None,"xmlns",namespaces["xmlns"]),
348-
"xmlns:xlink": ("xmlns","xlink",namespaces["xmlns"])
349-
}
337+
replacements=adjustForeignAttributesMap
350338

351339
fororiginalNameintoken["data"].keys():
352340
iforiginalNameinreplacements:

‎html5lib/treeadapters/__init__.py‎

Whitespace-only changes.

‎html5lib/treeadapters/sax.py‎

Lines changed: 44 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,44 @@
1+
from __future__importabsolute_import,division,unicode_literals
2+
3+
fromxml.sax.xmlreaderimportAttributesNSImpl
4+
5+
from ..constantsimportadjustForeignAttributes,unadjustForeignAttributes
6+
7+
prefix_mapping= {}
8+
forprefix,localName,namespaceinadjustForeignAttributes.values():
9+
ifprefixisnotNone:
10+
prefix_mapping[prefix]=namespace
11+
12+
13+
defto_sax(walker,handler):
14+
"""Call SAX-like content handler based on treewalker walker"""
15+
handler.startDocument()
16+
forprefix,namespaceinprefix_mapping.items():
17+
handler.startPrefixMapping(prefix,namespace)
18+
19+
fortokeninwalker:
20+
type=token["type"]
21+
iftype=="Doctype":
22+
continue
23+
eliftypein ("StartTag","EmptyTag"):
24+
attrs=AttributesNSImpl(token["data"],
25+
unadjustForeignAttributes)
26+
handler.startElementNS((token["namespace"],token["name"]),
27+
token["name"],
28+
attrs)
29+
iftype=="EmptyTag":
30+
handler.endElementNS((token["namespace"],token["name"]),
31+
token["name"])
32+
eliftype=="EndTag":
33+
handler.endElementNS((token["namespace"],token["name"]),
34+
token["name"])
35+
eliftypein ("Characters","SpaceCharacters"):
36+
handler.characters(token["data"])
37+
eliftype=="Comment":
38+
pass
39+
else:
40+
assertFalse,"Unknown token type"
41+
42+
forprefix,namespaceinprefix_mapping.items():
43+
handler.endPrefixMapping(prefix)
44+
handler.endDocument()

‎html5lib/treebuilders/dom.py‎

Lines changed: 1 addition & 64 deletions
Original file line numberDiff line numberDiff line change
@@ -1,7 +1,7 @@
11
from __future__importabsolute_import,division,unicode_literals
22

33

4-
fromxml.domimportminidom,Node,XML_NAMESPACE,XMLNS_NAMESPACE
4+
fromxml.domimportminidom,Node
55
importweakref
66

77
from .import_base
@@ -220,69 +220,6 @@ def serializeElement(element, indent=0):
220220

221221
return"\n".join(rv)
222222

223-
defdom2sax(node,handler,nsmap={'xml':XML_NAMESPACE}):
224-
ifnode.nodeType==Node.ELEMENT_NODE:
225-
ifnotnsmap:
226-
handler.startElement(node.nodeName,node.attributes)
227-
forchildinnode.childNodes:
228-
dom2sax(child,handler,nsmap)
229-
handler.endElement(node.nodeName)
230-
else:
231-
attributes=dict(node.attributes.itemsNS())
232-
233-
# gather namespace declarations
234-
prefixes= []
235-
forattrnameinlist(node.attributes.keys()):
236-
attr=node.getAttributeNode(attrname)
237-
if (attr.namespaceURI==XMLNS_NAMESPACEor
238-
(attr.namespaceURIisNoneandattr.nodeName.startswith('xmlns'))):
239-
prefix= (attr.nodeName!='xmlns'andattr.nodeNameorNone)
240-
handler.startPrefixMapping(prefix,attr.nodeValue)
241-
prefixes.append(prefix)
242-
nsmap=nsmap.copy()
243-
nsmap[prefix]=attr.nodeValue
244-
delattributes[(attr.namespaceURI,attr.nodeName)]
245-
246-
# apply namespace declarations
247-
forattrnameinlist(node.attributes.keys()):
248-
attr=node.getAttributeNode(attrname)
249-
ifattr.namespaceURIisNoneand':'inattr.nodeName:
250-
prefix=attr.nodeName.split(':')[0]
251-
ifprefixinnsmap:
252-
delattributes[(attr.namespaceURI,attr.nodeName)]
253-
attributes[(nsmap[prefix],attr.nodeName)]=attr.nodeValue
254-
255-
# SAX events
256-
ns=node.namespaceURIornsmap.get(None,None)
257-
handler.startElementNS((ns,node.nodeName),node.nodeName,attributes)
258-
forchildinnode.childNodes:
259-
dom2sax(child,handler,nsmap)
260-
handler.endElementNS((ns,node.nodeName),node.nodeName)
261-
forprefixinprefixes:
262-
handler.endPrefixMapping(prefix)
263-
264-
elifnode.nodeTypein [Node.TEXT_NODE,Node.CDATA_SECTION_NODE]:
265-
handler.characters(node.nodeValue)
266-
267-
elifnode.nodeType==Node.DOCUMENT_NODE:
268-
handler.startDocument()
269-
forchildinnode.childNodes:
270-
dom2sax(child,handler,nsmap)
271-
handler.endDocument()
272-
273-
elifnode.nodeType==Node.DOCUMENT_FRAGMENT_NODE:
274-
forchildinnode.childNodes:
275-
dom2sax(child,handler,nsmap)
276-
277-
else:
278-
# ATTRIBUTE_NODE
279-
# ENTITY_NODE
280-
# PROCESSING_INSTRUCTION_NODE
281-
# COMMENT_NODE
282-
# DOCUMENT_TYPE_NODE
283-
# NOTATION_NODE
284-
pass
285-
286223
returnlocals()
287224

288225

0 commit comments

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