xueqingwell / py-dom-xpath Goto Github PK

[Description]

Since XPathContext tries to retrieve the default namespace URI from the 
document elements' xmlns attribute, it doesn't work in the following situation 
where a) xmlns attribute is modified afterward b) the default namespace URI for 
the document without namespace specifications is implied by the DOM parser.

I was stuck on this in the case (b) because I have a DOM implementation other 
than minidom (actually backed by the WebKit DOM through PyObjC).

Patch attached.

[Code for reproduction]

from xml.dom.minidom import parseString
from xpath import XPathContext

doc = parseString('''
<foo xmlns="http://foo.example.com/">
  <item>foo</item>
  <item xmlns="http://bar.example.com/">bar</item>
</foo>
''')
doc.documentElement.setAttribute('xmlns', 'http://bar.example.com/')
assert doc.documentElement.namespaceURI == 'http://foo.example.com/'
assert doc.documentElement.getAttribute('xmlns') == 'http://bar.example.com/'
nodeset = XPathContext(doc).find('//item', doc)
assert len(nodeset) == 1
assert nodeset[0].namespaceURI == 'http://foo.example.com/' # boom!


[What version of the product are you using? On what operating system?]

SVN head

This project appears to be all but completely dead, but for the benefit of 
others:

The latest versions of pip will not install due to the following errors:

  py-dom-xpath an externally hosted file and may be unreliable
  py-dom-xpath is potentially insecure and unverifiable.

The installation can be forced with the following

pip install --allow-external py-dom-xpath --allow-unverified py-dom-xpath .

What steps will reproduce the problem?
1. Consider the document "<foo><text>hi</text></foo>"
2. The query "//text" will result in a syntax error, though it seems to be 
valid.

What is the expected output? What do you see instead?
A list of one element.

What version of the product are you using? On what operating system?
version 0.1 on openSUSE 11.3

Please provide any additional information below.
I have tested query on a couple of online websites (like this one: 
http://www.futurelab.ch/xmlkurs/xpath.en.html) and they work fine. It seems 
that the library doesn't let "text" be part of a query because it is a keyword.

What steps will reproduce the problem?
1. Find or create an XML file containing an element named 'text'.
2. Try to access that element's content via XPath via the findvalue() function
For example, I'm loading lexicon data from a LIFT file within the context of an 
'entry' element. This line fails:
lexeme = context.findvalue('lexical-unit/form/text/descendant-or-self::text()', 
entry)

xpath.exceptions.XPathParseError: Syntax error:
lexical-unit/form/text/descendant-or-self::text()
----------------------^

What is the expected output? What do you see instead?
I was hoping it could handle this kind of input file, since I think it's valid 
XML. But apparently text (even without parentheses) is some kind of reserved 
word, either in XPath or py-dom-xpath, I'm not sure.

What version of the product are you using? On what operating system?
Version 0.1, fron 2/22/2009, on WinXP SP3 under Python 2.7. 

Please provide any additional information below.
Is there an alternate syntax for this kind of thing? Or is there an easier 
workaround than what I've come up with? (I.e. making a copy of the source file, 
renaming all of its text elements to textt, and writing my XPaths to match 
that. Tedious.)

Python 2.6, py-dom-xpath from trunk


ERROR: test_string_number_nan (__main__.TestStringFunctions)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "functions.py", line 127, in test_string_number_nan
    result = xpath.find('string(0 div 0)', self.doc)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/__init__.py", 
line 21, in api_function
    return f(*args, **kwargs)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/__init__.py", 
line 150, in find
    return XPath.get(expr).find(node, **kwargs)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/__init__.py", 
line 21, in api_function
    return f(*args, **kwargs)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/__init__.py", 
line 113, in find
    return self.expr.evaluate(node, 1, 1, context)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/expr.py", line 
729, in evaluate
    result = self.steps[0].evaluate(node, pos, size, context)
  File "/home/user/.local/lib/python2.6/site-packages/xpath/expr.py", line 
368, in new_f
    args = [convert(x) for x in args]
  File "/home/user/.local/lib/python2.6/site-packages/xpath/expr.py", line 
104, in string
    elif int(v) == v and v <= 0xffffffff:
ValueError: cannot convert float NaN to integer

----------------------------------------------------------------------
Ran 77 tests in 0.118s

FAILED (errors=1)

What steps will reproduce the problem?
1. find any attribute with regex
2. using [fn:matches(@* , "^a*")]

What is the expected output? What do you see instead?
I would like to see any node have attribute start with "a" if it dont use 
start-with.
such as 
<BBB type="aaa"/>
<BBB type="abb"/>

What version of the product are you using? On what operating system?
version 0.1
Windows XP

Please provide any additional information below.

I have xml content like this
<AAA>
<BBB type="aaa"/>
<BBB type="abb"/>
<BBB type="baa"/>
</AAA>

removed parentheses from declaration of class XPath on line 80. an empty paren 
set doesn't work on Python 2.4, but everything else does. it would be too bad 
to not have backward compatibility just for that.

Does it really change anything?
[email protected]

What steps will reproduce the problem?
1. Get a dom object
2. Use a xpath which contains "node" to find a node.
xpath.find("//node[@name='xxx']", element)
3. Get an Exception:
*** XPathParseError: Syntax error:
//node
------^

Why I can't use a xpath contains "node"?

CDATA-Sections are lost in the string_value function:

>>> str = '<asdf><![CDATA[lalaal<hallo>lasdfasdf]]></asdf>'
>>> import xml.dom.minidom as minidom
>>> doc = minidom.parseString(str)
>>> import xpath
>>> from xpath.expr import string_value
>>> string_value(doc.firstChild.firstChild)
>>>
>>> string_value(doc)
u''
>>>



I expected:
u'lalaal<hallo>lasdfasdf'


What version of the product are you using? On what operating system?
Python 2.5, Windows 7


Attached is a patch that checks for CDATA-Sections and inserts their content 
into the string-value.

python2.7

import xpath
from xml.dom import minidom

xml = 
minidom.parseString("<Rerservations><Rerservation><endDate>2011-03-15T11:00:00+0
8:00</endDate></Rerservation></Rerservations>")
xpath.find(xml,"/Reservations/endDate")


Traceback (most recent call last):
  File "<pyshell#8>", line 1, in <module>
    xpath.find(xml,"/Reservations/Reservation/endDate")
  File "C:\Python27\lib\site-packages\xpath\__init__.py", line 23, in api_function
    raise e
XPathParseError: Syntax error:
<xml.dom.minidom.Document instance at 0x014A53C8>
^

What steps will reproduce the problem?
1. search for //AAA/BBB/CCC-DDD

What is the expected output? What do you see instead?
I'd expect to get the count of element "CCC-DDD", but it's not found.

What version of the product are you using? On what operating system?
Running version py-dom-xpath-0.1.tar.gz with Python 2.7.2 under Debian 4.1.1-21.

Please provide any additional information below.
I'm not sure if the "-" is interpreted as an operator. I also tried to escape 
it, but that didn't work either. I hope there's still someone watching this 
project from time to time, so I can get a reply ;)