Mailman 3 May 2005 - lxml - The Python XML Toolkit

[lxml-dev] Q: XSLT parameters?
by Paul Everitt 18 May '05

18 May '05

I had the impression that lxml supported passing parameters to transformations, but I can't find it documented nor in tests. Is this on the TODO list or already implemented? --Paul

2 3

[lxml-dev] lxml 0.6 released
by Martijn Faassen 15 May '05

15 May '05

Hi there, I've released lxml 0.6 just now, containing the fixes to a bunch of recent important bugs. Thanks everybody for the bug reporting! It can be found here: http://codespeak.net/lxml Regards, Martijn

2 2

[lxml-dev] XPath bug in lxml (or libxml2)
by Henrik Thostrup Jensen 14 May '05

14 May '05

Hi I was testing some xpath querying in lxml when I found a rather strange syndrom. When doing several queries against two or more trees I get an exception like this: Traceback (most recent call last): File "/home/htj/code/annis/trunk/resourcedb.py", line 122, in query result_node = doc.xpath(path) File "etree.pyx", line 214, in etree._ElementTree.xpath File "etree.pyx", line 944, in etree.XPathEvaluator.evaluate SyntaxError: Error in xpath expression. The strange thing is that this only happens during the second query run (and the queries are exactly the same), and if I have more than one tree to query. Otherwise it works fine. Ideas? I can send an example/testcase if that will help. Best regards, Henrik

3 6

[lxml-dev] Malloc error test case
by Paul Everitt 13 May '05

13 May '05

I just checked in a ridiculously weird test case in paul/ tostringmalloc. It shows some weird interaction that is very sensitive to order of operations and input documents. In the course of tracking this down, I found a weird idiosyncrasy in effbot's ElementTree API: text = lxml.etree.tostring(tree.getroot()) text = style.tostring(tree) The "tostring" method gets passed an Element when used as a module function but an ElementTree when used on as an XSLT method. Grrrr. --Paul

2 3

[lxml-dev] unicode issues with XSLT
by newsletter＠theflow.de 13 May '05

13 May '05

Hello, thanks for this cool library, I really like it :) I replaced the following libxml/libxslt function: def apply_xslt(input_str, stylesheet): """input_str is a string, stylesheet is a file""" doc = libxml2.parseDoc(input_str) styledoc = libxml2.parseFile(stylesheet) style = libxslt.parseStylesheetDoc(styledoc) result = style.applyStylesheet(doc, None) result_str = style.saveResultToString(result) style.freeStylesheet() doc.freeDoc() result.freeDoc() return result_str with this one: def apply_xslt(input_str, stylesheet): """input_str is a string, stylesheet is a file""" doc = lxml.etree.parse(StringIO(input_str)) styledoc = lxml.etree.parse(stylesheet) style = lxml.etree.XSLT(styledoc) result = style.apply(doc) return style.tostring(result) Now I have some issues with Unicode, which I cannot identify easily. I get UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 600: ordinal not in range(128) and similar when dealing with the result string. Everything works fine in the libxml version, though. Any ideas what/where the error could be? I've read about some unicode fixes in the Changelog, so maybe there are some additional errors in XSLT? thanks a lot, Florian

4 6

[lxml-dev] segmentation fault getting attrib on element
by Julien Anguenot 09 May '05

09 May '05

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, When doing this : >>> import lxml.etree >>> doc = lxml.etree.parse('./publication.xpdl') >>> root = doc.getroot() >>> root.attrib Segmentation fault But if : >>> import lxml.etree >>> doc = lxml.etree.parse('./publication.xpdl') >>> root = doc.getroot() >>> print root.attrib.keys() ['Id', 'schemaLocation'] It works while accessing the keys / values / items of attrib. It's not harmful but wanted to report this if it's hidding something. Or is it normal ? Config : - -------- Python 2.3.5 libxml2-2.6.16-3 libxml2-python-2.6.16-5_17.rhfc3.at libxslt-1.1.11-1 libxslt-python-1.1.11-1 libxslt-devel-1.1.11-1 on FC3. J. - -- Julien Anguenot | Nuxeo R&D (Paris, France) CPS Plateform : http://www.cps-project.org mail: anguenot at nuxeo.com; tel: +33 (0) 6 72 57 57 66 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.6 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFCe70mGhoG8MxZ/pIRAhjQAJ4gqKOBzmkShNRI6w7vRanZfRa9sgCfbGoK UfJ1hfstnZj/qWfDrtL1yyo= =IT0p -----END PGP SIGNATURE-----

5 8

Re: [lxml-dev] segmentation fault getting attrib on element
by Marc-Antoine Parent 09 May '05

09 May '05

I can reproduce it on mine, and the failure case is much simpler than yours: <?xml version="1.0" encoding="UTF-8"?> <Package xmlns:a="http://www.example.com/" a:b="c"/> Note that removing either attribute makes it pass. Marc-Antoine Parent

1 0

[lxml-dev] new testcase area for lxml
by Martijn Faassen 09 May '05

09 May '05

Hi there, I've just opened a new testcase area for lxml. In order to add testcases there, you need a login in codespeak (or convince someone with a login to add a testcase for you). There are two kinds of test cases -- a simple directory based one for those who do not know how to write unit tests, or preferred, a unit test area. Broken testcases (unittest or not) should go in the testcase area, not in the lxml test suite itself, those tests should always pass in a fresh checkout. What I (and hopefully others) will do is examine test cases, translate them into unit tests (if necessary), add them to the lxml testsuite and then fix the bug until the test passes, and remove the test case from this area again. You can checkout the testcase area here: http://codespeak.net/svn/lxml/testcase/ the README.txt contains instructions on what to add where. Suggestions are of course welcome! Regards, Martijn

1 0

[lxml-dev] XInclude not turned on by default?
by Paul Everitt 07 May '05

07 May '05

Hi all. I believe XInclude isn't turned on by default in lxml. Here is a test case: exsl-test.xml ----------------------- <?xml version="1.0" encoding="UTF-8"?> <doc> <a> <xi:include href="b.xml" xmlns:xi="http://www.w3.org/2001/XInclude"/> </a> </doc> b.xml ----------- <?xml version="1.0" encoding="UTF-8"?> <bstuff> <title>Stuff from b</title> </bstuff> exsl-test.xsl ----------------------- <?xml version="1.0" encoding="UTF-8"?> <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"> <xsl:template match="/"> output <xsl:for-each select="//bstuff"> <xsl:value-of select="title"/> </xsl:for-each> </xsl:template> </xsl:stylesheet> exsl-test.py ----------------------- import lxml.etree xmldocf = open("exsl-test.xml", "r") xmldoc = lxml.etree.parse(xmldocf) xsldocf = open("exsl-test.xsl", "r") xsldoc = lxml.etree.parse(xsldocf) xsl = lxml.etree.XSLT(xsldoc) r = xsl.apply(xmldoc) print xsl.tostring(r).encode("utf-8") 1) The output from xsltproc: paul$ xsltproc --xinclude exsl-test.xsl exsl-test.xml <?xml version="1.0"?> Stuff from b 2) The output from the python program: paul$ python ./exsl-test.py <?xml version="1.0"?> output 3) IMPORTANT NOTE: If the word "output" doesn't appear above in the .xsl, then lxml gives a bus error (on OS X, at least). Meaning, if there is no data produced, there's a bus error trying to make the string. --Paul

2 1

[lxml-dev] HELP: Some truncated output and malloc errors
by Paul Everitt 06 May '05

06 May '05

Hi all. I'm having some problems perhaps related to usage, perhaps not. Below is a script: =================================== import os, lxml.etree delivdir = os.path.split(os.path.abspath(__file__))[0][:-3] class Test: def __init__ (self): # Load the compiler-blank and compiler docs compilerblankdoc = loadfile("lib/blank-compiled.xml") compilerdoc = loadfile("lib/compiler.xsl") # Compile the compiler-blank into our style doc for repeated use compilerstyle = lxml.etree.XSLT(compilerdoc) self.compiledstyle = compilerstyle.apply(compilerblankdoc) self.processor = lxml.etree.XSLT(self.compiledstyle) def transformToString(self, inputdoc): """Run the compiled stylesheet against the input doc""" result = self.processor.apply(inputdoc) return self.processor.tostring(result).encode("utf-8") def loadfile (fn): """Load a file and return an xml dom""" f = open(delivdir + fn, "r") doc = lxml.etree.parse(f) f.close() del f return doc def main (): """Used for command-line debugging""" d = Test() sampledoc = loadfile("tests/sampledoc1.xml") result = d.transformToString(sampledoc) print result return d if __name__ == "__main__": main() =================================== When I run it, the string that is produced (in the main() function) seems very truncated relative to inspection by walking the nodes manually. Also, I get a bunch of: python(8231) malloc: *** Deallocation of a pointer not malloced: 0x1845235; This could be a double free(), or free() called with the middle of an allocated block; Try setting environment variable MallocHelp to see tools to help debug python(8231) malloc: *** Deallocation of a pointer not malloced: 0x1845117; This could be a double free(), or free() called with the middle of an allocated block; Try setting environment variable MallocHelp to see tools to help debug etc. When I run the same stylesheet and transform via xsltproc from the command line, all works ok. --Paul

2 2