Mailman 3 September 2006 - lxml - The Python XML Toolkit

[lxml-dev] 1.0.4 released
by Stefan Behnel 09 Sep '06

09 Sep '06

Hi all, I just released lxml 1.0.4 to cheeseshop. It's a bug-fix release that fixes a crash bug in the replace() function. It also adds an extend() method to Element to make it more list-like. 1.1 final will still have to wait a little longer until the filename issue has settled. If I find the time, it will be out next week. I know, I keep saying things like that, but we had a couple of bug reports last week, so I'm happy that we can solve some more problems before the release. Have fun, Stefan

1 0

[lxml-dev] Building debian packages
by Paweł Pałucha 07 Sep '06

07 Sep '06

Hi, I've got a small patch against current SVN to file doc/build.txt: * info about changing md5sum in *.dsc file * command `dpkg -x ...` is incorrect - it should be `dpkg-source -x ...` Paweł Index: doc/build.txt =================================================================== --- doc/build.txt (wersja 32034) +++ doc/build.txt (kopia robocza) @@ -217,7 +217,8 @@ * remove the unpacked directory * tar.gz the lxml SVN version and replace the orig.tar.gz that lies in the directory -* do ``dpkg -x lxml-...dsc`` and cd into the newly created directory +* check md5sum of created tar.gz file and place new sum and size in dsc file +* do ``dpkg-source -x lxml-...dsc`` and cd into the newly created directory * run ``dch -i`` and add a comment like "use trunk version", this will increase the debian version number so apt/dpkg don't get confused * run ``dpkg-buildpackage -rfakeroot -us -uc`` to build the package

2 1

Re: [lxml-dev] File name encoding problems
by Stefan Behnel 07 Sep '06

07 Sep '06

Hi Paweł, Paweł Pałucha wrote: > Your patch is almost ok, I mean it produces: > > File "p.py", line 5, in ? > d = lxml.etree.parse('/tmp/ó.xml') > File "etree.pyx", line 1774, in etree.parse > File "parser.pxi", line 878, in etree._parseDocument > File "parser.pxi", line 798, in etree._parseDocFromFile > File "parser.pxi", line 512, in etree._BaseParser._parseDocFromFile > File "parser.pxi", line 583, in etree._handleParseResult > File "etree.pyx", line 195, in etree._ExceptionContext._raise_if_stored > File "parser.pxi", line 275, in etree._parser_resolve_from_python > File "apihelpers.pxi", line 492, in etree.funicode > UnicodeDecodeError: 'utf8' codec can't decode bytes in position 5-8: invalid data > > but if I change line 275 in parse.pxi (etree._parser_resolve_from_python) from: > url = funicode(c_url) > to: > url = c_url > > parsing works ok. Does it make sense? It's not a perfect solution, as it may mean that the user ends up with a UTF-8 encoded URL to resolve. That's rarely a problem, as most URLs are plain ASCII anyway, except for filenames. URLs are passed as UTF-8 encoded strings if they occurred in a document. The funicode() call is meant to handle that. Filenames are passed as byte strings encoded for the local filesystem if they were provided by the user and are UTF-8 encoded if they originate from a document, so that makes custom resolvers more difficult to handle. At least, it's always a byte encoded string, it's just not obvious to the user which encoding was used... As it works with ASCII URLs and also supports the case of locally encoded filenames, I think it's an acceptable tradeoff. I'll put it into the trunk. Stefan

1 0

[lxml-dev] help: lxml causes wxpython app to crash at exit
by Oliver Schoenborn 06 Sep '06

06 Sep '06

Hi, just wondering if anyone else has had same problem and what solution they found: 1. system is RHEL 4, with python 2.3, and latest wxpython from rpm on wxpython website; python and wxpython work fine (demo etc) 2. latest lxml (http://codespeak.net/lxml/) 1.0.3 installed (with latest libxml2 and libxslt built from source) and also built lxml from source (with latest pyrex etc) 3. when I open pyalamode (a wx app with a python shell window), then do "from lxml import etree" then exit, a seg fault occurs; the core dump says it happens during finalization of the python interpreter, possibly when terminating some thread. 4. same thing happens from my own wx app (if I uncomment a line that says "from lxml import etree" there is no segfault after exit) 5. I can use lxml for validation etc from within the wx app, all is fine; the problem is *only* when exiting the wx app 6. this segfault does NOT happen on Windows, 7. it does NOT happen if I do the above from within python shell in xterm 8. the problem does NOT happen when just do "import lxml", lxml.etree is the problem; etree is a binding to some C code in lxml/etree.c; this file is generated by pyrex Upgrading to python 2.4 will not be easy so I want to make sure I exhaust other solutions first. Thanks, Oliver

2 1

[lxml-dev] File name encoding problems
by Paweł Pałucha 06 Sep '06

06 Sep '06

Hi! I'm not sure if it's really an lxml problem, but it looks like... I'm using lxml version 1.0.3 (from Debian package). My default system encoding is ISO-8859-2. I have a simple program: #!/usr/bin/python # -*- coding: iso-8859-2 -*- import lxml.etree d = lxml.etree.parse('/tmp/ó.xml') The problem is with letter 'ó' (or any other non-ascii letter) in file name. While running program I get something like that: Traceback (most recent call last): File "./p.py", line 6, in ? d = lxml.etree.parse('/tmp/łódź.xml') File "etree.pyx", line 1615, in etree.parse File "parser.pxi", line 687, in etree._parseDocument File "apihelpers.pxi", line 343, in etree._utf8 AssertionError: All strings must be Unicode or ASCII If I try to write path using UTF-8 encoding, the file cannot be found (because the path with UTF-8 encoded name does not exists). This did not happen with python 2.3 - the problem is only with python2.4. There's a workaround - reading file to StringIO object and then parsing XML from that object. It works fine but it's silly. I would be very grateful for any suggestions. Paweł Pałucha

2 2

[lxml-dev] Pure Text Elements as XSLT results impossible?
by Hans-Jürgen Hay 05 Sep '06

05 Sep '06

XSLT transforms that should result in a TEXT only Element silently result in some kind of 'None' result class with the text seemingly lost. Propably caused by Text Elements beeing second class in ElementTree. This seems to be very wrong or am I. Regards Hans from lxml import etree xsl = """<?xml version="1.0"?> <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:output method="text" /> <xsl:template match="/"> TestText </xsl:template> </xsl:stylesheet>""" trans = etree.XSLT(etree.XML(xsl)) str(etree.tostring(trans(etree.XML('<test/>')))) --->> 'None' insted of 'TestText' ---------------------------------------------------------------------- from lxml import etree xsl = """<?xml version="1.0"?> <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:template match="/"> <dummy/> TestText </xsl:template> </xsl:stylesheet>""" trans = etree.XSLT(etree.XML(xsl)) etree.tostring(trans(etree.XML('<test/>'))) -->> '<dummy/> TestText' (works) ---------------------------------------------------------------------- from lxml import etree xsl = """<?xml version="1.0"?> <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:template match="/"> TestText <dummy/> </xsl:template> </xsl:stylesheet>""" trans = etree.XSLT(etree.XML(xsl)) etree.tostring(trans(etree.XML('<test/>'))) --->> '<dummy/>' ('TestText' in front missing)

2 1