lxml.etree._utf8 crash - Page 2

Agama · 08-02-2012, 09:39 AM

Quote:

Originally Posted by nickredding

I've done a lot for calibre. You just don't know it because I don't trumpet it around.

My post had nothing to do with how much you have done for calibre, (I may well be using features that you worked on), but was simply a comment on your attitude regarding the lxml issue.

nickredding · 08-02-2012, 10:08 AM

Quote:

Originally Posted by kovidgoyal

Because pot, illegal characters are already stripped from both the title and the text_summary in the __init__ method of Article class. So your exception isn't because of illegal characters.

That is not what the traceback says, and I confirmed that with a try/except traceback at line 177 in templates.py

kovidgoyal · 08-02-2012, 10:16 AM

A try except does not confirm anything other than that an exception occurs there. something that is already evident from the traceback. The question is why the exception occurs. Since non printable ascii chars and null chars have already been removed, you need to figure out why the exception is occurring and if you do figure that out, come up with a patch that prevents it happening in general.

nickredding · 08-02-2012, 10:44 AM

And the message in the original traceback means nothing? My purpose in placing the try was to confirm where the error arose. I then checked the periodical being processed and sure enough the garbage characters showed up in the web browser. So I'm not sure what further diagnosis cab be done except to look at the code you say is stripping out these characters. If you point me at the code i will have a look.

kovidgoyal · 08-02-2012, 11:17 AM

Error messages often dont mean what they say. People rarely update error messages to cover every condition that can trigger the error.

Article.__init__ in feeds/__init__.py

"Garbage characters" doesn't mean much. The fact that the characters are not interpretable by a human does not neccessarily mean anything. lxml chokes on unicode strings that ascii control codes or null characters. Both those are taken care of already. If there is some other well defined character set that lxml chokes on, that is not valid in a unicode text string, then that can be added, though I suspect that is not the case.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
0.7.30 crash	nickredding	Calibre	1	11-27-2010 01:40 PM
Pseudo-crash w/V 6.39	petercreasey	Calibre	12	02-11-2010 05:59 AM
calibre-0.6.31, mechanize and lxml	taurnil	Calibre	5	01-01-2010 07:47 AM
calibre python-lxml problem on ubuntu	carpii	Calibre	5	11-29-2008 05:34 AM
upgrade failed - but not python-lxml fault	alexxxm	Calibre	7	10-06-2008 09:36 AM

08-02-2012, 10:16 AM	#18
kovidgoyal creator of calibre Posts: 43,857 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	A try except does not confirm anything other than that an exception occurs there. something that is already evident from the traceback. The question is why the exception occurs. Since non printable ascii chars and null chars have already been removed, you need to figure out why the exception is occurring and if you do figure that out, come up with a patch that prevents it happening in general.

08-02-2012, 10:44 AM	#19
nickredding onlinenewsreader.net Posts: 324 Karma: 10143 Join Date: Dec 2009 Location: Phoenix, AZ & Victoria, BC Device: Kindle 3, Kindle Fire, IPad3, iPhone4, Playbook, HTC Inspire	And the message in the original traceback means nothing? My purpose in placing the try was to confirm where the error arose. I then checked the periodical being processed and sure enough the garbage characters showed up in the web browser. So I'm not sure what further diagnosis cab be done except to look at the code you say is stripping out these characters. If you point me at the code i will have a look.

08-02-2012, 11:17 AM	#20
kovidgoyal creator of calibre Posts: 43,857 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Error messages often dont mean what they say. People rarely update error messages to cover every condition that can trigger the error. Article.__init__ in feeds/__init__.py "Garbage characters" doesn't mean much. The fact that the characters are not interpretable by a human does not neccessarily mean anything. lxml chokes on unicode strings that ascii control codes or null characters. Both those are taken care of already. If there is some other well defined character set that lxml chokes on, that is not valid in a unicode text string, then that can be added, though I suspect that is not the case.