project-cdsware-users@cern.ch archives


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: bibrank problem due to old pystemmer


  • From: Samuele Kaplun <Samuele.Kaplun@xxxxxxx>
  • Subject: Re: bibrank problem due to old pystemmer
  • Date: Mon, 13 Aug 2007 16:17:48 +0200

Dear Kam-ming,

In data lunedì 13 agosto 2007, Kam-ming Ku scribacchiaste dette parole:
> Mine is 0.93.  When I run bibrank -f50000, it gives:

This means bleeding edge ;-)

> [root@eproxy1 cds-invenio]# /opt/cds-invenio/bin/bibrank 11443
> 2007-08-13 21:21:41 --> Task #11443 started.
> 2007-08-13 21:21:41 --> Running rank method: journal impact factor.
> 2007-08-13 21:21:41 --> No new records added since last time method was run
> 2007-08-13 21:21:41 --> Reading knowledgebase file:
> /opt/cds-invenio/etc/bibrank/demo_jif.kb 2007-08-13 21:21:41 --> Number of
> lines read from knowledgebase file: 5 2007-08-13 21:21:41 --> Number of
> records available in rank method: 0 2007-08-13 21:21:41 --> Running rank
> method: word similarity
> 2007-08-13 21:21:41 -->
> Exception caught: 'module' object has no attribute 'algorithms'
>   File "/usr/lib/python2.4/site-packages/invenio/bibrank.py", line 138, in
> task_run_core func_object(key)
>   File "/usr/lib/python2.4/site-packages/invenio/bibrank_word_indexer.py",
> line 1210, in word_similarity return word_index(run)
>   File "/usr/lib/python2.4/site-packages/invenio/bibrank_word_indexer.py",
> line 801, in word_index tags = get_tags(config) #get the tags to include
>   File "/usr/lib/python2.4/site-packages/invenio/bibrank_word_indexer.py",
> line 878, in get_tags if config.get(function,"stemming") and
> stem("information", "en") != "inform": File
> "/usr/lib/python2.4/site-packages/invenio/bibindex_engine_stemmer.py", line
> 47, in stem if lang and is_stemmer_available_for_language(lang):
>   File
> "/usr/lib/python2.4/site-packages/invenio/bibindex_engine_stemmer.py", line
> 42, in is_stemmer_available_for_language _stemmers[thread_ident] =
> _create_stemmers()
>   File
> "/usr/lib/python2.4/site-packages/invenio/bibindex_engine_stemmer.py", line
> 55, in _create_stemmers for lang in Stemmer.algorithms():

This is due to a recent change in CVS bleeding edge code. We just reintroduced
PyStemmer from a new incarnaction (http://snowball.tartarus.org). It's still
called PyStemmer but has a different api (this means a different interface
from developing point of view) from the obsoleted year-2002-PyStemmer
(http://sourceforge.net/projects/pystemmer/), and that's why you received
this error.
The best way to fix this on your machine is to remove from the system the
obsolete PyStemmer you will surely have installed in the past.
When we will release a stable version of CDS Invenio you will definitly find
indications on how to install the new PyStemmer by Snowball.

Best regards,
	Samuele


--
.O.
..O
OOO