[BioPython] Re: [Biopython-dev] PubMed module

Yasushi MASUDA ymasuda at ethercube.com
Tue Dec 7 03:24:54 EST 2004


As far as I know/believe, PubMed.py, which handles communication with 
Entrez/PubMed, imports Bio.WWW.RequestLimiter. It seems to limit your 
access to NCBI based on time delay. 

The default delays are 5.0 seconds for Dictionary(), 2.0 seconds for 
search_for(), 120.0 seconds for download_many(). Those delays are 
injected just *before* each iterative request (even in single query) 
by RequestLimitter so that your program will sleep before retriving 
individual entry every time.

Thus, if you write your program in well-behaved manner 
(no parallelize of requests by threading, no use of broken system 
timer, etc.), there should be no problem.
If you are interested in reading Python code doing it, please take a 
look into $pythonlib/site-packages/Bio/WWW/__init__.py or 
$ptyhonlib/site-packages/Bio/PubMed.py.


"Hisahiko Sato" <hisahikosatou at hotmail.com> wrote:

> I am considering using pubmed class of BioPython to access Pubmed.
> 
> I have a question for using this.
> Is there any limitation for using this class?
> I am afraid too much use of this class would recognize as traffic attack to
> ncbi site from my company.
> I heard ncbi would limit the internet access if there are too much access.

---
Yasushi Masuda
http://www.cubelab.com/ymasuda/




More information about the BioPython mailing list