[Biophp-dev] Back on track

Dong Gregorio biophp-dev@bioinformatics.org
Tue, 22 Apr 2003 08:09:11 +0800

Hello all!

Sorry I've been away for the "mail" for a while...

Sean, I've finally read about (at NCBI's site) the eSearch,
eFetch, eUtils you were mentioning last time.  From what I
gather, you'd need CURL to actually save the data returned
by the NCBI site to your local CPU.  And then you'd need
to parse the data which is in XML format.  Is that correct?

So, my next question is: what are you working on at the 
moment: the CURl part or the XML parsing part?  Both?
Have you finally got PHP's XML parser functions (based
on expat) to work?  Or are you using something else?



Sean wrote:


On Sat, 19 Apr 2003 14:53:34  
 S Clark wrote:
>Thanks, firstly, to Greg Tyrelle for supplying a working parser example.  I 
>believe I've "got it" now...
>Later today I'll be uploading the working (though minimal and still sloppily
>written) ESearch module components to my "point-n-laugh" section.  Following
>that, I'll work on getting the XML parser and interface for POST'ing queries
>to webservers more properly separated and abstracted out for "semi-official"
>release.  Once that's done, the more detailed work to get minimal useful
>ESummary and EFetch modules working should not require much additional
>After that I'll be getting back to the NCBI Blast interface (which you can
>find a more-or-less working unrefined version of in the "point-n-laugh"
>section, which is effectively capable of submitting a sequence and parameters
>for a sequence query, and more-or-less properly waiting for the results
>to come back.  It just doesn't do any PARSING of the results at all...)
>A sort of "personal milestone" that I'm aiming for is to be able to put 
>together an example "tell me about this sequence" application, which would
>examine results of an NCBI Blast query and, using information pulled from
>the closest matches that turn up, report things like probable taxonomic
>classification, type of sequence, and pointers to likely-related pubmed
>articles, and so on.  
>As always, comments and suggestions are welcome.  Thanks.

Need a new email address that people can remember
Check out the new EudoraMail at