[Biophp-dev] Finally got it out of my system!

Serge Gregorio biophp-dev@bioinformatics.org
Tue, 10 Jun 2003 13:20:24 +0800


Hello all!

>> Nonetheless, I proceeded to write some PHP code 
>> that generates SQL INSERTs from data displayed on >> web forms.  You can find them at:
>>
>>    http://genephp.sf.net/parse_na_gb_par.html   [1]
>>    http://genephp.sf.net/parse_na_gb_res.php    [2]
>>    http://genephp.sf.net/insert_na_gb_res.php   [3]
>
>Do these scripts work with the BioSQL database or >with the tables you wrote?  I suggest to stick to the >BioSQL layout and write the scripts necessary to >insert a (biophp) seq object into this structure and >to read them back out again.  No need to reinvent the >wheel.

Uh okay.  But it'll take some time for me to figure out this new wheel.  Also, in db design, there is no one right 'wheel'.  Sometimes, you have to reinvent it to suit your peculiar needs (speed vs. expressive power vs. etc.).  Finally, BioSQL does not seem to cover Enzymes, Biological Pathways, Promoters, etc., well at least, that's my impression. 

That's why I'm exploring the idea (a webscript) that 
makes it easy create alternate db structures for the various biological databases.  If I could even add a GUI, so much the better.  More on that later.

>PEAR and adodb basically do the same thing.  Since >you really do not need to 'install' anything to get >either PEAR or adodb to work (you simply have to >download one or the other) and I don't think we are  
>ready to spread our resources by using both of them, >it makes sense to choose one or the other.  

Ah, but at the back of my mind, I had this idea that 
one supports some databases that the other does not.
(Like, ADODB supports ODBC, PEAR does not.)  So my suggestion was meant to be able to interface to as many databases as possible by having both of them.

But if this assumption isn't necessarily true (e.g. 
95 % of the databases that one supports is also supported by the other, and that the remaining 5 % isn't worth the trouble), then let's go with one (may I vote for ADODB?).

>> If anyone would also like to work on a DATABASE SCHEMA for data  
>> extracted by the various parser scripts (there are 30 formats  
>> supported fully/partially at http://genephp.sf.net/io_set.html), well,  
>> that would be swell...

>All great and nice, but I don't see any of the >code...  What are you parsing into?  Are you using 

I'm using some code from there, but not all.

>Sometimes I have the feeling we are not really >working together here...

Not at all... all I need is a CVS that I can write to.
Also, I've incorporated some of the code contributed by Frankie (levenshtein) in one of the demo scripts I think... 

Also, lately I've detoured into a different approach.
I write scripts that do something useful in the quickest way I can (coding conventions be damned), 
and then go back and fix the code into nice little BioPHP modules (coding conventions be praised). =)

The code is quick and dirty and might be criticized 
as being "Windows-like". *guffaw* 

I'll post them once I've fixed them up... 

Lastly, you're all welcome to contribute demo scripts. But take note of SF's policies: no CURL, no script that allows the public to INSERT to the database, no script that's resource-intensive, etc. 

Cheers!

Serge


Need a new email address that people can remember
Check out the new EudoraMail at
http://www.eudoramail.com