Re: [PyWrapper-devel] MOBY info

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

milko,
you know there are variables in tapir.=20
but the problem is that the numbers used for the summary section can =
only be produced when the entire model section is over!
thats a vicous circle. you cant use these numbers before the output is =
done. so you cant use them to create the output!
m

	-----Original Message-----
	From: pyw...@li... =
[mailto:pyw...@li...] On Behalf Of =
Skofic A. Milko (IPGRI)
	Sent: Donnerstag, 16. November 2006 11:34
	To: D=F6ring, Markus
	Cc: PyWrapper Developers mailing list; Javier privat
	Subject: Re: [PyWrapper-devel] MOBY info
=09
=09
	I understand you have to parse a stream and not load the whole thing in =
memory or it would be impossible to handle, but is there any way to get =
the <summary> section someway?

	If I am correct, in the output model you have a <structure> part that =
defines the model onto which you want to map the TAPIR output, in the =
<mapping> section you map elements of the <structure> section to data =
output section inside of the <search> part of the TAPIR response. Is =
there any way to access elements outside of the schema the database is =
wrapped to?

	For instance:

	<search>
	<Datasets>
	<Dataset>
	...
	</Dataset>
	</Datasets>
	<summary start=3D"0" totalMatched=3D"25008" next=3D"50"/>
	<search>

	Are there any "system variables" to which the start=3D"0" =
totalMatched=3D"25008" next=3D"50" could be written to once you get to =
the end of the response?

	Thanks!

	Milko

	On Nov 16, 2006, at 10:09 , D=F6ring, Markus wrote:

		thats right.
		the summary section is at the end of the model result section cause =
otherwise pywrapper would have to know the numbers in advance which =
means keeping it in memory and no streaming. thats something we wanted =
to avoid when designing tapir.
		=20
		markus
		=20

			-----Original Message-----
			From: pyw...@li... =
[mailto:pyw...@li...] On Behalf Of =
Skofic A. Milko (IPGRI)
			Sent: Donnerstag, 16. November 2006 00:56
			To: Javier privat
			Cc: PyWrapper Developers mailing list
			Subject: Re: [PyWrapper-devel] MOBY info
		=09
		=09
			I think a count can be very useful, like getting the number of =
elements conforming to some condition. Also, it can be useful to have an =
idea of how many elements one must get, then use that number when =
setting the secondary paging parameters.

			I agree that search is enough for now, that covers most of the needs.

				The count operation is not available for use with biomoby. There is =
no way in TAPIR, apart of using XSLT on top of it, to modify the =
response dynamically of a search result that has the count set to true.

			Does this mean that in an output model you don't have access to the =
response structure of a TAPIR result? for instance to this:
			<summary start=3D"0" totalReturned=3D"0" totalMatched=3D"25008" />=20

			Thanks!

			Milko

			On Nov 16, 2006, at 0:39 , Javier de la Torre wrote:

				Hi Milko,=20

				As you correctly said the moby2TapirFilterXslt just transform a =
biomoby message into a tapir filter, just the filter! not a full TAPIR =
request. Take a look at the XSLT, is pretty simple.

				As it is implemented right now the only operations that can be done =
is a search...

				I implemented it like because with search is the only operation that =
you can do dynamic outputmodels and the only thing that it is needed to =
trasnform is the filter.

				The count operation is not available for use with biomoby. There is =
no way in TAPIR, apart of using XSLT on top of it, to modify the =
response dynamically of a search result that has the count set to true.

				So, again, it is not possible to create biomoby services in =
pywrapper that uses a COUNT. What kind of services do you envision to do =
this? If you really need it then we can think about other possibilities =
but I dont know if this is necessary...

				Cheers.

				On 15/11/2006, at 19:50, Skofic A. Milko (IPGRI) wrote:

					Javier, podr=EDas decirme cuando me podr=EDas dedicar un poco de =
tiempo para comprender unas cosas de como implementas el MOBY?=20

					Te indico las cosas que no me son claras:

					El  <moby2TapirFilterXslt> que hace exactamente, transforma el =
input Moby mensaje en el filtro te b=FAsqueda TAPIR? Si eso es la =
verdad, solamente se necessita crear la parte <filter> de la b=FAsqueda, =
lo otro es automatico?

					Si yo quer=EDa hacer un MOBY service que me da el COUNT de cuantos =
registros ha encontrado como puedo hacer? El problema es que esta =
informaci=F3n me la devuelve el protocolo TAPIR, no el esquema que se =
utiliza.

					Yo har=EDa esta query;

					<?xml version=3D'1.0' encoding=3D'UTF-8'?>
					<request>
					<header>
					<source =
accesspoint=3D"http://tapir.grinfo.net:8080/pywrapper/pywrapper?dsa=3DEUR=
ISCO" sendtime=3D"2006-11-13T12:54:29.43"/>
					</header>
					<search count=3D'true' start=3D'0' limit=3D'0'>
					<externalOutputModel =
location=3D"http://eurisco.ecpgr.org/models/output/MCPD_EURISCO_1.00.xml"=
/>
					<filter>
					<like>
					<concept id=3D'GENUS@mcpd_eurisco_100' />
					<literal value=3D'B*' />
					</like>
					</filter>
					<orderBy>
					<concept id=3D'ACCENUMB@mcpd_eurisco_100' />
					</orderBy>
					</search>
					</request>

					Y TAPIR me devolver=EDa:

					<?xml version=3D"1.0" encoding=3D"utf-8"?>
					<response xmlns=3D"http://rs.tdwg.org/tapir/1.0">
					  <header>
					    <source sendtime=3D"2006-11-15T19:26:50.57" =
accesspoint=3D"http://tapir.grinfo.net:8080/pywrapper/pywrapper?dsa=3DEUR=
ISCO_TEST">
					      <software version=3D"2.1" name=3D"PyWrapper" />
					    </source>
					  </header>
					  <!-- CONTENT START -->
					  <search>
					    <summary start=3D"0" totalReturned=3D"0" totalMatched=3D"25008" =
/>
					  </search>
					  <!-- CONTENT END -->
					  <diagnostics></diagnostics>
					</response>

					Ent=F3nces tendr=EDa que hacer un output model con la estructura =
del <response> mapeado a el MOBY type?

					Ciao!

					Milko

		=09
		=09
			Milko A. Skofic (IPGRI)
			International Plant Genetic Resources Institute
			Via dei Tre Denari, 472/a
			00057 Maccarese (RM)
			ITALY
			E-mail: m.s...@cg...
			Tel: +30 06 6118286