exist-open

[Exist-open] XProc (Calabash) Status?

[Gary K. <ge...@ru...>]

The last update to xquery-xproc-xmlcalabash-module was about a year ago. The
README itemizes some current limitations. I would like to know if the
following limitations are still applicable, and if so can you suggest any
workarounds:

 

"Function signature will change soon to accept XML for pipeline, output, as
well as specify input/output/parameter ports and options... for now it's
primitive"

 

"p:xquery has no context with eXist, this is a big limitation, but there are
several ways around this"

 

I'm new to XProc, and have a lot of studying and experimentation to do. All
I really want to accomplish is to link multiple XQuery scripts together so
that the XML generated in one script is used as input in a FLWOR in the next
script, without explicitly creating temporary documents in the database. I'm
_assuming_ that is possible with XProc, but please correct me if I'm wrong
and I'll abandon this line of attack. Thanks!

 

--Gary Kopp

 

 

Re: [Exist-open] XProc (Calabash) Status?

[Ari N. <ari...@gm...>]

Hi,

I think Norm is busy with work and real life, but I know the Calabash XProc
3.0 version is in the works. You should look into Morgana XProc for XProc
3.0 support in the meantime.

As for running a sequence of XQueries with one providing the input to the
next, I assume you'd want to do it in an eXist context? If so, XProc
remains a rather painful option as it's not well supported.

I'd go with an XQ wrapper instead. I did something very similar to run XSLT
stylesheets in sequence in eXist. I had done it in XProc and Calabash on
the file system but getting XProc to work in eXist proved to be too much of
a pain so I reimplemented the pipeline in XQ instead.

My XProc and XQuery are both on GitHub:
https://github.com/sgmlguru/xproc-batch

It should be fairly uncomplicated to modify the XQuery to run XQueries
instead.

Best,

/Ari


On Wed, 31 Mar 2021, 21:38 Gary Kopp, <ge...@ru...> wrote:

> The last update to xquery-xproc-xmlcalabash-module was about a year ago.
> The README itemizes some current limitations. I would like to know if the
> following limitations are still applicable, and if so can you suggest any
> workarounds:
>
>
>
> “Function signature will change soon to accept XML for *pipeline*,
> *output*, as well as specify input/output/parameter ports and options...
> for now it's primitive”
>
>
>
> “p:xquery has no context with eXist, this is a big limitation, but there
> are several ways around this”
>
>
>
> I’m new to XProc, and have a lot of studying and experimentation to do.
> All I really want to accomplish is to link multiple XQuery scripts together
> so that the XML generated in one script is used as input in a FLWOR in the
> next script, without explicitly creating temporary documents in the
> database. I’m _*assuming*_ that is possible with XProc, but please
> correct me if I’m wrong and I’ll abandon this line of attack. Thanks!
>
>
>
> --Gary Kopp
>
>
>
>
> _______________________________________________
> Exist-open mailing list
> Exi...@li...
> https://lists.sourceforge.net/lists/listinfo/exist-open
>

Re: [Exist-open] XProc (Calabash) Status?

[Gary K. <ge...@ru...>]

Thanks for the suggestion, Ari. Let’s see if I understand your approach, now that I’ve looked at your code. It seems the interface between pipeline steps is a temporary document stored in the database in one step, and read in the next step using an explicit URI. Am I seeing that right?

 

--Gary

 

From: Ari Nordström <ari...@gm...> 
Sent: Wednesday, March 31, 2021 4:31 PM
To: Gary Kopp <ge...@ru...>
Cc: exist-open <exi...@li...>
Subject: Re: [Exist-open] XProc (Calabash) Status?

 

Hi,

 

I think Norm is busy with work and real life, but I know the Calabash XProc 3.0 version is in the works. You should look into Morgana XProc for XProc 3.0 support in the meantime.

 

As for running a sequence of XQueries with one providing the input to the next, I assume you'd want to do it in an eXist context? If so, XProc remains a rather painful option as it's not well supported. 

 

I'd go with an XQ wrapper instead. I did something very similar to run XSLT stylesheets in sequence in eXist. I had done it in XProc and Calabash on the file system but getting XProc to work in eXist proved to be too much of a pain so I reimplemented the pipeline in XQ instead.

 

My XProc and XQuery are both on GitHub: https://github.com/sgmlguru/xproc-batch

 

It should be fairly uncomplicated to modify the XQuery to run XQueries instead.

 

Best,

 

/Ari

 

On Wed, 31 Mar 2021, 21:38 Gary Kopp, <ge...@ru... <mailto:ge...@ru...> > wrote:

The last update to xquery-xproc-xmlcalabash-module was about a year ago. The README itemizes some current limitations. I would like to know if the following limitations are still applicable, and if so can you suggest any workarounds:

 

“Function signature will change soon to accept XML for pipeline, output, as well as specify input/output/parameter ports and options... for now it's primitive”

 

“p:xquery has no context with eXist, this is a big limitation, but there are several ways around this”

 

I’m new to XProc, and have a lot of studying and experimentation to do. All I really want to accomplish is to link multiple XQuery scripts together so that the XML generated in one script is used as input in a FLWOR in the next script, without explicitly creating temporary documents in the database. I’m _assuming_ that is possible with XProc, but please correct me if I’m wrong and I’ll abandon this line of attack. Thanks!

 

--Gary Kopp

 

 

_______________________________________________
Exist-open mailing list
Exi...@li... <mailto:Exi...@li...> 
https://lists.sourceforge.net/lists/listinfo/exist-open

Re: [Exist-open] XProc (Calabash) Status?

[Ari N. <ari...@gm...>]

Hi Gary,

You don't actually save anything before the pipeline is finished, unless 
$debug is set.

The manifest listing the XSLTs is first read into a sequence. That 
sequence is then used in a recursive function for the actual transforms, 
one XSLT at a time, for as long as there is an XSLT left in the 
sequence. The output from one recursion is used as the input to the next.

Plus, since I may have to feed steps with parameters, there's stuff in 
there to handle parameters set in the manifest.

You can test it using the XSLT-pipelines example repo at 
https://github.com/sgmlguru/xslt-pipelines. The example works both in 
XProc on the file system and in eXist using XQuery.

Oh, and I spoke about the pipelined XSLT approach at Balisage last year: 
https://www.balisage.net/Proceedings/vol25/html/Nordstrom01/BalisageVol25-Nordstrom01.html.

Best,

/Ari

On 2021-04-01 01:27, Gary Kopp wrote:
>
> Thanks for the suggestion, Ari. Let’s see if I understand your 
> approach, now that I’ve looked at your code. It seems the interface 
> between pipeline steps is a temporary document stored in the database 
> in one step, and read in the next step using an explicit URI. Am I 
> seeing that right?
>
> --Gary
>
> *From:* Ari Nordström <ari...@gm...>
> *Sent:* Wednesday, March 31, 2021 4:31 PM
> *To:* Gary Kopp <ge...@ru...>
> *Cc:* exist-open <exi...@li...>
> *Subject:* Re: [Exist-open] XProc (Calabash) Status?
>
> Hi,
>
> I think Norm is busy with work and real life, but I know the Calabash 
> XProc 3.0 version is in the works. You should look into Morgana XProc 
> for XProc 3.0 support in the meantime.
>
> As for running a sequence of XQueries with one providing the input to 
> the next, I assume you'd want to do it in an eXist context? If so, 
> XProc remains a rather painful option as it's not well supported.
>
> I'd go with an XQ wrapper instead. I did something very similar to run 
> XSLT stylesheets in sequence in eXist. I had done it in XProc and 
> Calabash on the file system but getting XProc to work in eXist proved 
> to be too much of a pain so I reimplemented the pipeline in XQ instead.
>
> My XProc and XQuery are both on GitHub: 
> https://github.com/sgmlguru/xproc-batch 
> <https://github.com/sgmlguru/xproc-batch>
>
> It should be fairly uncomplicated to modify the XQuery to run XQueries 
> instead.
>
> Best,
>
> /Ari
>
> On Wed, 31 Mar 2021, 21:38 Gary Kopp, <ge...@ru... 
> <mailto:ge...@ru...>> wrote:
>
>     The last update to xquery-xproc-xmlcalabash-module was about a
>     year ago. The README itemizes some current limitations. I would
>     like to know if the following limitations are still applicable,
>     and if so can you suggest any workarounds:
>
>     “Function signature will change soon to accept XML for /pipeline/,
>     /output/, as well as specify input/output/parameter ports and
>     options... for now it's primitive”
>
>     |“p:xquery| has no context with eXist, this is a big limitation,
>     but there are several ways around this”
>
>     I’m new to XProc, and have a lot of studying and experimentation
>     to do. All I really want to accomplish is to link multiple XQuery
>     scripts together so that the XML generated in one script is used
>     as input in a FLWOR in the next script, without explicitly
>     creating temporary documents in the database. I’m _/assuming/_
>     that is possible with XProc, but please correct me if I’m wrong
>     and I’ll abandon this line of attack. Thanks!
>
>     --Gary Kopp
>
>     _______________________________________________
>     Exist-open mailing list
>     Exi...@li...
>     <mailto:Exi...@li...>
>     https://lists.sourceforge.net/lists/listinfo/exist-open
>     <https://lists.sourceforge.net/lists/listinfo/exist-open>
>