Automate XSLT update in gems

[CAMOBAP]

Intro

Thanks to @opoudjis for the suggestion of improvement which can help us to keep xslt up-to-date in our gems

Problem

We need to always keep eye on xslt update & propagate them to gems

Proposed solution

Implement separate GHA workflow which will be triggered by release (or other even on this repo), the steps of this flow

• build XSLTs with make xslts
• commit xslts which was generated according to the mapping bellow:

   cp xslt/iec.international-standard.xsl ../metanorma-iec/lib/isodoc/iec
   cp xslt/iso.international-standard.xsl ../metanorma-iso/lib/isodoc/iso
   cp xslt/iso.amendment.xsl ../metanorma-iso/lib/isodoc/iso
   cp xslt/itu.recommendation.xsl ../metanorma-itu/lib/isodoc/itu
   cp xslt/itu.recommendation-annex.xsl ../metanorma-itu/lib/isodoc/itu
   cp xslt/itu.resolution.xsl ../metanorma-itu/lib/isodoc/itu
   cp xslt/ogc.abstract-specification-topic.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.best-practice.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.change-request-supporting-document.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.community-practice.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.community-standard.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.discussion-paper.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.engineering-report.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.policy.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.reference-model.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.release-notes.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.standard.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.test-suite.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.user-guide.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.white-paper.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/ogc.other.xsl ../metanorma-ogc/lib/isodoc/ogc
   cp xslt/un.plenary.xsl ../metanorma-unece/lib/isodoc/un
   cp xslt/un.plenary-attachment.xsl ../metanorma-unece/lib/isodoc/un
   cp xslt/un.recommendation.xsl ../metanorma-unece/lib/isodoc/un
   cp xslt/csd.standard.xsl ../metanorma-csd/lib/isodoc/cc
   cp xslt/csa.standard.xsl ../metanorma-csand/lib/isodoc/csa
   cp xslt/rsd.standard.xsl ../metanorma-rsd/lib/isodoc/ribose
   cp xslt/m3d.report.xsl ../metanorma-m3d/lib/isodoc/m3aawg
   cp xslt/gb.recommendation.xsl ../metanorma-gb/lib/isodoc/gb
   cp xslt/iho.specification.xsl ../metanorma-iho/lib/isodoc/iho
   cp xslt/iho.standard.xsl ../metanorma-iho/lib/isodoc/iho
  

Open questions

• Which event should trigger xslts update? Release event or some else even? - Use push: on master as trigger event

cc @opoudjis @Intelligent2013 @ronaldtse

[CAMOBAP]

@Intelligent2013 please let me know how do you think generally about the topic? Also maybe you have something to say for the open question?

[Intelligent2013]

@CAMOBAP when I update XSLTs, I change them in a separate branch to make sure that all PDF generate ok. After merge branch to master, XSLTs in xslt_src folder are ready to further use. I don't use a git versions. So I think you can triggered workflow on push or merging to master branch.

[CAMOBAP]

@Intelligent2013 thanks for your input, push to master as the event, works perfectly also

[opoudjis]

@CAMOBAP So has this functionality been implemented? That behaviour does not seem to be happening...

[CAMOBAP]

@opoudjis hmm, let me check

[Intelligent2013]

Ping @CAMOBAP.

[ronaldtse]

@Intelligent2013 I think @CAMOBAP is on holiday, he's returning the week after.

[CAMOBAP]

@opoudjis @ronaldtse @Intelligent2013

1. Should we rename the files below?

• xslt_src/csd.standard.core.xsl -> xslt_src/cc.standard.core.xsl
• xslt_src/csd.standard.presentation.core.xsl -> xslt_src/cc.standard.presentation.core.xsl
• xslt_src/m3d.report.core.xsl -> xslt_src/m3aawg.report.core.xsl
• xslt_src/m3d.report.presentation.core.xsl -> xslt_src/m3aawg.report.presentation.core.xsl
• xslt_src/mpfd.circular.core.xsl -> xslt_src/mpfa.circular.core.xsl
• xslt_src/mpfd.circular.presentation.core.xsl -> xslt_src/mpfa.circular.presentation.core.xsl
• xslt_src/mpfd.compliance-standards-for-mpf-trustees.core.xsl -> xslt_src/mpfa.compliance-standards-for-mpf-trustees.core.xsl
• xslt_src/mpfd.compliance-standards-for-mpf-trustees.presentation.core.xsl -> xslt_src/mpfa.compliance-standards-for-mpf-trustees.presentation.core.xsl
• xslt_src/mpfd.guidelines.core.xsl -> xslt_src/mpfa.guidelines.core.xsl
• xslt_src/mpfd.guidelines.presentation.core.xsl -> xslt_src/mpfa.guidelines.presentation.core.xsl
• xslt_src/mpfd.standards.core.xsl -> xslt_src/mpfa.standards.core.xsl
• xslt_src/mpfd.standards.presentation.core.xsl -> xslt_src/mpfa.standards.presentation.core.xsl
• xslt_src/mpfd.supervision-of-mpf-intermediaries.core.xsl -> xslt_src/mpfa.supervision-of-mpf-intermediaries.core.xsl
• xslt_src/mpfd.supervision-of-mpf-intermediaries.presentation.core.xsl -> xslt_src/mpfa.supervision-of-mpf-intermediaries.presentation.core.xsl
• xslt_src/rsd.standard.core.xsl -> xslt_src/ribose.standard.core.xsl
• xslt_src/rsd.standard.presentation.core.xsl -> xslt_src/ribose.standard.presentation.core.xsl

2. As far as I understand we should copy any produced xslt/*.xsl except *.presentation.*, is that correct?

[ronaldtse]

@CAMOBAP rename: yes please, an excellent find 😉 Also need to ping @opoudjis that the XSL names will change.

As far as I understand we should copy any produced xslt/*.xsl except .presentation., is that correct?

@Intelligent2013 can you please help answer? Thanks!

[CAMOBAP]

rename: yes please, an excellent find 😉 Also need to ping @opoudjis that the XSL names will change.

@ronaldtse @opoudjis I can change it in specific processors too. To be on the same page, should I care about potential backward compatibility issues or anything what should I check after this rename?

[ronaldtse]

I can change it in specific processors too.

@CAMOBAP thank you. I don't think there are side effects but let's see what @opoudjis says.

[Intelligent2013]

@CAMOBAP I didn't look at changes in PR yet, therefore answer about current situation.

1. Should we rename the files below?

Each xslt filename has a parts related to xml structure. For example, csd.standard.presentation.core.xsl:

• csd - <csd-standard xmlns="https://www.metanorma.org/ns/csd">
• standard - <doctype>standard</doctype>

These rules are using for PDF generating in mn-native-pdf in gh-pages branch.

If we change csd to cc, then we need some external determined rules for relationship between xslt filename and source xml file.

2. As far as I understand we should copy any produced xslt/*.xsl except .presentation., is that correct?

@CAMOBAP in xslt_src folder there are two kind of core.xsl files:

• with presentation suffix - it should be used for PDF
• without presentation suffix - obsoleted xslt files and can be omitted.

So in xslt/*presentation.xsl should be used.

[CAMOBAP]

@Intelligent2013 thanks for the explanation, I asked about XSLT with presentation suffix because I didn't manage to find them in our metanorma-* processors, maybe @opoudjis can comment on this also.

[opoudjis]

Yes, I rename the *.presentation suffixes to strip them when I copy them into the individual gems.

There are indeed rules for allocating the multiple XSLTs of a flavour; they are often but not always doctype. The rules are implemented in the lib/isodoc/*/pdf_converter.rb files.

[Intelligent2013]

I propose to remove non-presentation xslt in mn-native-pdf, and rename *presentation.core.xsl into *core.xsl (+ update Makefile).

[Intelligent2013]

@CAMOBAP I'll do it in feature/update-xslts-in-processors branch if do you mind.

[CAMOBAP]

I propose to remove non-presentation xslt in mn-native-pdf, and rename *presentation.core.xsl into *core.xsl (+ update Makefile).

@Intelligent2013 Good idea!

@CAMOBAP I'll do it in feature/update-xslts-in-processors branch if do you mind.

Yes please, just let me know once you are done

[Intelligent2013]

@CAMOBAP Done.

[CAMOBAP]

It works!

Several commits were pushed:

• metanorma/metanorma-itu@eb52782
• metanorma/metanorma-gb@d9ec531

For itu new XSLT files was added, so I prepared PR for it metanorma/metanorma-itu#273

Closing this one