Closed matyaskopp closed 7 months ago
Was dev merged into data / main? Because there were changes to the schema to accommodate CHES in devel.
Was dev merged into data / main? Because there were changes to the schema to accommodate CHES in devel.
yes, devel is both in the main and data branches: https://github.com/clarin-eric/ParlaMint/commits/data
Mixture of state
and label
is not allowed in TEI: (https://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-state.html)
<content>
<sequence>
<elementRef key="precision" minOccurs="0" maxOccurs="unbounded"/>
<alternate>
<elementRef key="state" minOccurs="1" maxOccurs="unbounded"/>
<sequence>
<classRef key="model.headLike" minOccurs="0" maxOccurs="unbounded"/>
<classRef key="model.pLike" minOccurs="1" maxOccurs="unbounded"/>
<alternate minOccurs="0" maxOccurs="unbounded">
<classRef key="model.noteLike"/>
<classRef key="model.biblLike"/>
</alternate>
</sequence>
<alternate minOccurs="0" maxOccurs="unbounded">
<classRef key="model.labelLike"/>
<classRef key="model.noteLike"/>
<classRef key="model.biblLike"/>
</alternate>
</alternate>
</sequence>
</content>
Drat, missed that one. It doesn't show up in the build, as I only validate against ParlaMint RNG schemas. The worst is that we could manage without label, as it only encoded the CHES name of the party.
I can't fix this now, as it would mean recompiling all the corpora from scratch (CHES/Wiki/ministers get inserted firs, then all the rest of the build). I guess this means we will have to live with failing samples and non-TEI conformant corpora for 4.0. Sorry about this.
org//state/label
is still in relaxNG schema:
https://github.com/clarin-eric/ParlaMint/blob/f7170e567d94acb50c0ccd9956d8ad9a7a1c1524/Schema/ParlaMint.rng#L200-L204
OK, this has now been fixed, as evidenced by new conversion log files. I also removed all Samples (as they are failing with the new schema) and started adding new samples as they become available. So, closing.
not sure what happened, only countries with CHES fail