most oddDecl elements must be descendant of <schemaSpec> or <specGrp>#2855
most oddDecl elements must be descendant of <schemaSpec> or <specGrp>#2855sydb wants to merge 7 commits into
<schemaSpec> or <specGrp>#2855Conversation
issue #2306. Still have to write tests.
elementSpec-should-not-be-in-head.
|
With @sydb and @HelenaSabel : We tested this and determined that there are 981 cases that violate these constraints within p5subset.xml (spec elements are children of divs there). Therefore we chose to postpone this merge until after the February release, so that we can also add wrapping specGrps to make p5subset.xml valid (for which another ticket already exists), and check that there's no fallout from that change in the Stylesheets ODD processing. |
|
Coming back to this: it seems that there are a lot of irrelevant changes to the way the XInclude namespace is declared and used in the PR, which is a bit confusing. @sydb Was there any intention here, or is this just meaningless fallout from the use of an XML parser/serializer that does things differently? Could we remove those changes from the future PR to make it easier to evaluate? |
4 participants
Address #2306 by:
<classSpec>,<constraintSpec>,<dataSpec>,<elementSpec>,<macroSpec>,<outputRendition>, or<specGrpRef>is not inside a<schemaSpec>, a<specGrp>, or a<egXML>(due to testing-swaps-namespace thing).<specGrp><elementSpec>-inside-<head>)In 9–10 months time the deprecation warning should become an error.
Note: This is currently a DRAFT PR, as I have not yet added a note to Appendix G (deprecations). Hope to get to that tonight, but figure the rest can be looked at without that.