cubicweb-seda #11447416 generate schema from xsd file [validation pending]

We should be able to generate the schema from the XSD file, with the following tweaks:

  • agent related information will be generated from agent definition from the saem_ref cube
  • extension points will be discarded at that point
done in0.1.0
load left0.000
closed by<not specified>
patch[xsd2yams] cleanup XSD2YAMS datastructure and fix mapping for DataObjectRefType [applied][xsd2yams] refactor to clearly separate code generation from intermediary representation. [applied][xsd2yams] extract possible values for KeywordType in a csv file. [applied][xsd2yams] store description level as concept instead of string [applied][xsd2yams] fix handling of attributes defined in parent class [applied][xsd2yams] store storage and appraisal code as concept instead of string [applied][xsd2yams] fix simpleType handling [applied][xsd2yams] add constraint to match particular scheme for related concept [applied][xsd2yams] skip still more mandatory elements [applied][xsd2yams] properly handle cardinalities [applied][xsd2yams] map PhysicalId to string instead of concept [applied][xsd2yams] skip ArchiveUnitRefId / SignedObjectId [applied][xsd2yams] remove agent related information by directing writer/signer/validator to Agent [applied][xsd2yams] skip more mandatory elements [applied][xsd2yams] remove useless conditions [applied][xsd2yams] fix ArchiveUnitRefId handling [applied][xsd2yams] no user_cardinality attribute if there is a single possible choice [applied][xsd2yams] fix crash when generating the whole schema, introduced by 735e8b7ee50a [applied][xsd2yams] don't skip data object reference id anymore [applied][xsd2yams] don't skip attachment anymore [applied][xsd2yams] fix signature elements handling [applied][xsd2yams] don't skip custodial_history_file and data_object_reference anymore [applied][xsd2yams] don't skip restrictionValue anymore [applied][xsd2yams] don't skip unit and weight anymore [applied][xsd2yams] don't skip language anymore [applied][xsd2yams] skip Filename [applied][xsd2yams] skip various elements supporting custom metadata according to the media type [applied][xsd2yams] skip some BinaryDataObjectType sub-elements and reintroduce message digest entity [applied][xsd2yams] test and fix support for some attributes of CompressedType [applied][xsd2yams] dispatch tests on erroneous elements given to collect in the callee [applied][xsd2yams] merge _collect into collect makes a simpler api [applied][xsd2yams] skip KeywordContent element [applied][xsd2yams] start skipping mandatory elements [applied]backport a fix for too long index name. Related to #11447416 [applied]add basic code generation that generate a 'compilable' schema [applied]