Node to set the funder acronym for transformation is not usable together with TDS and transformator service. The funder acronym must be cabled into the transformation rules. For Croatian projects we must create 2 distinct TDS
Allow assignment of INGEST wf to aggregator of data repos
fixed wf template for datacite OAI sets
removed old transformation for DOAJ journal list
Mappings for re3data and opendoar are now TDS rules and not anymore xslt in the classpath
Removed old XSLT for doaj article titles: the relative TDS rule must be used instead.
Mapping for journaltitles and irdb are now TDS rules and not anymore xslt in the classpath
Incremental transformation for Datacite OAI sets
Deleted old xslt: now that we have incremental transformation we have to use TDS rules
fixed wf template
context for miur funder
Funder wfs now use xslt as TDS rule for transformations
using dsm_datasources in sql expressions
fixed arc target node name
To support incremental update of projects in the database, useful functionality of IncrementalTransformationJobNode has been extracted to a superclass. Wf template for entity registries has been updated accordingly. Wf templates for other types of ds that were already in incremental mode have been updated for the new param name inherited from the superclass.
RCUK: handle cases when the /funder/name is blank
Incremental transformation also for entity registries
Added wf parameter for dashboard visibility
merged incremental transformation to trunk
Re-introduced <FIELD name="_dnet_resource_identifier_">: it is required if we want to enable the updates on the database.
fixing mappings for datasource subjects, languages, contenttypes
using dsm_datasources instead of datasources table, fixed datasource data migration rules for languages and od_contenttypes
modified context and transformation for Academy of Finland in accord with #3119#note-5
removed column 'trust' from re3data mapping
fixed params in datasource entity reg aggregation workflow
merged branch dsm into trunk
Adjusted trust levels to give priorities to organizations from re3data and OpenDOAR rather than EC databases. Among corda and corda h2020, the latter wins.
Fixes #3313: record trust is now a user parameter.
Pangaea dataet: identifiers must be plain DOI (not URL) or the portal generates a wrong link.
change the funderId from DFGF to DFG
re-introduced persons for prod. New file created with the no-person mapping for h2020
revert previous commit: opendoar was not affected
ticket #2955: ensure entity registry set the proper provenanceactionclass for the entity they create
transformation for dfg funder
context for dfg funder
XSLTs for pangaea removed from the datacite folder. They must be only in their dedicate folder of the classpath.For the pangaea by project, the format of the id slightly changed and the mappings have been changed accordingly
commented new field contactfullname in table projects. To remove the comment when new db version in beta
context for Academy of Finland funder
transformation for Academy of Finland funder
transformation for sgov funder
trasformation for sgov context generation
more accurate mapping for contactfullname
avoid to produce empty xml records when a 'canceled' project is found, it breaks the workflow. Fixed also the organization identifier costruction and the separator character used to parse column 'project number'
normalise spaces when extract project contact info
project contact person moved in project's description
getting rid of person entities #2893
changed to match the change in the db
change for upgrade in the db
implements #2385#note-66: FWF projects whose status is canceled must be skipped.
changed the name and description funding stream level 1
fix bug in date format
trtansformation for conicyt context
transformation for conicyt funder
fixed nsPrefix length
Context for Tubitak funder
Transformation for Tubitak funder
Properly handle dates with xslt 2
SFI OA mandates starts from 2009 (#1506#change-23022)
Fixed WT rule for OA mandate and project aliases
WT projects have OA mandate if they started after 2015.
set the datasource prefix
integrated (hopefully) all required changes from dnet40
OpenAIRE mdbuilder basically does nothing if the record has already an oaf:entity (for claim of records already in OpenAIRE)
Do not generate a new objIdentifier if there is one, even if it is not in the dri namespace (for claims, see r47589 )
Claims metadata are collected from the file system, not from the database.
Let's forget the repositoryId
Corrected description of node
codebase used to migrate to java8 the production system