Process order

[SbstnErhrdt]

5.4. Sequence of Processing
In August 2015 the strategy in the way that the records in one exchange delivery are to be processed
has been changed – in order to resolve issues with the stability of the “doc-id” – the unique and stable
identifier to the publication-identifier that is supported at level. This identifier
is expected to remain stable throughout the entire life of a publication. There will however be
circumstances under which the value of the “doc-id” may fluctuate for a given publication-identifier.
These circumstances will only present themselves in a relatively limited number of scenarios, in the
case of complicated issues resolved manually – by intellectual intervention.
The new strategy is to cater for scenarios where:
• the same publication-identifier is in the exchange TWICE – with different values in “doc-id”
• the same doc-id is in the exchange TWICE – with different publication-identifiers
once with status=’D’ – once with status=’C’ or status=’A’
To allow for managing these scenarios in a transparent way, the following solution has been opted for
in cases where one publication-identifier is in the exchange twice:
• concentrate records with status=”D” in a dedicated package
-DeleteRekey-PubDateAndBefore-*
• include this dedicated package in the CreateDelete package
-CreateDelete-PubDateAndBefore-*
• sequence of processing must be:
process records in -DeleteRekey-PubDateAndBefore-*
before records in -CreateDelete-PubDateAndBefore-*
before records in -Amend-PubDateAndBefore-*

https://link.epo.org/web/T09.01%20ST36%20User%20Documentation%20vs%202.5.8.pdf