enrich ISBNs (Alma)

[hagbeck]

Currently Alma exports only the ISBNs which are in the origin MARC record.

Please use morph functions <isbn to="isbn10"/> and <isbn to="isbn13"/> to ensure that both versions of ISBNs find their way in the lobid-resources.

Example: http://alma.lobid.org/resources/990030520380206441.json

[dr0i]

If an isbn10 exists a corresponding isbn13 is generated.
Also, an isbn is normalized by removing semnatic sugar, so that:

-  "isbn" : [ "3-642-32079-1" ],
+  "isbn" : [ "3642320791", "9783642320798" ],

Going to deploy.

[dr0i]

Deployed, see e.g. #1184.
Closing.

[dr0i]

c6e59ec introduced duplicate isbns. This cannot be seen in the test data (https://github.com/hbz/lobid-resources/blob/master/src/test/resources/alma/(DE-605)HT019246898.json) but in the new index (http://alma.lobid.org/resources/search?q=HT019246898&format=json). That's because the tests are not working atm (see #1124) and we have to check manually by doing a mvn failsafe:integration-test -Dit.test=AlmaMarc21XmlToLobidJsonTest; git diff.
Reopened.

[TobiasNx]

I assume that the source data already has ISBN-10 and ISBN-13. It would make sense to add the <unique />-function at the end of each transformation.
Also this seems not to be a problem of the testing but of working multiple branches at the same time. The test file you refer to (https://github.com/hbz/lobid-resources/blob/master/src/test/resources/alma/(DE-605)HT019246898.json) does not exist in the 1180-enrichIsbn branch. It was introduced in the 1171-incorrectCurrentLocation. (https://github.com/hbz/lobid-resources/commits/master/src/test/resources/alma/(DE-605)HT019246898.json)

Due to the merging of multiple PRs at the same time the test data does not reflect all changes but only those that are pushed with the specific branches. Perhaps we should update the testfile after merging the master in the feature branch before merging the PR.

[TobiasNx]

I updated the morph and open a PR now.

[dr0i]

Indeed, the test resource HT019246898.json was added at a later stage. It proves to be a very good test case also for isbn, because this test case (a resource having both, isbn10 and isbn13) was not present before.

Due to the merging of multiple PRs at the same time the test data does not reflect all changes ... Perhaps we should update the testfile after merging the master in the feature branch before merging the PR.

Yes, in this case it's true (working tests would come in handy here ;) ).

[dr0i]

Deployed, see http://alma.lobid.org/resources/search?q=HT019246898&format=json. Closed.