Property talk:P6394
Documentation
identifier assigned by the ELNET consortium of Estonian libraries. Format: 7 digits, "x" or another digit
List of violations of this constraint: Database reports/Constraint violations/P6394#Single value, SPARQL
List of violations of this constraint: Database reports/Constraint violations/P6394#Unique value, SPARQL (every item), SPARQL (by value)
List of violations of this constraint: Database reports/Constraint violations/P6394#Scope, SPARQL
Pattern ^([0-9]{7}[0-9x])$ will be automatically replaced to a\1. Testing: TODO list |
|
After formatter URL edit each link using it returns 404 Document not found
edit@Jahl de Vautban: after your edit https://www.wikidata.org/w/index.php?title=Property:P6394&diff=prev&oldid=2191167993 each link using the formatter URL returns 404: Document not found OCLC SRW/SRU Server Friedrich Schultz (talk) 06:45, 12 October 2024 (UTC)
- @Friedrich Schultz: thanks for the ping. My edit back then was to fix what you got otherwise with a11174882. However things have changed yet again and now the initial
a
seem to have been droped from the identifer as 11174882 correctly resolve. If @Epìdosis: can confirm my analysis I can launch a batch tonight to remove thea
on every item. —-Jahl de Vautban (talk) 08:10, 12 October 2024 (UTC)- @Jahl de Vautban: I confirm your analysis: on October 6 VIAF replaced ERRR IDs starting with "a" with IDs without "a"; I have now consequently adapted the property removing "a" from the examples and from the constraints; I have also enabled an autofix that will fix in a few days all the existing IDs (it is better to use the autofix instead of QuickStatements, because removing and readding the IDs would lose their references, which are not fundamental, but I think it would be anyway preferrable to keep them). --Epìdosis 08:17, 12 October 2024 (UTC)
- @Epìdosis: The IDs have changed back (again). Late on 2024-12-21, OCLC deleted ERRR IDs (lacking an initial
a
) from VIAF clusters, and replaced them with ERRR IDs containing an initiala
, early on 2024-12-22. Can the autofix be changed? Please see the history of World Trade Organization (Q7825), and the histories of VIAF clusters 312770601, 12149542594800300759, 6149542952400302734, and 40830893, as examples. Thanks. -- Dcflyer (talk) 02:52, 2 January 2025 (UTC)- @Dcflyer: Autofix inverted, it should work. Epìdosis 08:58, 2 January 2025 (UTC)
- @Epìdosis: The IDs have changed back (again). Late on 2024-12-21, OCLC deleted ERRR IDs (lacking an initial
- @Jahl de Vautban: I confirm your analysis: on October 6 VIAF replaced ERRR IDs starting with "a" with IDs without "a"; I have now consequently adapted the property removing "a" from the examples and from the constraints; I have also enabled an autofix that will fix in a few days all the existing IDs (it is better to use the autofix instead of QuickStatements, because removing and readding the IDs would lose their references, which are not fundamental, but I think it would be anyway preferrable to keep them). --Epìdosis 08:17, 12 October 2024 (UTC)
So:
When | formatter | example result not encoded | example result encoded |
---|---|---|---|
2019-01-24 [1] day of property creation | https://viaf.org/processed/ERRR|$1 | https://viaf.org/processed/ERRR|a11174882 | https://viaf.org/processed/ERRR%7Ca11174882 |
2024-06-29 [2] value with a | https://viaf.org/processed/ERRR$1 | https://viaf.org/processed/ERRRa11174882 | https://viaf.org/processed/ERRRa11174882 |
2024-10-12 [3] value without a | https://viaf.org/processed/ERRR$1 | https://viaf.org/processed/ERRR11174882 | https://viaf.org/processed/ERRR11174882 |
@Jahl de Vautban: "re My edit back then was to fix what you got otherwise with [https://viaf.org/processed/ERRR%7Ca11174882 a11174882]" - what did you got?
@Epìdosis: 1) re "I confirm your analysis" - the following text does not match - "on October 6 VIAF replaced ERRR IDs starting with "a"" - the formatter edit didn't remove "a". 2) Wherefrom did you take "October 6"? Friedrich Schultz (talk) 18:21, 12 October 2024 (UTC)
- October 6 comes from the history of a few VIAF clusters that I have checked randomly, e.g. https://viaf.org/viaf/12149542594800300759/: in all cases it was visible that on October 6 the ID starting with "a" was removed and the ID without "a" was added. The formatter URLs with and without | both work as of now (https://viaf.org/processed/ERRR11174882 and https://viaf.org/processed/ERRR%7C11174882); "a" has been removed not from the formatter URL (which, I confirm, did not contain it) but from the ID themselves. Epìdosis 07:56, 13 October 2024 (UTC)
- @Friedrich Schultz: ah interesting, probably the
|
isn't understood the same way by all browsers and some aren't/weren't able to process it, resulting in an error saying that the URL contained an unauthorized character. I'm not able to reproduce the error on the computer though (it does appears from time to time on mobile), and double checking on a cluster confirm that there is indeed this|
, though it doesn't appear to be mandatory for the URL to resolve; anyway, I have reverted my edits. --Jahl de Vautban (talk) 18:35, 15 October 2024 (UTC)
- @Friedrich Schultz: ah interesting, probably the
Duplicates due to different script
editWhilst a few years ago ERRR used to have separate authority records with the preferred form (MARC21 field 100) in Latin script or in Cyrillic script, the situation has changed firstly on 2021-12-05 (https://viaf.org/en/viaf/10381604) and secondly on 2022-04-10 (e.g. https://viaf.org/en/viaf/22655969), when ERRR deleted these duplicates, keeping the authority record with the preferred form in Latin script and merging the other one into it (the resulting authority record has the Cyrillic form as variant form in MARC21 field 400). Thus the great majority of single-value constraint violations (presently 2058, cf. https://w.wiki/EAuv) can be easily solved. --Epìdosis 11:49, 17 May 2025 (UTC)