Closed chenejac closed 3 years ago
Benjamin Gross said:
Perhaps this has been discussed, but another reason why this functionality is important is because if all language files are moved to firsttime (to enable editing in the UI), a language would have to be enabled in runtime.properties upon the first application startup to ever be enabled, since RDFFilesLoader skips RDF for languages not enabled. Sounds like this functionality would address that complication.
Andrew Woods said:
Agreed
Dominik Feldschnieders said:
[PR|[https://github.com/vivo-project/Vitro/pull/195]]
As I have already mentioned I had to adjust the procedure slightly. The case "else, if they differ, leave user's triplestore statement alone", I think would lead to problems if i activate a new langauge with firsttime files like we have in tbox (for example vitroAnnotations_en_US.n3). Now i check if some of these new/old triples are not or still in the user models and if so, i added/removed them.
One more thing: Since I have to load the firsttime files for the check, the log now always contains the output that these files are reloaded (from class RDFFilesLoader). Should I adjust the loading method in this class so that this does not happen? Or is the current output, which says that the firsttime check starts, enough?
Andrew Woods said:
Finished code review (see comments in GitHub), now for functionality testing.
William Welling said:
I have tested a few scenarios and it works as expected. Added details as an attachment.
Andrew Woods said:
See code and functionality review commits on pull-request.
Andrew Woods said:
[~accountid:6113d5cd4e8d8d0069f2e486]: In order to make testing of your pull-request... doable, I have resolved the above mentioned bug with: https://jira.lyrasis.org/browse/VIVO-1944
Dominik Feldschnieders said:
I added a commit to the PR. I have fixed the bug, so changes in the UI should no longer be overwritten by the changes from the files. But I'm not sure how to configure the backup-models so that they wont be queried via the SPARQL-API. I added an example in a comment in the PR.
Andrew Woods said:
Thank you, [~accountid:6113d5cd4e8d8d0069f2e486]. I will give it a look in the coming week. As far as you know, does having the backup-models show up in SPARQL-API queries have any negative effect on the running application?
Andrew Woods said:
[~accountid:6113d5cd4e8d8d0069f2e486]: I will need to confirm, but during yesterday's dev meeting there was suspicion that the backup models are visible in SPARQL-API queries because they may be saved in the content-triplestore instead of the configuration-triplestore. If your pull-request is not already doing so, we will need to put the backup-models in the configuration-triplestore.
Dominik Feldschnieders said:
Okay, i found the problem with that. The content-triplestore is probably the standard, now they are stored in the configuration-triplestore.
I have performed additional tests and found two issues, this first one was with later added langauge files where the "same" label but with different language tag were changed in the UI. This one is already fixed.
The second probably has nothing to do with this ticket. Can someone confirm that with two available languages after a firsttime startup, the page "Site Admin -> Page management -> Click the pencil icon next to the 'Capability Map' row" or any other page on this site, is not fully translated when the language is changed? I see this behavior also in the i18n-sprint branch.
Andrew Woods said:
The second probably has nothing to do with this ticket. Can someone confirm that with two available languages after a firsttime startup, the page "Site Admin -> Page management -> Click the pencil icon next to the 'Capability Map' row" or any other page on this site, is not fully translated when the language is changed? I see this behavior also in the i18n-sprint branch.
Yes, I see the same issue. There are several outstanding bugs, particularly in the siteAdmin pages: https://jira.lyrasis.org/issues/?filter=14602
Dominik Feldschnieders said:
I have adjusted the last things as requested in the PR, it should be ready now.
This issue depends on #1833
Andrew Woods (Migrated from VIVO-1918) said:
The data associated with firsttime files is found in three locations:
This task is to reload the RDF found in "firsttime" files per the following algorithm on start-up: