Bibcheck: Adds a temporary plugin

[GiorgosPa]

• Corrects the authors from some old atlas papers

[jalavik]

ping @jmartinm we should review this.

[jalavik]

@GiorgosPa Can you rebase this and remove/comment out the added entries in rules.cfg (to avoid activating the rules on PROD). Then I will merge this.

[kaplun]

Sorry to join late to the party :smile:. I haven't seen this PR... I think it would be nice to rename the atlas check to something more generic (and actually could be almost something to include into Invenio). In the end this is going to be a typical problem not just related with ATLAS but also other use cases (plus there is no atlas-specific filter in the configuration).

E.g. could be named author_typos.py or something similar, and also the various CFG_ATLAS thingy could be renamed to remove the word "ATLAS".

[jalavik]

Yeah, @GiorgosPa and I had a quick chat about this yesterday and we agreed to kill the KB and config variable parts actually. Instead turn it into a batch rule check_records(), that receives the source file for the mappings as parameter, then build the mapping once in memory and then replace all the values needed over each record - without the need of using a knowledgebase.

@dset0x This is a use case to think about actually for the new ~~matcher~~ checker, regarding "setting up" a rule (pre-loading of stuff etc.) once and run it on all things..

[jalavik]

That said.. for the generic case, maybe a stable author typos mapping in a KB would make sense, but indeed the check should not worry about creating and filling the KB then. Just check if it is there or not.

[GiorgosPa]

I updated the PR so that no configuration variable or database knowledge base is used. I also think that the dictionary building will happen only once what do you think @kaplun ?

[kaplun]

Thanks @GiorgosPa, so just one small note for me, but beside that, I think the plugin will now no longer have performance issues :+1:

[GiorgosPa]

Updated.

[jalavik]

@kaplun Do you think this one is okay to merge?