This is an old revision of the document!

The priority for a web translator should be chosen according to the following guidelines:

• 100 = specific journal/website (target should be a specific domain)
• 150 = multi-journal sites with different subdomains, but all journals would look practically the same
• 200 = multi-journal sites with different subdomains and some journals may have somewhat different layout
• 250 = CMS/Library Catalog with no specific domain, but either a very unique URL or unique checks in detectWeb (for all possible detection options). i.e. unlikely to be a false-positive
• 260 = CMS/Library Catalog with no specific domain, but somewhat unique detectWeb checks (mostly URL-based checks).
• 270 = CMS/Library Catalog with no specific domain and rather generic URL checks (conceivably, could be a false-positive)
• 300+ = Translators to be run on every page. Priority numbers should be in order of metadata quality.

(see https://github.com/aurimasv/translators/commit/5b4ef451a9c3c14fa9b2b8d5bce37a8bf7c1a3f2 for more information)

Moreover, the priority is very important for search translators if we have multiple translators for and identifyer (e.g. ISBN or doi). It's somewhat relevant for import translators - especially the various RDF versions which need to be run in the right order. It doesn't apply to export at all.