Trados Studio Plugin to connect Term-Resources from translate5

In large parts this is a Trados specific development.

Based on the "Terminology Provider API" of Trados Studio a Trados plug-in is developed, that 

• can connect translate5 termCollections in a similar way as Multiterm server DBs
• needs for the connection
  • the translate5 server-URL
  • a user name and password. This user must not have API rights in translate5, but 
  • after this is inserted, it is saved in Trados for future usage within this project 
  • translate5 is queried for all existing term-resources that can be accessed by this user. The resources are listed and the user can select them and add them as term-resources for the current Trados project (analogous to Multiterm server DBs).
• on opening a file for translation 
  • sends at first the first few segments for tagging to translate5 and gets back the terms (in stemmed form, like they exist in the segment) and referenced by ID the actual terms and definition as they are in the term-resource (see http://producthelp.sdl.com/SDK/TerminologyProviderApi/2017/html/ba9488ad-d5bb-44eb-911a-fa4bf5af3f75.htm)
  • later sends packages of 50 segments in batches to translate5 for tagging and receives them back
  • saves the result in a cache file for fast access. 
• on opening a segment by a translator
  • asks the cache file for tagging of the terms
  • if the cache file does not contain the segment so far, directly asks the translate5 API
  • if translate5 is not reachable for some time or delivers an error, the user is shown the error in Studio

The translate5 API is extended to receive batches of segments for tagging them against one or more term-resources and delivers back REST-objects in the way needed by the above linked search-query of the Trados API. All other necessary REST-calls already exist in translate5. In addition a user with the role "Term search" must be allowed to access the translate5 API (get a session) and browse and tag terminology in the above described way (but nothing else via API).