problem

In t5memory it still may happen that a TM is "damaged" and must be reorganized.

clarify with Orest how a TM will respond in such a case. Currently it returns:

[error] => stdClass Object
        (
            [method] => POST
            [url] => http://bla:1984/otmmemoryservice/TMID/fuzzysearch
            [code] => Error Nr. 5008
            [error] => EQF0399: The dictionary or Translation Memory 'NAME' needs to be organized; it is either damaged or it was created with an older version of OpenTM2.
 Please use the organize function to organize the dictionary or Translation Memory.
        ) 

solution

1. If we get such an error, we should trigger automatically the reorganize endpoint
2. Clarify with Orest what happens when t5memory gets the reorg call multiple times on a tm while it is reorganizing. If that is a problem we have to ensure that we send it only once.
3. In addition we need a CLI command t5memory:reorganize to trigger reorganize on a specific TM
4. as goodie: t5memory:list which lists all memories and its status

• Automatic reorganize, clarify multiple calls or ensure only one at a time
• Automatic reorganize should be triggered by certain error-codes of t5memory. What error codes should trigger reorganize, should be configurable via system configuration of level 1 (not configurable in the GUI). For now error codes 5008 and 5019 should be on the list.
• Auto reorganize should be only triggered once for a TM, afterwards it should automatically be checked, if the TM is now usable again (by sending a fuzzy request and a status request to it). If it is still not usable and returns an error instead of a valid result, translate5 should itself throw an error.