Translate
Although we’ve been working with this batch task for a number of years, now and then it still seems that ‘incorrect’ (i.e. not the most recent, for example) translation units are finding their way into updated versions of previously translated files, and I’m trying to understand the exact processes that Studio goes through here, particularly Apply PerfectMatch in combination with Pre-translate. The following scenario outlines how we work, and I’d very much appreciate any suggestion to the below questions…
Scenario:
We receive files from the customer to translate. The project contains a customer main TM (MTM) and a subordinated project TM (PTM). The external translator works into the PTM using the MTM as a lookup. The files are checked in-house with the PTM deactivated and updating the MTM. The customer then sends an updated version containing changes to the previously translated files:
A: Assuming that the initial file has been proofread in-house, we add this to the existing project using the Prepare task sequence with both TMs enabled. In the Apply PerfectMatch step, we select Apply PerfectMatch and lock. The previously translated and checked segments are locked and we can work on the new segments.
B: If the intial file hasn’t been proofread, we also add this to the existing project using the Prepare task sequence with both TMs enabled, but in the Apply PerfectMatch step, we select Use original translation origin and status. The proofreader then checks or amends the segments from the translator and sees possible TM matches.
- Given that Apply PerfectMatch uses a previously translated document and not a TM, does it make a difference using the Prepare or Prepare without project TM task sequences when adding the changes versions to the project?
- As above, does it make a difference whether the PTM is enabled when using the Prepare task sequence?
- According to the Studio 2015 product help, PerfectMatch translations are overwritten by Pre-translate: a) Is this only the case when the status is set to Use original translation origin and status (because the PerfectMatch segments have been locked and are ‘safe’)? b) Do the ApplyPerfect Match and Pre-translate tasks run in parallel or in the order that they occur in the task sequence (i.e. could changing the order to pre-translate first avoid PerfectMatch segments being overwritten)?
- We frequently get an error message (I’m afraid I don’t have it to hand) when processing these updated versions containing changes using the above-mentioned task sequences. The workaround that we’ve found is to run the Analyze Files task first and then Apply PerfectMatch. The problem with this is that the workaround only works running the tasks in this particular order (i.e. you can’t ApplyPerfectMatch first and then run the Analyze task) so we don’t know how many characters/words are actually PerfectMatches. Is there a known conflict here, or what might be the cause?
