Serena Marangoni wrote:
A client gave me a few documents (user manuals, lots of repetitions) so that I can craft a quote.
I've been given a copy of a previously translated manual, but it's a physical copy, not digital. If the client accepts the quote I'll OCR these and align them to prepare the TM. Right now I only have a digital file of both originals (the one I need to translate, and the one that was used as the source for the translation that's already done). Is there a way to create a "fake TM" (from the original of the one that's already translated) that allows me to count the content that actually needs to be translated?
Create a fake TM
Open the original source file in your favorite CAT tool
Copy source to target
Confirm and commit the segments to the TM. You now have a TM filled with the sentences of the original, it doesn't matter what's in the target.
Close the original source file
Open the new source file
Run the analysis against the TM populated as above. You'll have the analysis of the new source file with matches from the original source file taken into account.
Don't forget to charge for the time needed to populate the TM, OCR, align, etc. It does take time.