Trados + HTML codes
Thread poster: Ivana de Sousa Santos
I've got the following "problem" (it is to me, but I understand if you say "duh" by reading this).
I received a translation yesterday and there were 5 files to translate of around 10,000 words.
Two months ago I translated for the same agency the same kind of document (a website) and now they sent me the proofread TM to work on it (it's the same website).
Their analysis with Trados revealed a certain amount of new words (about 3000) and the rest are repetitions or X% match.
Up to here everything is OK for me.
I started the first file and everything was OK. All I had to do was practically alt+ because almost the whole text was in the TM.
The 2nd and 3rd files were OK too (they were actually very small; about 1 page)but as for the other ones I noticed there were words I couldn't tell what they were. I panicked, also because it was late and I could only speak to the agency's PM in the morning.
After spending a while reading the source texts (which are in French) I started figuring out what the words meant. For instance the word in the French text "précédente" meant "précédente" and the "é" stood for "é" and so on.
What did I do for the two next hours?
I changed every single word of the source texts that was like this (the text also had red codes and the time before I was told I should leave them as they were... and so I did).
I just couldn't figure out 2 words.
Then, I stayed up until 4:30 a.m. (because today I had an appointment at a doctor in Lisbon and knew I would be out the whole day and couldn't translate) doing my translation and I could finish it all (without revising it, of course; the deadline is Thursday afternoon and as I would be out today I left the revision and document comparison for tomorrow).
In the morning, before leaving to Lisbon I wrote an e-mail to the agency's PM (also telling him I wouldn't see his answer until I came back home at night) stating that the texts were not well written, that I spent 2 hours editing them, that I hoped next time they wouldn't be like that and asking him to clarify me the 2 words I couldnt' figure out.
I was astonished when I got home this evening and saw his answer: "the words are not wrong, they are in HTML code, that I should respect those codes and that they should be left the same".
I am a bit confused now: some of the sentences were already in the TM as 100% or other % match and with those HTML codes they have got no match.
I never worked with HTML codes.
My first question is: is it normal to receive a second translation with HTML codes when the previous one had none? (I'm not questioning the analysis done by the agency)
Second: Trados doesn't seem to recognize sentences that were translated before. Is it how it works?
Third: wasn't I supposed to change the French text so that some sentences would match?
Fourth: can I leave it as it is now? Or shall I redo the translation (on a copy/paste basis from what I have so far) and work on the original TM they sent me and not the one I have now and the text with words such as "précédente" and so on?
Fifth: did I misunderstand what the client wrote regarding my e-mail about this words or did he misundertand mine?
Sixth: the words I couldn't figure out (as I did with the others): won't I have to translate them? If I leave them like that, wouldn't they be in French afterwards in the middle of a Portuguese text??? (The PM told me one of the words stood for Nº, which is the same in Portuguese, but the other one "…" I don't even know what it is in French to translate into Portuguese).
They may be silly questions for those who know HTML codes but I had never seen it before (only heard about it and also was told not to change the HTML codes in red... now even the black words are like that). I didn't know there could be HTML codes "inside a word".
I would really appreciate to read some of your comments to know what to do now regarding the translation and if I shall answer the agency PM or not.
Many thanks in advance.
| || || |
| | Hynek Palatin
Local time: 05:42
English to Czech
The problem is that in the source text, some special characters were encoded as entities, not inserted directly. More on entities (see the last three links for a complete reference):
The TM contained the sentences with unencoded characters, so Trados couldn't find a match. It's a problem of the source files - they should have been prepared the same way as before. This way it's more work for you - but it's paid and not so hard: You can save the source file as HTML and open it in browser (in TagEditor, you can display Preview). Now, the HTML codes should be displayed as the special characters and you should be able to copy the sentences and find the translations in TM using concordance.
By the way, you were probably trying to give examples of the codes. As you can see, there are no codes in your message and they have been replaced by the characters themselves.
| Thank you Hynek... || Jun 21, 2006 |
... for the reference file you mentioned in your posting and your answer as well.
And yes, I was trying to give an example and I now realize it didn't work.
To report site rules violations or get help, contact a site moderator:
You can also contact site staff by submitting a support request »