CafeTran

Just encountered the following issue: Whenever I open the drop-down menu with the keyboard shortcut to insert placeables/non-translatables, the placeables are inserted correctly only if I insert them one after the other in exactly the same order as in the menu. However, if I insert for example item #1, then #2, and then #6, CTE inserts the wrong number. Same with other items at the bottom of the list.

Furthermore, some items on the list are not in parantheses, and not all numbers are shown on the list (seven instead of eight). This is annoying as I have to change the word order/sentence structure, and thus the order of the numbers, fairly often. Could this please be fixed? Thank you in advance!

That problem I described occurs with 10.8.3. On my other computer running 10.8.1 there is no such issue.

You are right, I just noticed this, too.

Do you have any news on this? Will this be fixed in the next update? Thanks!

Hi Igor, many thanks for your efforts. I am afraid the issue is not completely resolved, though... See screenshot attached. All the non-translatables in the list should be in parantheses, just like in the source text.

What is more, now, when inserting a non-translatable from the list into the target text, the non-translatable is not removed from the list anymore. Before the update, inserted NTs used to disappear from the list, which was very convenient, especially with segments having many NTs. Would it be possible to have this back, that the inserted NTs are no longer shown on that list? Thanks a lot! :)

Great, thanks! :)

There is still an issue with the way CTE handles non-translatables. I use a non-translatables-only glossary filled with regex to catch all kinds of numbers and number-letter combinations in parentheses. This is very convenient as it allows me to quickly insert even large sets of numbers in parentheses simply by pressing a shortcut followed by enter. I have more regex in that NT glossary that handle dimensions such as "0.3 mm" etc.

With version 10.8.1 all this worked perfectly, see screenshot 10.8.1 as an example: All single numbers and sets of multiple numbers in parentheses are correctly displayed in the list, one after the other as in the source text.

With CTE 10.8.3 something has changed, as can be seen in the screenshot:

- whenever a number or set of numbers in parentheses appears more than once, only the first instance of this number or set in parentheses is displayed correctly, all others are broken and displayed as individual numbers without parentheses.

- Dimensions such as "0.3 mm" are no longer recognized at all.

In 10.8.3 it is still somehow possible to select the right NT fragments, because if one selects and inserts a first instance of a duplicate fragment, the second instance of that fragment (that was previously displayed incorrectly) is, by some magic, suddenly shown correctly and insertable.

Has anything changed in the way CTE handles regex in such NT glossaries? It would be good if CTE could display NT fragments as in version 10.8.1 again. The way it is now in 10.8.3 is quite confusing; it actually looks as though the regex is bad, while it is not because it works perfectly in 10.8.1. The fact that "0.3 mm" is not recognized, while "90 °C" is, is also strange.

Thank you in advance for your help with this.

Thanks, I did not notice that update. I applied it, and it looks better, see screenshot attached. Maybe it would be possible to make all list entries appear in parentheses as in the source text, too? But I can use it the way it is now, for the time being.

And, would it be possible to make the NT list appear at or near the current cursor position, just like the prompter? The way it is now one always has to look too far to the left, away from the target text.

Anyway, thanks for your work, it's greatly appreaciated!

I just discovered something quite useful: In my current project I translated a segment containing ten times (5) and seven times (5a), among other non-translatable fragements that were not duplicates. When I opened the non-translatables list, it contained only one occurrence of each of these fragments, plus one more occurrence of each with the parantheses missing. When inserting the (5)'s and (5a)'s, Cafetran let me insert them the exact number of times they were present in the source text, then they were removed from the list. This is so much better than being presented with a huge list of redundant elements! Now Cafetran could just not show the list elements where the parantheses are missing, as these are not necessary anyway.