Start a new topic

REQ Regex tagger

Hi Igor, Four months ago you wrote that the regex tagger would be added soon. Any news on this? Thanks, Hans

I've moved this posting from the CafeTran Talk forum to this forum, since it's actually a request.

I wonder in what way it would differ or be preferable over the current non-translatable solution which also supports regular expressions optionally.  

Igor wrote:


>I wonder in what way it would differ or be preferable over the current non-translatable solution which also supports regular expressions optionally.  


Hello Igor,

Thank you for taking interest in this request! I'm happy to explain my reasons to request a regex tagger for CafeTran.

Introduction

This is what the memoQ manual says:

The Regex tagger is a secondary filter that is used after another document filter in a cascading filter configuration. It cannot be used directly to import a document; it can parse the text of a document that was imported by another filter. 

And here's an example from an SDL employee:


Screen%20Shot%202016-11-01%20at%2017.48.39.png



Why would I want to use a regex tagger?


  1. More consistency
  2. Less errors in the variable parts
  3. Less QA needed
  4. More higher matches
  5. More revenues
  6. Nicer work
  7. More free time
  8. Better life
I've prepared two documents: Vars.docx with a couple of sentences with variables and Vars tagged.docx, that show what happens when a regex tagger is used.

Vars.docx

1.png


2.png


3.png

4.png


5.png



Vars tagged.dox

6.png


When I proceed to the next segment, all others are autopropated:

7.png


Of course it would be nice to see the content of the tags in the Project segments pane.

docx
docx
(37 KB)

If you need any more information, please don't hesitate.


With CafeTran being as marvellous as it is, this new feature would be one of the last major improvements that may be necessary.

Hey Igor, do you need any more info on this? Did I convince you?

I think hiding selected phrases behind the tags is a bit exotic request seemingly based on the principle "other tools have it so CT should have it too". However, I find it interesting and probably will be added as an option to non-translatables which could be hidden this way. I can't tell you when as it is not the top priority although I can't wait to start coding it. :)

IK: ...seemingly based on the principle "other tools have it so CT should have it too"


A Swiss Army Knife is a wonderful tool, but if you overdo it...




RZSMkm3dPlPQuQEWtnIwQO8BC0KySWGV2g.jpg

...you'll need a friggin' manual to find the friggin' phillips scewdriver...


H.

A manual is important. I don't question it! Making the tool simple and intuitive is an art and even more important. Does anybody read iPhone manuals?

IK: A manual is important.


Then write one. Nobody else can.


H.

>I think hiding selected phrases behind the tags is a bit exotic request seemingly based on the principle "other tools have it so CT should have it too". 


Not here. I have a genuine interest in using the feature. And of course, it's not out of the blue that others have it.


>However, I find it interesting and probably will be added as an option to non-translatables which could be hidden this way. I can't tell you when as it is not the top priority although I can't wait to start coding it. :)


Thank you very much. Best of luck on this important day!

Hans Lenting: ...it's not out of the blue that others have it.


It's quite possible because they lack Auto-Assembly, and as a result, lack the possibility to adjust the settings for AA.


H.

Quite often I get SDLXLIFF files from a client that has a subsidiary. The manuals of this subsidiary are almost identical with those from the parent. Where the manuals of the client use 'Maier GmbH', those of the subsidiary write: 'Müller AG'.


I hope that CafeTran's implementation of the regex tagger will allow me to convert all instances of 'Müller AG' in an opened SDLXLIFF files to a tag. When I then use my TM for Maier GmbH, all instances of 'Maier GmbH' in matching segments will be tagged too, resulting in exact matches:


Untagged:


TM:

Herzlich Willkommen bei Maier GmbH!


Project:

Herzlich Willkommen bei Müller AG!


Tagged:


TM:

Herzlich Willkommen bei 1!


Project:

Herzlich Willkommen bei 1!

CafeTran´s Assembly function has much simpler "segment patterns" feature. Just define your pattern in your glossary or TM as follows:


German: Herzlich Willkommen bei {1}

Dutch: Dutch translation {1}


and then whatever phrases you have in your glossary that appear in the source segment at {1} position will be translated giving you the exact match.


I'm trying to understand this. So, you are saying:


  • When the TM contains a segment with the name of company A and
  • The project contains the nearly identical segment with the only difference being that instead of the name of company A, the name of company B is written
  • CafeTran will treat these segments as exact matches
  • All these segments will be automatically translated, whiteout te need of further user interference, when using Insert all exact matches?
Or at, leat, I hope that's what you are saying.

Anyway, this would probably require to define loads of segments patterns, for all sentences that contain company names?

Or am I once again totally misinterpreting things here ;).

Login to post a comment