Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11] >
TMLookup
Thread poster: FarkasAndras
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
Great! Feb 2, 2016

FarkasAndras wrote:

It looks simple enough to do. All I need to work out is whether it uses character references in addition to < and >. Probably & but there may be others.
Email me a TM, preferably something with ampersands, apostrophes, quotes and a couple of accented letters thrown in and I'll have a look.
If and when I get it done, you can paypal me a donation.

[Edited at 2016-02-02 12:43 GMT]


Wow, thanks! I will prepare and send you an example file ASAP.

Michael


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
OK, here it is! Feb 4, 2016

https://www.dropbox.com/s/7fkhe6r6t6wi19q/Felix-test-TM.ftm?dl=0
https://www.dropbox.com/s/rpb77yv4z5zd6g1/Felix-test-TM.tmx?dl=0

If you need me to add any more weird stuff, just ask!


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
ok Feb 4, 2016

Had a look, it should be plain sailing. No character references apart from &amp;, &lt; and &gt;, and the format appears to be straightforward (and in UTF-8, which is nice).
Oddly enough the tm in your screenshot contains the language codes (<source_language> etc.) but the one you uploaded does not. It doesn't have any tags that refer to any language. Not sure how or why felix does that. Mystery TM exports are not that great as the languages can get mixed up during import.


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
cool! Feb 5, 2016

FarkasAndras wrote:

Had a look, it should be plain sailing. No character references apart from &, < and >, and the format appears to be straightforward (and in UTF-8, which is nice).
Oddly enough the tm in your screenshot contains the language codes ( etc.) but the one you uploaded does not. It doesn't have any tags that refer to any language. Not sure how or why felix does that. Mystery TM exports are not that great as the languages can get mixed up during import.


Re the missing languages, not sure what I did there. Still learning how to use Felix. The TMs I would want to import into TMLookup would all have the correct languages specified. Just checked my latest project TMs, e.g., and they all have the correct:

<source_language>nl</source_language>
<target_language>en</target_language>

in the header.

[Edited at 2016-02-05 00:17 GMT]


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
done Feb 5, 2016

Here it is, report back if something's broken.
https://dl.dropboxusercontent.com/u/16377950/TMLookup.exe


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
Wow, thanks! Feb 6, 2016

FarkasAndras wrote:

Here it is, report back if something's broken.
https://dl.dropboxusercontent.com/u/16377950/TMLookup.exe


Tested it, and it seems to work perfectly! Will send a PayPal donation your way ASAP!


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
@András: Feb 18, 2016

Hey, how hard would it be to create a little tool that would be able to convert the old db into the new db format (without losing the 3rd column for TMX source file)(if/once it gets implemented)? I am getting rather attached to my massive db, and growing increasingly less enthusiastic about the prospect of re-importing all six million of my separate TMXs.

M


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
@András Feb 19, 2016

OK, another feature request.

I am experimenting with using memoQ as a repository for my entire TMX collection, so I can then use this neat trick to quickly produce "reduced-size TMX sub-sets", for lack of a better term, tailored specifically
... See more
OK, another feature request.

I am experimenting with using memoQ as a repository for my entire TMX collection, so I can then use this neat trick to quickly produce "reduced-size TMX sub-sets", for lack of a better term, tailored specifically to my current project in whatever CAT tool I happen to be using today.

Anyway, to cut a long story short: we already have the ability to specify the source (via the super-handy"Add file name as source information" feature) when importing TMXs into TMLookup. However, I would really love it if TMLookup could import a TMX property (I only need one!), or whatever those things are called officially, you know, like:

<prop type="Source">
<prop type="Subject">
<prop type="Client">
<prop type="Document title">

etc.
Collapse


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
Re: Feb 19, 2016

Re: auto-updating a db to a new format, it's not really necessary. You can just export your db into a tabbed txt file straight from TMLookup and import that txt into a new db. Start the export, go out for dinner, and then start the import before you go to bed:)
Obviously you need to free up enough space on the drive before the export. The .txt should be significantly smaller than the .db though.

FTS5 is coming closer to a stable release but it's not there yet. 1 to 3 mo is sti
... See more
Re: auto-updating a db to a new format, it's not really necessary. You can just export your db into a tabbed txt file straight from TMLookup and import that txt into a new db. Start the export, go out for dinner, and then start the import before you go to bed:)
Obviously you need to free up enough space on the drive before the export. The .txt should be significantly smaller than the .db though.

FTS5 is coming closer to a stable release but it's not there yet. 1 to 3 mo is still my prediction, much like cold fusion, which has been a decade and a half away since the 1970s;)


Re: tmx metadata fields, there is some half-assed support for it already. TMLookup looks for <prop type="Txt::*"> tags, so it won't pick up the ones in your example. I'll modify it to omit the Txt::, but it will still be half-assed: it will only pick up the first (?) such field in each TU, and I think it will throw an error for each TU that doesn't have a metadata field (if you try to import metadata, that is). The metadata field is read as the last one in the file. I.e. the third "language" in a bilingual tmx. Honestly, I'm not sure if I ever tested this feature (which is why it's not documented) but it should work.
Collapse


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 08:47
Member (2009)
Dutch to English
+ ...
Wow, that's fantastic! Feb 20, 2016

FarkasAndras wrote:

Re: auto-updating a db to a new format, it's not really necessary. You can just export your db into a tabbed txt file straight from TMLookup and import that txt into a new db. Start the export, go out for dinner, and then start the import before you go to bed:)
Obviously you need to free up enough space on the drive before the export. The .txt should be significantly smaller than the .db though.


Wow, for some reason I thought that the third column (source) didn't get exported to the tabbed text files, but they do, which is great news! Will run a test export.

Re: tmx metadata fields, there is some half-assed support for it already. TMLookup looks for tags, so it won't pick up the ones in your example. I'll modify it to omit the Txt::, but it will still be half-assed: it will only pick up the first (?) such field in each TU, and I think it will throw an error for each TU that doesn't have a metadata field (if you try to import metadata, that is). The metadata field is read as the last one in the file. I.e. the third "language" in a bilingual tmx. Honestly, I'm not sure if I ever tested this feature (which is why it's not documented) but it should work.


This is also good news, although I am still trying to understand the implications, and specifically how they relate to devising a bulletproof workflow between memoQ-exported TMXs and importing into TMLookup.

Preferably a two-way workflow, where I could conceivably:

(1) export my TM from Felix into a TMX, and manually add the correct Source property indicating the name of the TMX
(2) import several of these TMXs into a single memoQ TM for maintenance/etc.
(3) be able to export this whole memoQ TM to a single TMX
(4) import this TMX into TMLookup
(5) maybe even export my entire TMLookup db to a tabbed txt file at some point
... and all of this without losing the initial custom "Source" property and its TMX name

Currently, any custom properties called "Source" that go through memoQ, get exported as:

<prop type="x-Source">this is wkwkwk.tmx</prop>

In order to import the TU including its Source value, I need to change them to:

<prop type="Txt::Source">this is wkwkwk.tmx</prop>

Even if the a TU contains several different custom props, TMLookup happily grabs the first one starting with "Txt::", which is good. Not sure what would happen if you modify it to omit the Txt::. That is, which custom prop would TMLookup then choose?


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
- Feb 20, 2016

Michael Beijer wrote:


Wow, for some reason I thought that the third column (source) didn't get exported to the tabbed text files, but they do, which is great news! Will run a test export.

Internally, there is no difference between text and metadata columns. They are the same to TML. I considered differentiating (not indexing metadata columns to save disk space) but I decided against it.


Even if the a TU contains several different custom props, TMLookup happily grabs the first one starting with "Txt::", which is good. Not sure what would happen if you modify it to omit the Txt::. That is, which custom prop would TMLookup then choose?



It'd probably take the first prop type field. This is all because I don't have a real tmx parser and I just hacked together something that mostly works. It's kind of emberrassing but I don't have the intestinal fortitude to delve into the awfulness that is xml parsing so it will stay this way.


 
Dan Lucas
Dan Lucas  Identity Verified
United Kingdom
Local time: 08:47
Member (2014)
Japanese to English
2 bind variables when 3 are needed Mar 3, 2016

I'm trying out TM Lookup. I keep getting the following error trying to import Japanese/English unicode text exported from Excel. File seems well-formed tab-separated text, but hard to validate that kind of thing.

"Can't execute query: called with 2 bind variables when 3 are needed"

Any ideas where to start looking? Text is proprietary, unfortunately.

Regards
Dan


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
2 bind variables when 3 are needed... Mar 4, 2016

...means that TMLookup was expecting 3 text fields (either three languages or two languages and a meta column) and there were only 2 in the file in that row. Make sure you apply the right settings during the import (i.e. you don't try to read and import 3 columns from a 2-column file). The well-formed lines will still be imported but you need to click away the errors one by one, which gets old if there are 500 of them. I will improve the handling of this error if/when I get around to it.
<
... See more
...means that TMLookup was expecting 3 text fields (either three languages or two languages and a meta column) and there were only 2 in the file in that row. Make sure you apply the right settings during the import (i.e. you don't try to read and import 3 columns from a 2-column file). The well-formed lines will still be imported but you need to click away the errors one by one, which gets old if there are 500 of them. I will improve the handling of this error if/when I get around to it.

If the original is in Excel, Ctrl-down will quickly find any rows with empty cells and help avoid the issue.


[Edited at 2016-03-04 14:50 GMT]
Collapse


 
Danesh
Danesh
Local time: 12:17
English to Persian (Farsi)
What is the latest version of TMLookup? Jun 24, 2016

FarkasAndras wrote:

Fixed the TMX import bug and added a checkbox to filter out half-empty segments during imports:
https://dl.dropboxusercontent.com/u/16377950/TMLookup_1.54_win.zip



Dear András,
Should I use TMLookup 1.54 (you has given link to on this page) or TMLookup 1.31 (at http://www.farkastranslations.com/tmlookup.php)?

Yours,
Danesh

[Edited at 2016-06-24 09:30 GMT]


 
FarkasAndras
FarkasAndras  Identity Verified
Local time: 09:47
English to Hungarian
+ ...
TOPIC STARTER
1.54 Jun 25, 2016

You should use 1.54. It has quite a few new features.
I haven't updated the website in ages. I should get that sorted at last...


 
Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11] >


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

TMLookup







CafeTran Espresso
You've never met a CAT tool this clever!

Translate faster & easier, using a sophisticated CAT tool built by a translator / developer. Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools. Download and start using CafeTran Espresso -- for free

Buy now! »
TM-Town
Manage your TMs and Terms ... and boost your translation business

Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.

More info »