Inconsistencies / Number errors Not Caught!!!
Thread poster: Scott Saylor

Scott Saylor  Identity Verified
Local time: 00:35
Japanese to English
Mar 24, 2011



To my horror, I have discovered that Trados 2009 is not actually catching all of my number errors or translation inconsistencies, which I was counting on.

I have made sure there are no exclusions set in the verification tool. Can anyone tell me why I Trados wouldn't catch inconsistencies? Or blatent number errors (like pdf115 in the source and pdf553 in target, etc)?

I have selected / deselected all of the settings I could think of that might help, so now I am out of ideas. If anyone knows I would be very appreciative. Thank you!



Direct link Reply with quote
 

Scott Saylor  Identity Verified
Local time: 00:35
Japanese to English
TOPIC STARTER
Thought I knew but didnt Mar 24, 2011

As for the inconsistencies:

I have 2 translation memories open. In one memory a translation is listed as one thing and then in the other memory its listed as another. Since both memories are open, the verification sees only that the translation is consistent as far as it being the same in at least one of the memories -- regardless of the fact that within the document the segments are inconsistent.

I will test this to see. (later: Nope that's not it!)

As for the numbers, I still need to know why. (and for the inconsistencies because the test showed it didnt matter which memory was open)

[Edited at 2011-03-24 01:56 GMT]


Direct link Reply with quote
 

Jerzy Czopik  Identity Verified
Germany
Local time: 09:35
Member (2003)
Polish to German
+ ...
Your examples are not numbers Mar 24, 2011

Numbers are 1,000 or 56 or 1,000,222 but not PDF115 or PM20. These are alphanumerical variables, which are unfortunately not recognized by Studio nor by any othere Trados version I know.
But Studio keeps track of numbers in usual format, this should work.
Also the inconsitences should be detected and this even without any translation memory. The verification work just on document level AFAIK. If it does not work properly, check your program version and please report here.


Direct link Reply with quote
 

Samuel Murray  Identity Verified
Netherlands
Local time: 09:35
Member (2006)
English to Afrikaans
+ ...
Automatic number alternation versus... Mar 24, 2011

Jerzy Czopik wrote:
Numbers are 1,000 or 56 or 1,000,222 but not PDF115 or PM20. These are alphanumerical variables, which are unfortunately not recognized by Studio nor by any othere Trados version I know.


[I'm a WFC user, so I don't know for certain, but...] Yes, but surely Studio should check for consistency of so-called untranslatable strings, and that would include any "word" that includes a number, such as abc111, or any "word" that is directly against a nonpunctuation/non-tag character... right? Or can Studio not check for such inconsistencies?


Direct link Reply with quote
 

Susanne Rieg  Identity Verified
Local time: 09:35
English to German
+ ...
Using XBench Mar 24, 2011

It is exactly because of these "gaps" in the performance of Studios QA checks that we use Apsic XBench as a second check on the sdlxliff files. XBench is able to catch letter/number combination errors. Also, the tool can check several sdlxliff files together in order to find inconsistencies.

Sanne


Direct link Reply with quote
 

Tomás Cano Binder, BA, CT  Identity Verified
Spain
Local time: 09:35
Member (2005)
English to Spanish
+ ...
SDL can surely improve that! Mar 24, 2011

Amazing.

My main product checks that and is very efficient (well, sometimes even a bit picky, but you can disable the feature if you wish), so it should not be too difficult to implement in future versions of Studio.

Having to run Xbench to cover such a deficiency sounds plain ridiculous. We don't have the time to configure and run Xbench 10 times a day!


Direct link Reply with quote
 

Stanislav Pokorny  Identity Verified
Czech Republic
Local time: 09:35
English to Czech
+ ...
Confirmed Mar 24, 2011

Tomás Cano Binder, CT wrote:
My main product checks that and is very efficient (well, sometimes even a bit picky, but you can disable the feature if you wish), so it should not be too difficult to implement in future versions of Studio.


I can confirm MemoQ performs much better at this.


Direct link Reply with quote
 

Adam Bojan  Identity Verified
Poland
Local time: 09:35
Dutch to Polish
+ ...
Use the regex of QA module Mar 24, 2011

Use the regular expression module of QA in project settings.
Go to "project settings">verification>regular exp... and tick the box "search reg..."
Then define a regex that will search for missing numbers:
in your case this will do:
in description: write any name
in regex source: [0-9]{1,8}
in regex target: $0
in condition: choose: "grouped searched expression, report if source matches but not target"
action: add item
OK and it should now work.
This topic was discussed here:
http://www.proz.com/forum/sdl_trados_support/163403-help_with_regex_needed_to_check_numbers_sdl_studio.html

more help on regex eg. here: http://blog.sdl.com/blog/2011/01/just-use-this-regex-expression.html
for me the problem with numbers is:
I have to specify the regex QA definitions for each project (template). It is not possible to save/export/import them (sic!).
When I check the box "numbers" for checking numbers by QA I get mostly (sometimes 100%) only false positives (eg with dates that are "not properly localised" but by who?).
There is always a case of number+letter string that will not be picked by the defined regex (see the discussion for how difficult it is to prepare an adequate regex).


Direct link Reply with quote
 

PMz
You can export profile Mar 25, 2011

Adam Bojan wrote:


I have to specify the regex QA definitions for each project (template). It is not possible to save/export/import them (sic!).


Actually you can export whole QA Checker profile - in the Profiles tab.

Regards


Direct link Reply with quote
 

Scott Saylor  Identity Verified
Local time: 00:35
Japanese to English
TOPIC STARTER
Thanks! Mar 25, 2011

Hello All,

Thank you for your many helpful responses. I am going to look into them. Some of the stuff may take a while for me to digest, but I am looking forward to implementing your good ideas. Thank you!

Scott


Direct link Reply with quote
 

Jing Nie
China
Local time: 15:35
Member (2011)
English to Chinese
+ ...
A question about Regular Expression Apr 3, 2011

I have a material just like this.

English Sentence 1
Chinese translation 1
English Sentence 2
Chinese translation 2
English Sentence 3
Chinese translation 3
......
......
English Sentence 100000
Chinese translation 100000

I want to use Notepad++ to change it to like this:
English Sentence 1< Tab >Chinese translation 1
English Sentence 2< Tab >Chinese translation 2
English Sentence 3< Tab >Chinese translation 3
......
English Sentence 100000< Tab >Chinese translation 100000

So I can convert it into TMX.

But I do not know how to write Regular expression in Notepad++, search and replace and get the result I need.
I know i can handle it in MS WORD, but it is too large, I want to know how to hand it in Notepad++

[修改时间: 2011-04-03 03:25 GMT]


Direct link Reply with quote
 

Jaroslaw Michalak  Identity Verified
Poland
Local time: 09:35
Member (2004)
English to Polish
Short answer Apr 3, 2011

Jing, I think such questions are better asked in a new topic...

A short answer: Notepad++ does not handle \r\n in the regexp search, so you have to do it in two stages:
1. Replace \r\n with a placeholder (e.g. @) with Extended search.
2. Use regex find: ([^@]+)@([^@]+)@ replace: \1\t\2\r\n


Direct link Reply with quote
 

Jing Nie
China
Local time: 15:35
Member (2011)
English to Chinese
+ ...
Many Thanks for your help. Apr 3, 2011

Hi Jabberwock,

Many Thanks. I will try.


Jabberwock wrote:

Jing, I think such questions are better asked in a new topic...

A short answer: Notepad++ does not handle \r\n in the regexp search, so you have to do it in two stages:
1. Replace \r\n with a placeholder (e.g. @) with Extended search.
2. Use regex find: ([^@]+)@([^@]+)@ replace: \1\t\2\r\n


Direct link Reply with quote
 

Jing Nie
China
Local time: 15:35
Member (2011)
English to Chinese
+ ...
it really works. Apr 3, 2011

I downloaded a freeware named MadEdit.

Search : ([^\r\n]+)\r\n([^\r\n]+)\r\n
Replaced with: $1\t$2\r\n

Very quick and neatly, thanks a lot for your help.

Jabberwock wrote:

Jing, I think such questions are better asked in a new topic...

A short answer: Notepad++ does not handle \r\n in the regexp search, so you have to do it in two stages:
1. Replace \r\n with a placeholder (e.g. @) with Extended search.
2. Use regex find: ([^@]+)@([^@]+)@ replace: \1\t\2\r\n


Direct link Reply with quote
 


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


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

Inconsistencies / Number errors Not Caught!!!

Advanced search







Protemos translation business management system
Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!

The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.

More info »
PerfectIt consistency checker
Faster Checking, Greater Accuracy

PerfectIt helps deliver error-free documents. It improves consistency, ensures quality and helps to enforce style guides. It’s a powerful tool for pro users, and comes with the assurance of a 30-day money back guarantee.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search