Consistency check on 'close to' 100% matches

Question

Hello forum members. 
 I am working on a big manual with lots of repetition across 33 files. Very often, repeated sentences are not 100% matches, but 99% matches because they contain a different number of tags or because some invisible element, such as in one segment there are non-breaking spaces, or one segment had an extra space or a typo, or a different spelling. These fuzzy are both internal and with the pre-existing TM. 
 Of course, in my translation they should all read the same. How do I run a consistency check that includes, say 98% to 100% matches? 
 
 For the sake of clarity, my translation must read the same whether the source read: 
 
 We are a company specializing in cars, and trucks. [AmE, Oxford comma] 
 We are a company specializing in cars and trucks. [AmE, no Oxford comma] 
 We are a company specialising in cars, and trucks. [BrE, Oxford comma] 
 We are a company specialising in cars and trucks. [BrE, no Oxford comma] 
 We are a company specializing in cars and trucks. [AmE, no Oxford comma and double spacing] 
 We are a company specializing in cars, and trucks [AmE, Oxford comma, no full stop]

Alison Field · Accepted Answer

Hi Paola 
 I know of no way you can run a 'consistency check' between segments that are not 100% source matches. 
 However, I can tell you what I always do in this type of instance. 
 With all files open simultaneously in a 'virtual merge' (open as if they are one document)... 
 Each time I realise an almost-identical phrase is going to occur, I use the filter on the Review tab (or the Advanced filter) and paste into it only the source text that is identical in each case, so, with the above sentence it would be: 
 We are a company specialising in cars 
 Then, I would type how I wanted the translation to be in all cases into my target segment, copy the whole target sentence, and confirm it (Ctrl+enter), which auto-propagates all other instances so long as you have Auto-propagation enabled on under File > Options > Editor > Auto-propagation (I also have it set to auto-propagate 'from first segment'). 
 The software would then automatically take me to the next empty target segment, where I would paste the copied target translation and confirm it. 
 Repeat until all instances are covered. 
 Using this method, there is no need to run any check because I know I've used exactly the same translation each time. 
 This method can save a lot of time once you get into the habit, if you have documents that has many semi-identical source segments. 
 I know this is a workaround but it means you don't have to wonder whether each new version of a semi-identical source segment has really been translated the same as the previous when using Ctrl+T with a less than 100% match coming from the TM, especially if the TM is large. 
 I hope this helps... 
 All the best 
 Alison

Alison Field · Answer

Diana Palenzuela Paul 
 In case anybody needs to know this... I'm surprised I did not notice all these years ago that 'specialising' and 'specializing' were spelt differently in the example that Paola shared 
 So, the above suggestion works better if you use the following: 
 We are a company speciali.ing in cars 
 either with the Review tab Display filter (which already 'sees' regex characters), or with the Advanced Display Filter 2.0 with the Regular Expression box ticked... 
 The regex character . matches any single character 
 I hope this helps future readers 
 All the best, 
 Ali

Trados Studio > 1. Trados Studio

Consistency check on 'close to' 100% matches

Top Replies