Help Desk for Team Members
 
 

This section offers some practical, step-by-step instructions or advice to our colleagues, who may grapple with the issues that we have come across and solved successfully.

In the course of our work, we have received help from numerous fellow experts and professionals. Now it is our turn: here we are presenting simple and advanced tips and tricks, recommending tools and procedures, as well as warning of pitfalls to avoid.

  1. How can I log into your SharePoint extranet?
  2. How can I show all the text (including the hidden parts) in MS Word documents?
  3. How do I track changes in MS Word documents?
  4. How do I review or insert my queries, comments and revisions in MS Word documents?
  5. I have located a Web dictionary and I want to download it and use it off-line.
  6. When converting files I get strange codes. What to do?
  7. How can I convert a Microsoft glossary to the Trados MultiTerm format?
  8. How to count ALL the words in a document, including the text boxes?
  9. How to segment internal matches with Trados or other CAT tools?

 

 

2. Showing all the text in MS Word documents


When answering queries or reviewing documents (especially the ones which have been translated in Trados), please display the hidden text (usually with the source language):

To do so, in MS Word select Tools (Narzędzia) and Options (Opcje). You will see a dialog window similar to this one:

Click the boxes Hidden text (Tekst ukryty) or All (Wszystko).

 

3. Tracking changes in MS Word documents

Before reviewing a document, you need to make sure that all the changes introduced by you will be easy to find by the author of the document or another reviewer.

Microsoft Word ver. 9.0 (also known as "Word 2000")

In the menu bar of Microsoft Word 9.0 select Tools (Narzędzia) | Track changes (Rejestruj zmiany) | Highlight changes (Wyróżniaj zmiany). You will see a following dialog box:

then select the fields: Track changes while editing (Rejestracja zmian podczas edycji) and Display changes on screen (Wyróżnianie zmian na ekranie) and click OK.

Microsoft Word ver. 10 and 11 (also known as "Word XP", "Word 2003")

Select: Tools (Narzędzia)| Track changes (¦ledĽ zmiany) from the Microsoft Word menu bar. This will also add the Reviewing toolbar to your toolbars area above the ruler bar.

You can now modify the text, while keeping track of the changes:

Notes:

  1. DO NOT DELETE ANY OF THESE HIDDEN CONTROL CODES.
  2. It is enough to mark your comment once - we will implement it later in other similar cases.
  3. A tip: when editing the beginning of the target language sentence (e.g. W ciągu in the picture above), make sure you type in the correct character format: move the insertion point to the right by one character (after the letter W), press Backspace and start typing or deleting from there.

4. Reviewing, answering and inserting queries in MS Word documents

You may need to display the hidden text first. Then display the list of comments.

Microsoft Word ver. 9.0 (also known as "Word 2000")

To review and edit annotations (also known as "comments"), double-click the colored mark showing such a note.

The window will split, enabling you to access the comment. All the annotations should be reviewed in the revision pane, by adding your reaction beneath the original comment or query.

Microsoft Word ver. 10 and 11

To show the revision pane in MS Word 11, on the Reviewing (Recenzja) toolbar click the Show (Pokaż) drop-down menu. Select the Revision Pane (Okienko recenzowania) item.

Click a comment in the revision pane to edit and locate the referred text in the upper window pane.

You may wish to insert your own comments, for example when you are only suggesting a change.

Select Insert (Wstaw)| Comment (Komentarz) from the Microsoft Word menu bar, or click Comment (Komentarz) button on Reviewing (Recenzja) toolbar.


The window will split, as shown above, allowing you to type your comments in the bottom pane.

 

 

 

 

 

5. Converting online dictionaries.

Please check this section later. For the time being, please download our 2002 ATA presentation which discusses the methods.

 

6. Macros for codes conversion

Please check this section later. Some macros are available in our 2002 ATA presentation package. 

 

7. Converting Microsoft glossaries to MultiTerm

 

Tools:
Microsoft Word, Microsoft Access, Microsoft Windows

Preparation time:
One day

Ingredients:
Microsoft glossaries

Problem:

It is useful to convert the Microsoft glossaries to Trados Workbench or other CAT format.

 

Solution:

Download Microsoft glossaries

Download the zipped files appropriate for your language combination from ftp://ftp.microsoft.com/developr/msdn/newup/glossary and unpack them in a temp directory.

You will get a collection of CSV files.

Delete tabs

Open each CSV file in Microsoft Word. Change all the tab marks:
^t
to any unused symbol, for example to:
>-->
Save each file afterwards.

 

Import into Access

Import each CSV file into Microsoft Access.

Create an empty Access file and use the File | Get external data | Import command. (Do not use the Open command - it would link to the CSV files only. We want to import them instead.)

When importing, you need to define an appropriate import filter. In the Import File Wizard dialog window click the Advanced button and define an import specification which imports Columns 1, 3, 5, 8 only and uses the comma as the delimiter and quote marks as the text qualifier:

Save this filter by pressing the Save As... button. Reuse it with the other files by clicking the Specifications... button at each import.

Check source files

You should see a list of tables with five columns, the first column containing the indexing numbers ("primary key") only:

Notes:

1. Some of the files (e.g. po_iex501) do not import correctly, on account of their wrong internal structure. Ignore them.

2. Some of the longer strings will be lost. It is not a bug, it is a feature of Microsoft Access. (There is a way around it, but the procedure is too long to explain here.)

 

 

Export to txt

Click the table with the imported file to display the table with the imported strings.

Select File | Export, to export each table as text delimited by tabs without the quote marks. In the MS Access export definition window (similar to the import window, see above), order the columns (fields) in this order: Field 5, Field 8, Field 1, Field 3, that is: Context (e.g. "button"), Windows program name (e.g. "WindowsXP"), Source (e.g. "I am a button"), Target ("Jestem przyciskiem"). (You can order the fields in the export window by... overtyping the names of the exported columns.)

Join glossaries

Join all the resulting txt files into one. Go to DOS, and type e.g.:
copy *.txt all.doc

Convert to Trados WorkBench

(You may also add and fill down the data in the columns in Access in this way.) Open the TXT file in Word. Write and run a macro to convert the file to the Trados Workbench import format. Use two additional context fields:

               <TrU>
               		<Txt L=Context>TXT
               		<Txt L=Source>Microsoft: Visio 2003
               		<Seg L=EN_US>Negative impedance both-way amp
               		<Seg L=PL_01>Dwukier. wzm. o ujem. imped.
               </TrU>
               <TrU>
               		<Txt L=Context>TXT
               		<Txt L=Source>Microsoft: Visio 2003
               		<Seg L=EN_US>Negative impedance both-way amplifier.
               		<Seg L=PL_01>Dwukierunkowy wzmacniacz o ujemnej impedancji.
               </TrU>
               <TrU>
               		<Txt L=Context>TXT
               		<Txt L=Source>Microsoft: Visio 2003
               		<Seg L=EN_US>Amplifier with bypass
               		<Seg L=PL_01>Wzmacniacz z obej¶ciem
               </TrU>
Import to Workbench

Test the import process on a small chunk of the txt file. (Consult the Trados manual if in doubt.) You should get a nice database with four fields:

Notes:
  1. Unfortunately, only the first 255 characters of each field are imported.
  2. Instead of using Word, you can use Microsoft Access macros or commands to perform some of these operations.
  3. Do not try to do it in Excel - it is not capable of importing long text databases.
  4. You may try using the MsGloss2TWB program instead. Be careful: version 1.0.3 of this program crashes on some malformed glossary files, strips all additional info from the records: the source and the context of a string, and even does not handle Unicode properly (the resulting files do not have proper coding, at least in the case of CE languages).

 

8. Counting all the words in a document, including the text boxes

Tools:
Microsoft Word 9.0 or newer

Preparation time:
Half an hour

Ingredients:
Word document
Special macro

Problem:
The usual Word Statistics tools in Microsoft Word does not include the text in the text boxes. (Actually, the Word Help file warns you about this "feature", advising you to manually select the text boxes and count the words again. Not too practical, isn't it?).

Solution:
Instead, you may try this macro.

Short instructions:
  1. Check your Word Startup directory. In Word select Tools | Options | File Locations | Startup.
  2. Download the macro package
  3. Unzip it to your Startup directory, established in Point 1
  4. Restart Word
  5. Ignore the warnings about the new macros or disable them

You will now have access to the CountAllWordsinActiveDocument.Main macro, which you may easily add to your Tools menu or the Word toolbar by selecting the "Customize" command. (See the Word Help file for further instructions.)

 

 

9. Internal matches - Trados and other CAT programs

Tools:
Trados (or other CAT program not capable of exporting internal matches)

Preparation time:
Varies

Ingredients:
Source documents
Piece of paper
Shot of vodka
Primary school rudiments of the set theory

Problem:
Many CAT tools cannot mark similar segments (sentences) in the analyzed documents. For example, if five Help documents have similar sentences each:

"Company policies and procedures vary by business unit, so please see your local Loss Prevention representative or manager for details about your specific area. "
"Company policies and procedures do not vary, so please do not contact your local Loss Prevention representative or manager for details about your specific area."
"Policies and procedures of Company X may vary, so please see your local Loss Prevention representative or manager about the details for your specific area"
etc.

It would be the best for the same translator to handle all such sentences, so as to retain not only procedural but also stylistic consistency. On the other hand, with the looming deadline and the large size of the project, you may need to create a team of several translators working on the same set of source files. (The problem is somehow alleviated when they share the same translation memory, yet discrepancies will arise when one of them decides to update their translation of this segment afterwards.)

Solution:
Here is a short instruction on how to tag such repeating or similar segments in the source text, using Trados or other similar CAT tools:

  1. Analyze all the files
  2. Export all the repeating (frequent) segments to a memory file Frequent_Memory. (In Trados use Export Unknown Segments | 2 or more Occurrences).
  3. Create new clean memory Frequent_TM.
  4. Import the Frequent_Memory file created in point 2 to the new Frequent_TM memory.
  5. Check if the hidden text exists in the original file.
  6. Select the appropriate color for the (pre)translated text, e.g. green (in Trados: Options | Translation Memory Colors)
  7. Pretranslate the work files with memory Frequent_TM, at e.g. 80% match rate.  (In Trados use the option "80" % or higher match value). Make sure that you disable the Mark unknown sentences option. You will get a file where the segments that repeat or are similar to one another are marked with the source/target ("Trados") tags.
  8. If there was no hidden text beforehand, delete all the hidden text. You will get the original file, but with the repeating sentences marked with the (green) color selected in Trados (see point 6.)
  9. To make things easier for the translator, replace green text with the DO_NOT_TRANSLATE style.
  10. At the end, replace the DO_NOT_TRANSLATE style with normal character style, but with e.g. violet font, and translate only such repeating segments.