Speaking the Language of Business
GeoEdu

Educational Resource Center

Webinars Whitepapers

Registration Now Open

October 29, 2014

8:00am - 4:00pm

The 4th Annual Global Communications Conference

The Bottom Line
@
Sheraton Hotel - Madison, WI
Welcome reception the evening of October 28th
Registration now open: Register now

Read More

File Tailoring for Localization Projects

If you're a project manager, chances are you’re familiar with greatly varying quotes and proposals that seem to be out of control. The Geo Group’s goal is to help you (the client) understand the localization process better, giving you the skills to prepare your files for localization, allowing you to ‘take back control’ over those ‘out of control’ quotes. Tailoring your files can maximize efficiency when it comes time to translate a product and prepare it for specific regional distribution, saving everyone involved time and money. In order to make your files more localization friendly, it is imperative that they be “search & replace” (find and replace) friendly, monolingual or at most bilingual, and free of surprises within the translatable text (either supply specs or custom coding). However, before we get to the meat and potatoes of file tailoring for localization, we will briefly review the process itself.

To begin, localization is the process of preparing, designing, or programming software or other media for use in a specific region or language through the use of area-specific components and translation. The process of localization is comparable to manufacturing a product in the sense that much like in product manufacturing, raw materials are entered into a process as “input.”  In the case of manufacturing, the input is raw metal, which is then sent through the process of smelting and forge work.  In the case of localization, untranslated text is the input and it is put through a process involving translation and localization.  Both of these processes have end results, or “outputs” and they are similar in that they are both polished, high quality products; the difference is that one output is tangible (tools, utensils, etc) and the other is intangible (translated and localized text).

Specifically, when a localization agency or translator opens a file with a translation tool (software) to begin the process, the following actions are executed by the tool:
 

  1. it searches the file for translatable text (many files contain lots of untranslatable material, such as tags or stand-alone numbers);
  2. it segments the translatable text into translation-friendly chunks (such as sentences), called translation units;
  3. it prompts the translator to enter replacement text (a.k.a. the translation) for each translation unit;
  4. if a translation memory exists (a database of previously translated text) the tool will also search for matches that it can offer to the translator.

In the following sections of this white paper, we go more in depth about what traits are necessary in creating optimal, tailored files for localization.

When tailoring files for a localization agency, make sure the source text (often referred to as untranslated or translatable text) is replaceable. This bears the most importance in the localization process. It means that the source text needs to allow for search & replace or you will find that your files need prep work before the localization process can even begin. There are two main reasons why a localization agency would charge prep time to a client due to improperly prepared files, such as:
 

  1. the files provided for localization hold the source text in one field and require the translation to be placed in another; or
  2. the files contain fields with source text, requiring fields to be added with the appropriate translations.


When a localization agency receives files with these attributes, prep work has to be done and the hours you will be charged for begin to accumulate quickly. This is due to the agency having to add the field(s) to hold the translation, copying the source text into the target field(s), and ensuring the source field(s) remains protected (does not get replaced). This is usually what we mean by “make sure your files allow for search and replace.” It is suggested that you do this before sending files to an agency if you want to be sure that little or no prep time is charged to your localization project.

It is important to remember that any localization agency prefers to receive monolingual files (containing just source text that will be replaced with target text) or at most bilingual files (source text plus a copy of the source text that will be replaced by target text). That usually means one file per target language in the end.

The goal here is to have multiple translations being worked on at one time or in parallel, which is a huge time saver. If an agency receives one file for all languages for translation, it greatly extends the project timeline. We have found that in some cases, the programmers at the client’s end find it easy to work with multilingual files, for example:

stringid1    
[English]=some English text
[Spanish]=the Spanish translation
[Japanese]=the Japanese translation
...
stringid2
[English]=some other English text
[Spanish]=the next Spanish translation
[Japanese]=the next Japanese translation
...

This format appears easy enough to understand. Unfortunately, if we are asked to populate a file like this with 27 languages (or even 2), it becomes a bit more challenging. When we receive files written like this for software, we then have to pass the file on to the corresponding translator one at a time. Our goal is to end up with all translations merged into one file to easily facilitate search and replace into their corresponding target language files. This cannot be done without writing custom scripts to:
 

  1. split the original file into multiple files allowing for easy search and replace, and
  2. merge the individual translations back into the original format.


How much programming is required depends on the actual files. There is a general rule: the more “idiosyncratic” (or distinctive) the file format is, the more prep time required.

Systems that are set up to pull text from bilingual files can easily be switched to pull from parallel monolingual files. This means that instead of writing instructions that say “open the multilingual database file, look for stringid XYZ, and pull the Japanese version,” you can write instructions like “open the Japanese database file and pull the string with id XYZ.”

In the previous section, we discuss what makes the localization process more streamlined. Above, we mentioned “idiosyncratic” file formats. This refers to not only the structure of the file, but also to smaller details such as variables included in the translatable text.

When we receive files without any specifications, we need to spend time looking through them to find all possible forms of variables. Programmers have been known to cook up surprising things….

Variables can appear as:

 1, $1, %1, %%1, #1, {1}, _1_, [1], <1>

In general, the appearance of variables (like the ones above) is ok, but it would be of tremendous help to your localization provider (while saving you time and money in the process) if the files came with notes regarding such details.

In summary, an ideal file tailored for localization has the following traits:
 

  1. allows for search & replace;
  2. is monolingual or at most bilingual;
  3. does not contain surprises within the translatable text - either supply the agency with specifications on coding or avoid using custom coding.


Remembering all of these steps in the process of file tailoring for localization can help shorten timelines, reduce costs, and create better products. You might even be considered a hero.

The Geo Group Corporation is comfortable taking files as they come, in any form. We are happy to analyze files and develop strategies for their successful set-up and localization. We would also be more than happy to provide free consulting to help you achieve the ideal file format. Our business model does not rest on consulting services and we do not aim to earn money with custom coding. If we can help our clients prepare for the smoothest possible workflow, we free ourselves to focus on our core service—providing high-quality, polished translations at the price, timeline and quality level you need. Please follow this link for more information on the what The Geo Group can do to help you achieve your localization needs.

 

Back to Top

Back to Whitepapers

 

 

Back to Whitepapers Archives