Understanding Pseudo-Localization and Its Role for Localization Testing

Table of Contents

Pseudo-localization is a key technique in localization testing, crucial for identifying and resolving common i18n issues including embedded strings, concatenation and other. It allows you to actively test an application’s internationalization readiness before actual localization to ensure it is fit for global markets.

Exploring the Pseudo-Localization Process

Known by several names—pseudo-translation, test translation, round-trip test translation, translation simulation, or dummy translation—pseudo-localization simulates translation. It automatically replaces text with placeholder characters, keeping non-translatable elements intact and mimicking different issues such as varying text length, characters, and language direction. 

A pseudo-locale is like a regular locale, like de-DE for German translation, but instead of a translation, it modified the source strings in the following way:

A source string like:

Order placed successfully!

will be localized as:

[Öŕðéŕ þļåçéð šûççéššƒûļļý!————- П國カ내]

  • A start character, , is added at the beginning of the string. 
  • The string expanded to reflect the possible width of target locales strings.
  • Some characters from other writing systems are added to check for encoding or font issues. 
  • An end character, ], is added to show there that the source string ends to help detect concatenations.

Key Issues Identified by Pseudo-Localization

Pseudo-localization helps identify various issues in software localization before actual translation begins. Some of the key issues it helps to flag include:

  • Embedded strings that can’t be sent to translation
  • UI layout issues caused by text expansion
  • String concatenation which can make correct translation difficult or impossible due to the different order of parts or sentence structure in different languages.
  • Badly formed message syntax, such as a sentence split into multiple parts displaying as multiple bracketed messages.
  • Character set and encoding issues, ensuring that the product supports more than one character set or encoding standard.

How Pseudo-Localization Works: A Step-by-Step Guide

Let’s explore how you can perform pseudo-localization for localization testing. For this demonstration, we’ll use our demo website, Rebel Outfitters, along with Localyzer

First, we’ll examine the user interface in its source locale, which, in this instance, is English/US.

Interface of Rebel Outfitters with a navigation bar, showing a selection of starship models for sale.

Next, we run the application using a pseudo-locale (‘esperanto’), where you’ll notice the English strings have undergone pseudo-localization.

At this stage, you can see several internationalization issues, which are highlighted with red numbers.

Interface of Rebel Outfitters with a navigation bar, showing a selection of starship models for sale. The text is pseudo-localized with a mixture of English and placeholder characters to simulate localization issues, with red numbered indicators highlighting potential issues in the UI.

Let’s review each one:

  1. Text expansion: The end characters have been truncated, indicating a likely UI issue around the space set for the text. The UI may need to be refactored to accommodate for longer text in languages such as German.
  2. Character encoding: When the pseudo-localized text is showing as mojibake, such as �, this likely indicates a character encoding or a font issue. The application does not support non ASCII characters or non Latin-1 characters.
  3. Embedded String / Hardcoded String: If the text shows in the original source locale, (here English) as opposed to being pseudo-localized, it indicates a likely hardcoded string which has not been externalized into a resource file. That string cannot be sent to translation and will show in the interface as the original source string. This is a common internationalization issue.
  4. Concatenation: The pseudo-localization shows that two strings have been put together to make up Y-Wing Galactic Fighter, since the end character , ] , shows up after Galactic Fighter and a start character , [ , shows up before Fighter.

When these issues are fixed, the application is running once again and we see that pseudo-locale shows no issues anymore.

Interface of Rebel Outfitters with a navigation bar, showing a selection of starship models for sale. The text is pseudo-localized with a mixture of English and placeholder characters to simulate localization issues.

This example was created using Localyzer, a Lingoport tool that automates the process of UI string translations directly from repositories and streamlines localization testing. 

Here’s a detailed look at the pseudo-localization process with Localyzer:

  • String Retrieval: Localyzer automatically and continuously scans your software repositories to fetch new strings.
  • Pseudo-Localization Testing: Upon retrieving strings, it applies pseudo-localization rules to test them.
  • Automatic Recognition: When a string is added to a resource file, Localyzer immediately identifies it and processes it for pseudo-localization.

Check out this 7 mins, video by Olivier Libouban, showing in details how to run pseudo localzition.

Pseudo-Localization Challenges in Localization Testing 


As demonstrated, pseudo localization is a straightforward and effective method to identify errors relatively early in the development cycle. Nevertheless, it does not provide complete coverage, and several challenges still remain.

Not connected to the code:  Although developers can see an error, the error doesn’t connect to that part of the code.

Limited to text:  Pseudo-localization won’t catch all i18n issues. It primarily focuses on text, and may not adequately address issues related to other aspects of localization, such as date and time formats, currency, or cultural differences.

False positives: Applying expansion to areas of the UI where text length has already been limited by other systems can cause false positives, leading to unnecessary adjustments.

Not a replacement for real testing: Pseudo-localization is a tool to help identify potential issues, but it is not a substitute for actual testing with real translations and localized content.

To address these challenges, a tool like Globalyzer could be beneficial. Globalyzer offers the following advantages:

  1. It shows where in the code each issue is located, providing a clear path to resolution.
I18n issue Globalyzer report card. Showing issue status, priority, type, and code line when the issue can be found.

2. Globalyzer catches a wide range of i18n issues, not just those limited to text. 

Summary dashboard of a Globalyzer Report Card showing remaining i18n issues: 64 Embedded Strings, 16 Locale-Sensitive Methods, 0 General Patterns, 2 Static File References with a total of 82

3. It marks false positives and offers AI-generated suggestions for fixing or ignoring them, streamlining the debugging process.

Conclusion

To summarize, pseudo-localization offers a proactive approach to identifying internationalization (i18n) issues.

Integrating tools like Localyzer can automate the pseudo-localization process, providing immediate feedback and notifications on pseudo-localization results.

While pseudo-localization is effective, it is not without its limitations, particularly in areas beyond text translation.The incorporation of tools like Globalyzer can help to overcome these limitations.

With Localyzer and Globalyzer you can move one step ahead in your approach to i18n testing, making the process more efficient, accurate, and in sync with global product release strategies.

If you are looking for an extensive solution, Lingoport offers a full-scope i18n testing service.

To explore more about these products and services, contact us here.

Author

Kate Vostokova
Kate Vostokova
Kate is a seasoned B2B content marketing manager with a five-year journey in the localization industry. She is passionate about crafting various types of content to educate people about internationalization (i18n), localization, and the latest technological advances, including Large Language Models (LLMs).
Hand with mobile phone

Talk to Our Experts

Get you tailored consultation on i18n & L10n products and services.

Related Posts

Receive Latest News and Webinar Dates

Get the latest internationalization news delivered to your inbox.