You can use the Autofill tool to fill up empty text segments with matching texts from your Translation Memory or Machine Translation.
In this article, we will cover the following:
- how to access Autofill settings
- using Autofill with Translation Memory
- using Autofil with Machine Translation
- using Autofill with Translation Memory AND Machine Translation
How to access Autofill settings
Access Autofill settings via the “Translation Tools” submenu in your project settings.
Please note that, since Autofill settings are part of your project settings they apply for the selected project.
Autofill empty segments from Translation Memory
If ticking the checkbox LingoHub will check empty text segments of this project and automatically copy the text from the Translation Memory. The Translation Memory assigned to your project will used for autofilling.
Autofill settings allow you to decide if 100%, >90% or >80% text matches from the translation Memory shall be used to fill empty segments.
You can also select the status that will be assigned to autofilled segments. Selecting “MATCH STATUS” will assign the same status to the autofilled segment as used in the corresponding segment from the TM.
Only empty text segments can be autofilled. The status of the segment doesn’t matter, however it needs to be empty to be autofilled if the rules you specified apply.
Autofill doesn’t automatically apply for text segments created in the past. This means that empty text segments are autofilled if autofill is triggered.
Autofill is always triggered, if
1. you edit a text segment in project A that is used to fill your Translation Memory and the matching text segment in project B is empty. This only applies if the text changes are saved or
2. you add a new project locale (=language) to your project. All empty text segments will be autofilled from the Translation Memory if the specified rules apply (e.g. 100% match) or
3. a new empty text segment(s) is created. New text segment(s) can be created by uploading a file or creating a new text segment in the online editor.
A project can autofill itself. If your project is used to fill a certain Translation Memory and also receives suggestions from this Translation Memory, matching texts in your project can be autofilled.
Here’s a short example: You translate your website from English to German and use the phrase “Sign up” multiple times on your website. Your LingoHub project is used to fill the TM and also receives suggestions from it. “Autofill empty text segments from Translation Memory with text segments matching 100%” is activated. If your translator now translates “Sign up” the first time with “Anmelden” all other empty target language segments with the source phrase “Sign up” are autofilled with “Anmelden”.
Autofill empty segments from Machine Translation
If you tick the checkbox all empty segments will be autofilled with Machine Translation after activation. Autofill doesn’t automatically apply for text segments created in the past. This means that empty text segments are autofilled if autofill is triggered.
Autofill is always triggered, if
1. you add a new project locale (=language) to your project. All empty text segments will be autofilled from Machine Translation.or
2. a new empty text segment(s) is created. New text segment(s) can be created by uploading a file or creating a new text segment in the online editor.or
3. you edit the source text and no target text has been specified. Changes in your source text need to be saved that autofill is triggered.
Autofill from Translation Memory AND Machine Translation is selected
You can select both options to autofill empty segments from Translation Memory and Machine Translation. In this case LingoHub first checks empty segments and automatically copies the matching text from the TM if existent. This means LingoHub always tries to autofill empty text segments with TM first. If no matching texts exist in the Translation Memory the empty segments will be filled with texts from Machine Translation.
For additional autofill rules and how to share texts across different languages or dialects, check out this article.