The translation memory (TM) is a database of previously translated text used in the projects and suggests matching results while translating content, based on the default language of a project. This speeds up the translation process and improves consistency across texts and projects.
The TM is an organization-wide translation tool which can be assigned to different projects. Otherwise, if none of the projects share a translation memory, each project will rely on its own local translation memory.
Updating the translation memory is not necessary, as it is refreshed automatically with every new translation.
-
Click on Translation memory in the left-side menu to view the corresponding page.
-
Under the section, select the projects that will share the same organization-wide translation memory. To assign all existing projects at once, click on the checkbox at the top of the list.
-
Click on Update settings at the bottom of the page to apply the changes.
Translation memory sharing can also be enabled for each project individually in the project settings:
-
In the page, hover the mouse on the desired project and click on More/Project settings from the drop-down menu at the top.
The window will be displayed.
-
Select in the left-side menu of the window to view the corresponding settings.
-
Scroll down to find the option and check it to enable translation memory sharing for the project.
-
Click on Save at the bottom right of the window to apply the changes.
In the translation editor, suggestions coming from the translation memory are listed under the menu in the right sidebar.
Click on a listed suggestions to use it: the target language box is automatically filled with the text coming from the translation memory. Hover the mouse on each suggestion to open a side card preview with more details (e.g. status, related project, metadata).
The list of available suggestions is displayed by pressing the ArrowDown key when the target language field is selected.
The translation memory only suggests translations matching all the current language codes (locales) set in the project and it also finds suggestions when the source strings do not exactly match.
In case of multiple suggestions matching the source text, the best suggestion is listed first with other suggestions sorted highest to lowest percentage according to the quality of the suggestion. A colored progress bar next to each suggestion indicates how similar the suggested string is to the source language to be translated (100% match, fuzzy match, etc.).
Classic editor
In the translation editor, suggestions coming from the translation memory will be listed under the tab in the right sidebar.
To use one of the listed suggestions, click on the Apply button next to it in the tab: the target language box is automatically filled with the text coming from the translation memory.
The translation memory only suggests translations matching all the current language codes (locales) set in the project and it also finds suggestions when the source strings do not exactly match.
In case of multiple suggestions matching the source text, the best suggestion is listed first with other suggestions sorted highest to lowest percentage according to the quality of the suggestion. A colored progress bar next to each suggestion indicates how similar the suggested string is to the source language to be translated (100% match, fuzzy match, etc.).