Glossary
Glossary Tab (Detail Panel)
Shows glossary term management and matching for the current project. The tab has two modes, toggled via a segmented control at the top:
- Matched Terms: Shows only glossary entries whose source term appears in the currently selected segment's source text. This updates automatically when you select a different segment.
- All Terms: Shows all glossary entries in the project regardless of the current segment.
Each glossary entry displays:
- Source term and target term
- Mandatory badge for terms that must be used
- Prohibited badge (shown in red) for terms that must not be used
- Domain and Definition if provided
- An Apply button to insert the target term into the translation
Managing Glossary Entries
- Click the + button in the toolbar to add a new glossary entry. A sheet appears where you can enter the source term, target term, domain, definition, and toggle mandatory/prohibited flags.
- Right-click (context menu) on any entry to Edit or Delete it.
- Import entries in bulk via File > Import Glossary (CSV)....
Importing Glossary Entries
You can import glossary entries in bulk from a CSV file via File > Import Glossary (CSV).... The CSV file should have a header row with the following columns:
| Column | Required | Description |
|---|---|---|
| sourceTerm | Yes | The term in the source language |
| targetTerm | Yes | The term in the target language |
| domain | No | Subject domain (e.g., "Legal", "Medical") |
| definition | No | Definition or context for the term |
| mandatory | No | "true" if the term must be used in translations |
| prohibited | No | "true" if the term must not be used in translations |
Imported entries are added to the project's glossary and immediately become available for matching.
How Glossary Terms Are Used
Glossary terms interact with several features in tradumacos:
- Automatic matching: When you select a segment, the Glossary tab automatically shows all glossary entries whose source term appears in the segment's source text.
- QA checks: The Quality Assurance system includes two glossary-related checks:
- Terminology Violation — Flags segments where a mandatory term's source appears in the source text but the expected target translation is missing from the target text.
- Prohibited Term — Flags segments where the target text contains a term marked as prohibited in the glossary.
- LLM context: When using batch pre-translate or single-segment translation, relevant glossary terms are included in the LLM prompt as context, helping the AI model use the correct terminology.