Phrasebook

← Guide index · 中文 · ninth item in the navigation bar

The phrasebook collects common expressions: greetings, numbers, travel-guide style dialogues, fixed patterns from lessons. The division of labour with the Corpus page: sentences are for gloss analysis, phrases are for organising and presenting. A Lexicanter Phrasebook ends up here after import.

1. Layout

The left column lists categories (“All” plus each category with its count), the middle shows phrase cards (script line, text, pronunciation, translations, variants), and the inspector edits the current phrase. The search at the top covers text and translations, and only the current language is shown. Press a card and drag it onto another to move it there.

Words in the text that can be found in the lexicon (including inflected forms and stripped affixes) are underlined: hover to show the entry card, click to jump to the lexicon. Words are recognised exactly as in the corpus — inflected forms, affixes, paradigm affixes worked backwards, and morphemes too.

Fix on the card is the same code as in the corpus: a phrase stores its own analysis, so whichever piece of the segmentation is selected is the piece Fix changes (two occurrences of the same entry in one word are told apart), and picking a sense or a segmentation works here too. Search for an entry or a morpheme and click it and only what the word points at changes — the text itself is left untouched; to rewrite the text as well, tick Also rewrite the word in the text (off by default, and only offered when the whole word is being changed), after which the pronunciation and script line are recomputed from the rules (anything you marked irregular is left alone). Either way the whole thing undoes in one step. A piece that isn’t recognised says “Not found” on hover, and the same search box fills it in. Editing the text re-runs the analysis, keeping the words you have already confirmed.

2. Phrase fields

Deletion can be undone.

Translation workbench: click Workbench next to the translation in the inspector to build the text backwards from the translation, the same way as the corpus translation workbench: write the translation, drag the floating words up into order, click a word to pick its form and add affixes and particles in the attachment deck, then click Add to phrasebook. If the selected phrase’s text is still empty it is filled in; otherwise a new phrase is added to the current category. The words you picked are stored as the phrase’s analysis, and the pronunciation is filled in by the rules.

3. Connections to other modules

• Phonology: pronunciation rules.
• Script: a script line at the top of each card.
• Theme: text and translations use the corpus font slots.
• Command palette: Ctrl+K finds a phrase by text or translation and jumps to it.

4. Tips

• Import and Export in the title bar: import from a table (CSV / TSV / one per line); text, translations per language, category, tags and variants are guessed from the headers, and phrases whose text already exists are skipped. You choose the columns in the import panel in the main area while the inspector shows a live import preview (updating as you change settings, with changes highlighted and fading). JSON can be imported (also with a preview first) and exported, and CSV can be exported (following the current category and search).
• Categories can use “/” for hierarchical names (like “Travel/Directions”); the left column sorts them alphabetically.

5. Table format

Exported CSV files have exactly these columns and can be imported back directly; phrases whose text already exists are skipped. Format help at the top right of the import panel opens this section.