turn2unit import schema

[johnwdubois]

What to do Create a schema that makes it easy to create Units based on turns (when importing a OWPL file).

1. The schema is called owpl_turn2unit (stored in the Import folder).
2. The schema uses a Turn-based field as the Unit delimiter.
   • turnId (changes)
   • turnOrder (=1)
   • turnSeq (changes)
   • etc
3. If no Turn-based field is available, use the Participant field as the Unit delimiter. That is, create a new turn when there is a change in:
   • Speaker label [or Participant label] (changes)
   • ParticipantID (changes)
4. For each unit created under this import schema, set UnitType = turn.
5. (For the default import of SBC-type files, set UnitType = intonationUnit.)
6. Naming convention for One Word Per Line import schemas:
   • owpl_sbc.json
   • owpl_turn2unit.json

Further development

1. Use a similar import strategy (and data structure) for other types of units (as represented in OWPL files). Give each import strategy its own schema:
   • owpl_sentence2unit: set UnitType = sentence
   • owpl_pSent2unit: set UnitType = pSent
   • owpl_clause2unit: set UnitType = clause
   • owpl_para2unit: set UnitType = paragraph
   • owpl_tcu2unit: set UnitType = turnConstructionalUnit
2. Extend a similar import strategy to:
   • transcription import: test for where the speaker (participant) label changes.
   • prose import (para2unit): test for newline (hard return) character (at the end of each paragraph)

[terrydubois]

Import Screen: "Import Fields" window needs click range clamp.

[terrydubois]

Use "unit label" instead of "sentence" in Stacker options.