export_into_yaml.py

Entry point of the pipeline. Reads the construction spreadsheet export (CSV) and, for every row, writes a YAML file (data/constructions/yaml/cxn_<id>.yml), a draft CoNLL-C skeleton with just the header fields filled in (data/db_conllc(NON-definitivo)/cxn_<id>.conllc — still needs a human to fill in the actual construction rows, see the CoNLL-C format guide), and one raw example text stub per example sentence (data/db_esempi(NON-definitivo)/cxn_<id>_example_<n>.txt). Also fuzzy-matches each row’s free-text formal/functional tags against cc-database/cc-database.yaml comparative-concept IDs (Levenshtein distance ≤ 2).

Run from the repo root:

python3 adoc-tools/src/export_into_yaml.py data/db-export.csv

Flags (--cc-database, --yml-out, --conllc-out, --examples-out) override the defaults shown above if you want to write somewhere else — see --help.