majiayu000 tables CLI
Inspect a denmark statistics fact table before loading it into Postgres and view its official table metadata.
Run the CLI from .claude/skills/tables:
python scripts/tables.py --help
view — preview a fact table
python scripts/tables.py view FOLK1A --rows 5
python scripts/tables.py view FOLK1A --db-format # show the DB-ready schema
--rows/-ncontrols how many rows to display (default 10).--db-format/-druns the same processing used during ingestion (renames to lower-case, normalizes dtypes, parses date strings).
Output is always a column|dtype list followed by a pipe-delimited preview (TABLE.head(n)), which stays consistent whether you use the raw or DB-formatted view.
info — metadata snapshot
python scripts/tables.py info FOLK1A
Prints metadata for a single table, including id, title, description, unit and every dimension (text plus the coded values present in the metadata). The metadata is printed as XML. The XML is easy to feed into other tools or to scan quickly when deciding which dimensions to join.
python scripts/tables.py info FOLK1A --column "område"
Will print only the coded values ("id") and the text titles for the column specifiec. All unique values will be printed.
python scripts/tables.py info FOLK1A --normalize-col-names
Will normalize column names so they match the database schema. This means lowercasing col names and replacing å -> a, ø -> o, æ -> ae. So OMRÅDE becomes omrade.