jqueguiner / words2num2 / 25231306961

87%

push

github

Add auto-parse: currency + units + configurable separators

* converters/auto.py: auto_parse() and auto_parse_sentence() extract
  numeric values plus their unit from free-text expressions. Returns a
  Quantity dataclass with value, unit (canonical short), unit_long,
  kind, confidence, raw.
* formats.py: parse_number_string() with caller-overridable
  thousands/decimal separators, per-locale CLDR-style defaults for 50+
  locales, and an auto-detect heuristic when no hint is given.
* Currency: \$/€/£/¥/₹/₽/₩/₺ symbols + ISO codes (USD/EUR/...) as
  prefix or suffix; scale shortcuts (\$5m, \$1.5b, \$2.5t).
* Units: length (mm/cm/dm/m/km/in/ft/yd/mi/nm/µm), mass (mg/g/kg/t/lb/oz),
  temperature (°/°C/°F/K), time (ms/s/min/h/d), volume (ml/cl/dl/L/gal),
  percent. Word-form aliases for English ('forty-two kg' works).
* Disambiguation: prefer={"m": "mile"} overrides ambiguous unit tokens.
* Sentence mode preserves punctuation, adds expand=True to render the
  long unit form ('12.5 dollar' vs '12.5 USD').
* 55 new tests; full suite is 114 passing.

222 of 268 new or added lines in 3 files covered. (82.84%)

955 of 1099 relevant lines covered (86.9%)

5.21 hits per line