Documentation
¶
Overview ¶
Command gen builds data/names.tsv.gz: one-line glosses for the words in the built-in list that no dictionary defines.
The inputs are large -- the US census surname file is 9MB, the GeoNames city dump 40MB -- but only the rows matching an otherwise-undefined word are kept, so the output is tens of kilobytes. Because the word list is embedded and fixed, that gap is known ahead of time and this can run once, with the result committed.
Run with: go run ./data/gen
Sources:
US Census 2010 surnames public domain (a work of the US government) GeoNames cities1000 CC-BY 4.0 en.wiktionary.org name origins, taken as structured template facts
Click to show internal directories.
Click to hide internal directories.