Closed fb55 closed 11 years ago
Thanks! Could you please run grunt fetch
and amend the commit, adding the modified file in data/
?
gtm?
In 0fc3556bba9b542a149d6d97a269f961a5dea0ae I’ve added the missing parseError()
statement, and some tests. Thanks!
Since we can only remove numbers from data/decode-map-overrides.json
as long as they’re still listed in data/invalid-code-points.json
(see my inline comment), perhaps it’s easier to perform the optimization you suggested near the bottom of scripts/scrape-spec.js
, before the writeJSON
calls.
Nice :)
There was a bug in my patch, leading to incorrect results. I fixed this in a later commit; see https://github.com/mathiasbynens/he/compare/8bd18e6cdf4071f04c0ed9583f2d96b500db1da3...842c259d3923bf8ddd7a7fe79f77527cb81bbeb7#diff-5.
Probably as a simplification, the HTML5 tokenizer spec contains several characters that map to themselves. They aren't required for decoding, so this PR removes them.