Repository navigation

Fix Unicode parser correctness across historical and current releases - #110

Merged

mathiasbynens merged 45 commits into

node-unicode:mainfrom

JLHwung:fix/unicode-parser-correctness

Oct 25, 2026

Merged

Fix Unicode parser correctness across historical and current releases#110
mathiasbynens merged 45 commits into
node-unicode:mainfrom
JLHwung:fix/unicode-parser-correctness

Commits on Oct 25, 2026
- Fix algorithmic Hangul decomposition mappings
  
  JLHwung
  authored and
  mathiasbynens
  committed
  d7c5ce9
- Parse Unicode property records independently of separator spacing
  
  JLHwung
  authored and
  mathiasbynens
  committed
  0fcdd58
- Preserve six-digit code points in legacy property ranges
  
  JLHwung
  authored and
  mathiasbynens
  committed
  370bbff
- Honor historical Common defaults for the Script property
  
  JLHwung
  authored and
  mathiasbynens
  committed
  2f1132a
- Canonicalize historical Script property value names
  
  JLHwung
  authored and
  mathiasbynens
  committed
  51d8c08
- Canonicalize binary property aliases such as STerm
  
  JLHwung
  authored and
  mathiasbynens
  committed
  0a601a1
- Preserve Indic_Conjunct_Break as an enumerated property
  
  JLHwung
  authored and
  mathiasbynens
  committed
  bcf6560
- Exclude NFKC simple casefold mappings from binary properties
  
  JLHwung
  authored and
  mathiasbynens
  committed
  f573453
- Derive implicit joining types for historical ArabicShaping data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  eec7af7
- Infer mapping declaration value types from parsed data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  789253e
- Apply versioned bidi class defaults to unlisted code points
  
  JLHwung
  authored and
  mathiasbynens
  committed
  6f920e4
- Generate formal names for algorithmic character ranges
  
  JLHwung
  authored and
  mathiasbynens
  committed
  9416471
- Normalize Unicode data line endings before parsing
  
  JLHwung
  authored and
  mathiasbynens
  committed
  a8b6dbc
- Canonicalize historical White_space property names
  
  JLHwung
  authored and
  mathiasbynens
  committed
  231c4bb
- Preserve noncharacter records in legacy Unicode properties
  
  JLHwung
  authored and
  mathiasbynens
  committed
  fa99759
- Parse SpecialCasing comments and whitespace safely
  
  JLHwung
  authored and
  mathiasbynens
  committed
  7ec9f84
- Accept flexible separators in emoji and bidi bracket data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  5c64d10
- Strip inline comments before interpreting Unicode data fields
  
  JLHwung
  authored and
  mathiasbynens
  committed
  51db4f3
- Declare the generated root index as a default export
  
  JLHwung
  authored and
  mathiasbynens
  committed
  954fcbb
- Generate the No_Block default for gaps between blocks
  
  JLHwung
  authored and
  mathiasbynens
  committed
  77d35e9
- Apply the simple titlecase fallback for Unicode 4 and later
  
  JLHwung
  authored and
  mathiasbynens
  committed
  6f33a74
- Preserve explicit Non_Joining records in the default set
  
  JLHwung
  authored and
  mathiasbynens
  committed
  e723c07
- Ignore whitespace-only lines and indented parser comments
  
  JLHwung
  authored and
  mathiasbynens
  committed
  44481fb
- Split case-folding expansions on whitespace runs
  
  JLHwung
  authored and
  mathiasbynens
  committed
  6483160
- Split script extension values on whitespace runs
  
  JLHwung
  authored and
  mathiasbynens
  committed
  458c85e
- Accumulate discontiguous ranges in legacy Unicode blocks
  
  JLHwung
  authored and
  mathiasbynens
  committed
  e82c9ef
- Generate correction aliases from historical two-field data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  b570656
- Strip inline comments before parsing simple case mappings
  
  JLHwung
  authored and
  mathiasbynens
  committed
  dc18217
- Ignore diagnostic comparisons in legacy Unicode property dumps
  
  JLHwung
  authored and
  mathiasbynens
  committed
  b1c5a5b
- Skip historical Quick_Check values in normalization properties
  
  JLHwung
  authored and
  mathiasbynens
  committed
  5457e06
- Assign the Unicode 1.1.5 Unified Han range
  
  JLHwung
  authored and
  mathiasbynens
  committed
  666e247
- Let the narrower block entry win in legacy block data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  2e0a911
- Trim trailing spaces from character names
  
  JLHwung
  authored and
  mathiasbynens
  committed
  7a7548d
- Assign the Unicode 1.1.5 CJK compatibility ideographs
  
  JLHwung
  authored and
  mathiasbynens
  committed
  bd41213
- Assign supplementary private-use planes before Unicode 3.0.1
  
  JLHwung
  authored and
  mathiasbynens
  committed
  e06c8bb
- Generate the missing Unicode 1.1 Han character names
  
  JLHwung
  authored and
  mathiasbynens
  committed
  dbe4426
- Use PropList-3.1.1 for Unicode 3.1.1
  
  JLHwung
  authored and
  mathiasbynens
  committed
  3d2b3e3
- Apply the AL default for unlisted assigned Line_Break code points
  
  JLHwung
  authored and
  mathiasbynens
  committed
  2f7fb60
- Add Unicode 3.0 Line_Break data
  
  JLHwung
  authored and
  mathiasbynens
  committed
  b34b766
- Add composition exclusions for Unicode 3.0 through 4.0.1
  
  JLHwung
  authored and
  mathiasbynens
  committed
  92562cd
- Add Bidi mirroring data for Unicode 3.0.1 through 4.0.1
  
  JLHwung
  authored and
  mathiasbynens
  committed
  43e83a1
- Add Joining_Type data for Unicode 2.0 through 4.0.1
  
  JLHwung
  authored and
  mathiasbynens
  committed
  3d6da4e
- Reuse PropList-2.1.2 for Unicode 2.1.5
  
  JLHwung
  authored and
  mathiasbynens
  committed
  9813bb1
- Preserve ZERO WIDTH JOINER in legacy joining defaults
  
  JLHwung
  authored and
  mathiasbynens
  committed
  37bd6b5
- Confine historical data workarounds to the versions that need them
  
  JLHwung
  authored and
  mathiasbynens
  committed
  45d0131

Commits

Commits on Oct 25, 2026