Upgrade to Unicode 16.0

yeslogic · Oct 3, 2024 · 5bcf6ef · 5bcf6ef
1 parent eee2ea4
commit 5bcf6ef
Show file tree

Hide file tree

Showing 5 changed files with 128 additions and 170 deletions.
diff --git a/Makefile b/Makefile
@@ -1,6 +1,9 @@
 tables:
-	yeslogic-ucd-generate case-folding-simple --rust-match ../ucd-generate/ucd-15.0.0/ > src/case_folding_simple.rs
-	yeslogic-ucd-generate case-mapping ../ucd-generate/ucd-15.0.0/ > src/tables.rs
+	yeslogic-ucd-generate case-folding-simple --rust-match ../ucd-generate/ucd-16.0.0/ > src/case_folding_simple.rs
+	yeslogic-ucd-generate case-mapping ../ucd-generate/ucd-16.0.0/ > src/tables.rs
+	cargo fmt
+	sed -i.bak -E '/\(([0-9]+), &\[\1\]\)/d' src/tables.rs
+	rm src/tables.rs.bak
 
 
 .PHONY: tables

diff --git a/README.md b/README.md
@@ -10,7 +10,7 @@ unicode-case-mapping
   <a href="https://crates.io/crates/unicode-case-mapping">
     <img src="https://img.shields.io/crates/v/unicode-case-mapping.svg" alt="Version">
   </a>
-  <img src="https://img.shields.io/badge/unicode-15.0-informational" alt="Unicode Version">
+  <img src="https://img.shields.io/badge/unicode-16.0-informational" alt="Unicode Version">
   <a href="https://github.com/yeslogic/unicode-case-mapping/blob/master/LICENSE">
     <img src="https://img.shields.io/crates/l/unicode-case-mapping.svg" alt="License">
   </a>
@@ -19,7 +19,7 @@ unicode-case-mapping
 <br>
 
 Fast mapping of a `char` to lowercase, uppercase, titlecase, or its simple case folding
-in Rust using Unicode 15.0 data.
+in Rust using Unicode 16.0 data.
 
 Usage
 -----
@@ -79,10 +79,8 @@ some runs of repeated values in the first level array.
 Regenerating `tables.rs`
 ------------------------
 
-1. Regenerate with `yeslogic-ucd-generate` (see header of file).
-2. Add `#[allow(dead_code)]` to each table to prevent warnings.
-3. Delete entries that map to themselves. E.g. in Vim:
-   `:g/(\(\d\+\), &\[\1\])/d`.
+1. Regenerate with `yeslogic-ucd-generate` (run `make`).
+2. Add/restore `#[allow(dead_code)]` to each table to prevent warnings.
 
 [ucd-generate]: https://github.com/yeslogic/ucd-generate
 [to_uppercase]: https://doc.rust-lang.org/std/primitive.char.html#method.to_uppercase

diff --git a/src/case_folding_simple.rs b/src/case_folding_simple.rs
@@ -1,10 +1,10 @@
 // DO NOT EDIT THIS FILE. IT WAS AUTOMATICALLY GENERATED BY:
 //
-//   yeslogic-ucd-generate case-folding-simple --rust-match ../ucd-generate/ucd-15.0.0/
+//   yeslogic-ucd-generate case-folding-simple --rust-match ../ucd-generate/ucd-16.0.0/
 //
-// Unicode version: 15.0.0.
+// Unicode version: 16.0.0.
 //
-// yeslogic-ucd-generate 0.6.0 is available on crates.io.
+// yeslogic-ucd-generate 0.7.0 is available on crates.io.
 
 use std::num::NonZeroU32;
 
@@ -547,6 +547,7 @@ pub fn case_folding_simple(cp: u32) -> Option<NonZeroU32> {
             7302 => Some(NonZeroU32::new_unchecked(1098)),
             7303 => Some(NonZeroU32::new_unchecked(1123)),
             7304 => Some(NonZeroU32::new_unchecked(42571)),
+            7305 => Some(NonZeroU32::new_unchecked(7306)),
             7312 => Some(NonZeroU32::new_unchecked(4304)),
             7313 => Some(NonZeroU32::new_unchecked(4305)),
             7314 => Some(NonZeroU32::new_unchecked(4306)),
@@ -801,10 +802,12 @@ pub fn case_folding_simple(cp: u32) -> Option<NonZeroU32> {
             8138 => Some(NonZeroU32::new_unchecked(8052)),
             8139 => Some(NonZeroU32::new_unchecked(8053)),
             8140 => Some(NonZeroU32::new_unchecked(8131)),
+            8147 => Some(NonZeroU32::new_unchecked(912)),
             8152 => Some(NonZeroU32::new_unchecked(8144)),
             8153 => Some(NonZeroU32::new_unchecked(8145)),
             8154 => Some(NonZeroU32::new_unchecked(8054)),
             8155 => Some(NonZeroU32::new_unchecked(8055)),
+            8163 => Some(NonZeroU32::new_unchecked(944)),
             8168 => Some(NonZeroU32::new_unchecked(8160)),
             8169 => Some(NonZeroU32::new_unchecked(8161)),
             8170 => Some(NonZeroU32::new_unchecked(8058)),
@@ -1097,9 +1100,13 @@ pub fn case_folding_simple(cp: u32) -> Option<NonZeroU32> {
             42950 => Some(NonZeroU32::new_unchecked(7566)),
             42951 => Some(NonZeroU32::new_unchecked(42952)),
             42953 => Some(NonZeroU32::new_unchecked(42954)),
+            42955 => Some(NonZeroU32::new_unchecked(612)),
+            42956 => Some(NonZeroU32::new_unchecked(42957)),
             42960 => Some(NonZeroU32::new_unchecked(42961)),
             42966 => Some(NonZeroU32::new_unchecked(42967)),
             42968 => Some(NonZeroU32::new_unchecked(42969)),
+            42970 => Some(NonZeroU32::new_unchecked(42971)),
+            42972 => Some(NonZeroU32::new_unchecked(411)),
             42997 => Some(NonZeroU32::new_unchecked(42998)),
             43888 => Some(NonZeroU32::new_unchecked(5024)),
             43889 => Some(NonZeroU32::new_unchecked(5025)),
@@ -1181,6 +1188,7 @@ pub fn case_folding_simple(cp: u32) -> Option<NonZeroU32> {
             43965 => Some(NonZeroU32::new_unchecked(5101)),
             43966 => Some(NonZeroU32::new_unchecked(5102)),
             43967 => Some(NonZeroU32::new_unchecked(5103)),
+            64261 => Some(NonZeroU32::new_unchecked(64262)),
             65313 => Some(NonZeroU32::new_unchecked(65345)),
             65314 => Some(NonZeroU32::new_unchecked(65346)),
             65315 => Some(NonZeroU32::new_unchecked(65347)),
@@ -1369,6 +1377,28 @@ pub fn case_folding_simple(cp: u32) -> Option<NonZeroU32> {
             68784 => Some(NonZeroU32::new_unchecked(68848)),
             68785 => Some(NonZeroU32::new_unchecked(68849)),
             68786 => Some(NonZeroU32::new_unchecked(68850)),
+            68944 => Some(NonZeroU32::new_unchecked(68976)),
+            68945 => Some(NonZeroU32::new_unchecked(68977)),
+            68946 => Some(NonZeroU32::new_unchecked(68978)),
+            68947 => Some(NonZeroU32::new_unchecked(68979)),
+            68948 => Some(NonZeroU32::new_unchecked(68980)),
+            68949 => Some(NonZeroU32::new_unchecked(68981)),
+            68950 => Some(NonZeroU32::new_unchecked(68982)),
+            68951 => Some(NonZeroU32::new_unchecked(68983)),
+            68952 => Some(NonZeroU32::new_unchecked(68984)),
+            68953 => Some(NonZeroU32::new_unchecked(68985)),
+            68954 => Some(NonZeroU32::new_unchecked(68986)),
+            68955 => Some(NonZeroU32::new_unchecked(68987)),
+            68956 => Some(NonZeroU32::new_unchecked(68988)),
+            68957 => Some(NonZeroU32::new_unchecked(68989)),
+            68958 => Some(NonZeroU32::new_unchecked(68990)),
+            68959 => Some(NonZeroU32::new_unchecked(68991)),
+            68960 => Some(NonZeroU32::new_unchecked(68992)),
+            68961 => Some(NonZeroU32::new_unchecked(68993)),
+            68962 => Some(NonZeroU32::new_unchecked(68994)),
+            68963 => Some(NonZeroU32::new_unchecked(68995)),
+            68964 => Some(NonZeroU32::new_unchecked(68996)),
+            68965 => Some(NonZeroU32::new_unchecked(68997)),
             71840 => Some(NonZeroU32::new_unchecked(71872)),
             71841 => Some(NonZeroU32::new_unchecked(71873)),
             71842 => Some(NonZeroU32::new_unchecked(71874)),

diff --git a/src/lib.rs b/src/lib.rs
@@ -7,7 +7,7 @@
 //!
 //! // U+0307 is COMBINING DOT ABOVE
 //! assert_eq!(unicode_case_mapping::to_lowercase('İ'), ['i' as u32, 0x0307]);
-//! assert_eq!(unicode_case_mapping::to_lowercase('ß'), ['ß' as u32, 0]);
+//! assert_eq!(unicode_case_mapping::to_lowercase('ß'), [0; 2]);
 //! assert_eq!(unicode_case_mapping::to_uppercase('ß'), ['S' as u32, 'S' as u32, 0]);
 //! assert_eq!(unicode_case_mapping::to_titlecase('ß'), ['S' as u32, 's' as u32, 0]);
 //! assert_eq!(unicode_case_mapping::to_titlecase('-'), [0; 3]);
@@ -23,4 +23,4 @@ pub use case_mapping::{case_folded, to_lowercase, to_titlecase, to_uppercase};
 
 /// The version of [Unicode](http://www.unicode.org/)
 /// that this version of unicode-case-mapping was generated from.
-pub const UNICODE_VERSION: (u64, u64, u64) = (15, 0, 0);
+pub const UNICODE_VERSION: (u64, u64, u64) = (16, 0, 0);