Refactor unicode generation to fully use ucd-parse

ridiculousfish · Jan 18, 2024 · 9effd51 · 9effd51
1 parent b14bd46
commit 9effd51
Show file tree

Hide file tree

Showing 8 changed files with 626 additions and 627 deletions.
diff --git a/Cargo.toml b/Cargo.toml
@@ -35,3 +35,6 @@ utf16 = []
 [dependencies]
 hashbrown = "0.13.2"
 memchr = { version = "2.4.0", default-features = false }
+
+[patch.crates-io]
+ucd-parse = { git = "https://github.com/raskad/ucd-generate.git", branch = "add-derived-normalization-properties" }
diff --git a/gen-unicode/Cargo.toml b/gen-unicode/Cargo.toml
@@ -7,5 +7,5 @@ license = "MIT OR Apache-2.0"
 edition = "2021"
 
 [dependencies]
-codegen = "0.1.3"
+codegen = "0.2.0"
 ucd-parse = "0.1.12"
diff --git a/gen-unicode/README.md b/gen-unicode/README.md
@@ -7,14 +7,6 @@ This crate generates unicode tables and code specific for regress.
 1. Download the needed unicode source files:
 
     ```sh
-    curl -L http://ftp.unicode.org/Public/UNIDATA/CaseFolding.txt -o CaseFolding.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/extracted/DerivedBinaryProperties.txt -o DerivedBinaryProperties.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/DerivedCoreProperties.txt -o DerivedCoreProperties.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/extracted/DerivedGeneralCategory.txt -o DerivedGeneralCategory.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/DerivedNormalizationProps.txt -o DerivedNormalizationProps.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/emoji/emoji-data.txt -o emoji-data.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/PropList.txt -o PropList.txt
-    curl -L http://ftp.unicode.org/Public/UNIDATA/Scripts.txt -o Scripts.txt
     mkdir /tmp/ucd-15.0.0
     cd /tmp/ucd-15.0.0
     curl -LO https://www.unicode.org/Public/zipped/15.0.0/UCD.zip