diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml
index b0e7659b..fb2af07d 100644
--- a/.github/workflows/main.yml
+++ b/.github/workflows/main.yml
@@ -32,6 +32,9 @@ jobs:
     - name: Run tests
       run: cargo test --verbose
 
+    - name: Run tests with icu4x
+      run: cargo test --verbose --features icu
+
     - name: Rustfmt
       run: cargo fmt --all -- --check
 
diff --git a/harfrust/Cargo.toml b/harfrust/Cargo.toml
index 92e4418e..d791b355 100644
--- a/harfrust/Cargo.toml
+++ b/harfrust/Cargo.toml
@@ -17,12 +17,15 @@ bitflags = "2.9"
 bytemuck = { version = "1.22", features = ["extern_crate_alloc"] }
 core_maths = { version = "0.1", optional = true }
 smallvec = "1.14"
+icu_properties = { version = "2.2.0", optional = true }
+icu_normalizer = { version = "2.2.0", optional = true }
 
 [features]
 default = ["std"]
 std = ["read-fonts/std"]
 libm = ["dep:core_maths", "read-fonts/libm"]
 experimental_font_api = []
+icu = ["dep:icu_properties", "dep:icu_normalizer"]
 
 [dev-dependencies]
 harfrust = { path = "../harfrust", features = ["experimental_font_api"] }
diff --git a/harfrust/src/hb/algs.rs b/harfrust/src/algs.rs
similarity index 92%
rename from harfrust/src/hb/algs.rs
rename to harfrust/src/algs.rs
index 4a4b78fb..b7e1a421 100644
--- a/harfrust/src/hb/algs.rs
+++ b/harfrust/src/algs.rs
@@ -14,12 +14,6 @@ pub fn rb_flag_unsafe(x: u32) -> u32 {
     }
 }
 
-// FLAG_RANGE macro in harfbuzz.
-#[inline]
-pub fn rb_flag_range(x: u32, y: u32) -> u32 {
-    (x < y) as u32 + rb_flag(y + 1) - rb_flag(x)
-}
-
 // FLAG64 macro in harfbuzz.
 #[inline]
 pub const fn rb_flag64(x: u32) -> u64 {
diff --git a/harfrust/src/hb/aat/mod.rs b/harfrust/src/hb/aat/mod.rs
index bfd4ca3c..75efa155 100644
--- a/harfrust/src/hb/aat/mod.rs
+++ b/harfrust/src/hb/aat/mod.rs
@@ -81,7 +81,7 @@ pub struct AatTables<'a> {
     pub feat: Option<Feat<'a>>,
 }
 
-use crate::hb::algs::HB_CODEPOINT_ENCODE3 as encode3;
+use crate::algs::HB_CODEPOINT_ENCODE3 as encode3;
 
 /// Blocklist specific broken morx tables identified by the combination of
 /// morx, GSUB, and GDEF table lengths.
diff --git a/harfrust/src/hb/buffer.rs b/harfrust/src/hb/buffer.rs
index b889a5d5..f8af0264 100644
--- a/harfrust/src/hb/buffer.rs
+++ b/harfrust/src/hb/buffer.rs
@@ -1,11 +1,10 @@
 use super::hb_mask_t;
-use super::unicode::CharExt;
 use crate::hb::face::BasicFontMetrics;
 use crate::hb::glyph_metrics::GlyphMetrics;
 use crate::hb::glyph_names::GlyphNames;
 use crate::hb::set_digest::hb_set_digest_t;
 use crate::hb::tables::TableRanges;
-use crate::hb::unicode::Codepoint;
+use crate::unicode::{CharExt, Codepoint};
 use crate::U32Set;
 use crate::{script, BufferClusterLevel, BufferFlags, Direction, Language, Script, SerializeFlags};
 use alloc::{string::String, vec::Vec};
diff --git a/harfrust/src/hb/mod.rs b/harfrust/src/hb/mod.rs
index c6dba1de..0ee74c93 100644
--- a/harfrust/src/hb/mod.rs
+++ b/harfrust/src/hb/mod.rs
@@ -20,7 +20,6 @@
 #![allow(clippy::match_like_matches_macro)]
 #![allow(clippy::manual_range_contains)]
 
-mod algs;
 #[macro_use]
 pub mod buffer;
 mod aat;
@@ -72,11 +71,6 @@ mod tag;
 #[allow(clippy::collapsible_match)]
 mod tag_table;
 mod text_parser;
-#[rustfmt::skip]
-mod unicode_emoji_table;
-#[rustfmt::skip]
-mod ucd_table;
-mod unicode;
 
 use read_fonts::types::Tag as hb_tag_t;
 
diff --git a/harfrust/src/hb/ot/mod.rs b/harfrust/src/hb/ot/mod.rs
index 4b5783ba..d2568622 100644
--- a/harfrust/src/hb/ot/mod.rs
+++ b/harfrust/src/hb/ot/mod.rs
@@ -793,7 +793,7 @@ impl ClassDefInfo {
     }
 }
 
-use super::algs::HB_CODEPOINT_ENCODE3 as encode3;
+use crate::algs::HB_CODEPOINT_ENCODE3 as encode3;
 
 /// Blocklist specific broken GDEF tables identified by the combination of
 /// GDEF, GSUB, and GPOS table lengths. Nuke the GDEF tables to avoid
diff --git a/harfrust/src/hb/ot_layout.rs b/harfrust/src/hb/ot_layout.rs
index 2065d6d4..f91ff5f9 100644
--- a/harfrust/src/hb/ot_layout.rs
+++ b/harfrust/src/hb/ot_layout.rs
@@ -7,10 +7,9 @@ use super::font_funcs::FontFuncsDispatch;
 use super::ot::lookup::LookupInfo;
 use super::ot_layout_gsubgpos::OT;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
-use super::unicode::hb_unicode_funcs_t;
 use super::{hb_font_t, GlyphInfo};
 use crate::hb::ot_layout_gsubgpos::OT::check_glyph_property;
-use crate::hb::unicode::GeneralCategory;
+use crate::unicode::{hb_unicode_funcs_t, GeneralCategory};
 
 impl GlyphInfo {
     declare_buffer_var!(u16, 1, 0, GLYPH_PROPS_VAR, glyph_props, set_glyph_props);
diff --git a/harfrust/src/hb/ot_layout_gsubgpos.rs b/harfrust/src/hb/ot_layout_gsubgpos.rs
index a8771281..cc363eef 100644
--- a/harfrust/src/hb/ot_layout_gsubgpos.rs
+++ b/harfrust/src/hb/ot_layout_gsubgpos.rs
@@ -11,7 +11,7 @@ use super::ot_layout_common::*;
 use super::set_digest::hb_set_digest_t;
 use crate::hb::ot::{ClassDefInfo, CoverageInfo};
 use crate::hb::ot_layout_gsubgpos::OT::check_glyph_property;
-use crate::hb::unicode::GeneralCategory;
+use crate::unicode::GeneralCategory;
 use alloc::boxed::Box;
 use read_fonts::tables::layout::SequenceLookupRecord;
 use read_fonts::types::GlyphId;
diff --git a/harfrust/src/hb/ot_shape.rs b/harfrust/src/hb/ot_shape.rs
index 1a9c9e98..8d87d209 100644
--- a/harfrust/src/hb/ot_shape.rs
+++ b/harfrust/src/hb/ot_shape.rs
@@ -6,18 +6,11 @@ use super::ot_layout_gpos_table::GPOS;
 use super::ot_map::*;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
-use super::unicode::CharExt;
 use super::*;
 use super::{hb_font_t, hb_tag_t};
 use crate::hb::aat;
-use crate::hb::algs::{rb_flag, rb_flag_unsafe};
 use crate::hb::buffer::GlyphFlags;
-use crate::hb::unicode::hb_gc::{
-    HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER, HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER,
-    HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR, HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER,
-    HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER,
-};
-use crate::hb::unicode::GeneralCategory;
+use crate::unicode::{CharExt, GeneralCategory};
 use crate::BufferFlags;
 use crate::{Direction, Feature, Language, Script};
 use core::ptr;
@@ -664,12 +657,12 @@ impl OtShapeContext<'_, '_> {
 
             let gen_cat = info.general_category();
 
-            if (rb_flag_unsafe(gen_cat.to_u8() as u32)
-                & (rb_flag(HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER)
-                    | rb_flag(HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER)
-                    | rb_flag(HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER)
-                    | rb_flag(HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER)
-                    | rb_flag(HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR)))
+            if gen_cat.flag_unsafe()
+                & (GeneralCategory::LOWERCASE_LETTER.flag()
+                    | GeneralCategory::UPPERCASE_LETTER.flag()
+                    | GeneralCategory::TITLECASE_LETTER.flag()
+                    | GeneralCategory::OTHER_LETTER.flag()
+                    | GeneralCategory::SPACE_SEPARATOR.flag())
                 != 0
             {
                 i += 1;
@@ -767,7 +760,7 @@ impl OtShapeContext<'_, '_> {
             let rtlm_mask = self.plan.rtlm_mask;
 
             for info in &mut self.buffer.info[..len] {
-                if let Some(c) = info.as_codepoint().mirrored() {
+                if let Some(c) = info.as_codepoint().mirroring() {
                     if self.font_funcs.nominal_glyph(c).is_some() {
                         info.glyph_id = c;
                         continue;
diff --git a/harfrust/src/hb/ot_shape_fallback.rs b/harfrust/src/hb/ot_shape_fallback.rs
index 89e8a450..4ff54560 100644
--- a/harfrust/src/hb/ot_shape_fallback.rs
+++ b/harfrust/src/hb/ot_shape_fallback.rs
@@ -4,8 +4,8 @@ use super::buffer::{hb_buffer_t, GlyphPosition};
 use super::face::GlyphExtents;
 use super::font_funcs::FontFuncsDispatch;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
-use super::unicode::*;
 use super::{hb_font_t, Direction};
+use crate::unicode::*;
 
 struct FallbackShapeContext<'a, 'x, 'u> {
     plan: &'a hb_ot_shape_plan_t,
@@ -456,7 +456,7 @@ pub fn _hb_ot_shape_fallback_spaces<'a, 'x>(
     buffer: &'x mut hb_buffer_t,
     font_funcs: &'x mut FontFuncsDispatch<'a, '_>,
 ) {
-    use super::unicode::hb_unicode_funcs_t as t;
+    use crate::unicode::hb_unicode_funcs_t as t;
 
     let _ = plan;
     let len = buffer.len;
diff --git a/harfrust/src/hb/ot_shape_normalize.rs b/harfrust/src/hb/ot_shape_normalize.rs
index 9bf0a08d..7384e866 100644
--- a/harfrust/src/hb/ot_shape_normalize.rs
+++ b/harfrust/src/hb/ot_shape_normalize.rs
@@ -1,11 +1,9 @@
-use crate::hb::unicode::Codepoint;
-
 use super::buffer::*;
 use super::font_funcs::FontFuncsDispatch;
 use super::hb_font_t;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::{ComposeFn, DecomposeFn, MAX_COMBINING_MARKS};
-use super::unicode::{hb_unicode_funcs_t, CharExt};
+use crate::unicode::{hb_unicode_funcs_t, CharExt, Codepoint};
 use read_fonts::types::GlyphId;
 
 impl GlyphInfo {
@@ -96,7 +94,7 @@ fn decompose_unicode(
     _: &hb_ot_shape_normalize_context_t,
     ab: Codepoint,
 ) -> Option<(Codepoint, Codepoint)> {
-    super::unicode::decompose(ab)
+    crate::unicode::decompose(ab)
 }
 
 fn compose_unicode(
@@ -104,7 +102,7 @@ fn compose_unicode(
     a: Codepoint,
     b: Codepoint,
 ) -> Option<Codepoint> {
-    super::unicode::compose(a, b)
+    crate::unicode::compose(a, b)
 }
 
 fn output_char(buffer: &mut hb_buffer_t, unichar: u32, glyph: u32) {
diff --git a/harfrust/src/hb/ot_shaper.rs b/harfrust/src/hb/ot_shaper.rs
index 755e7605..70f2a07a 100644
--- a/harfrust/src/hb/ot_shaper.rs
+++ b/harfrust/src/hb/ot_shaper.rs
@@ -4,8 +4,8 @@ use super::font_funcs::FontFuncsDispatch;
 use super::ot_shape::*;
 use super::ot_shape_normalize::*;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
-use super::unicode::Codepoint;
 use super::{hb_tag_t, script, Direction, Script};
+use crate::unicode::Codepoint;
 use alloc::boxed::Box;
 use core::any::Any;
 
diff --git a/harfrust/src/hb/ot_shaper_arabic.rs b/harfrust/src/hb/ot_shaper_arabic.rs
index 8a395347..634de33c 100644
--- a/harfrust/src/hb/ot_shaper_arabic.rs
+++ b/harfrust/src/hb/ot_shaper_arabic.rs
@@ -1,4 +1,3 @@
-use super::algs::*;
 use super::buffer::*;
 use super::font_funcs::FontFuncsDispatch;
 use super::ot_map::*;
@@ -6,8 +5,8 @@ use super::ot_shape::*;
 use super::ot_shape_normalize::HB_OT_SHAPE_NORMALIZATION_MODE_AUTO;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
-use super::unicode::*;
 use super::{hb_mask_t, hb_tag_t, script, GlyphInfo, Script};
+use crate::unicode::*;
 use crate::Direction;
 use alloc::boxed::Box;
 
@@ -17,21 +16,21 @@ const HB_BUFFER_SCRATCH_FLAG_ARABIC_HAS_STCH: hb_buffer_scratch_flags_t =
 // See:
 // https://github.com/harfbuzz/harfbuzz/commit/6e6f82b6f3dde0fc6c3c7d991d9ec6cfff57823d#commitcomment-14248516
 fn is_word_category(gc: GeneralCategory) -> bool {
-    (rb_flag_unsafe(gc.to_u8() as u32)
-        & (rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL)))
+    (gc.flag_unsafe()
+        & (GeneralCategory::UNASSIGNED.flag()
+            | GeneralCategory::PRIVATE_USE.flag()
+            | GeneralCategory::MODIFIER_LETTER.flag()
+            | GeneralCategory::OTHER_LETTER.flag()
+            | GeneralCategory::SPACING_MARK.flag()
+            | GeneralCategory::ENCLOSING_MARK.flag()
+            | GeneralCategory::NON_SPACING_MARK.flag()
+            | GeneralCategory::DECIMAL_NUMBER.flag()
+            | GeneralCategory::LETTER_NUMBER.flag()
+            | GeneralCategory::OTHER_NUMBER.flag()
+            | GeneralCategory::CURRENCY_SYMBOL.flag()
+            | GeneralCategory::MODIFIER_SYMBOL.flag()
+            | GeneralCategory::MATH_SYMBOL.flag()
+            | GeneralCategory::OTHER_SYMBOL.flag()))
         != 0
 }
 
@@ -54,10 +53,10 @@ fn get_joining_type(u: Codepoint, gc: GeneralCategory) -> hb_arabic_joining_type
         return j_type;
     }
 
-    let ok = rb_flag_unsafe(gc.to_u8() as u32)
-        & (rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK)
-            | rb_flag(hb_gc::HB_UNICODE_GENERAL_CATEGORY_FORMAT));
+    let ok = gc.flag_unsafe()
+        & (GeneralCategory::NON_SPACING_MARK.flag()
+            | GeneralCategory::ENCLOSING_MARK.flag()
+            | GeneralCategory::FORMAT.flag());
 
     if ok != 0 {
         hb_arabic_joining_type_t::T
diff --git a/harfrust/src/hb/ot_shaper_arabic_table.rs b/harfrust/src/hb/ot_shaper_arabic_table.rs
index 8bdb8316..2f764925 100644
--- a/harfrust/src/hb/ot_shaper_arabic_table.rs
+++ b/harfrust/src/hb/ot_shaper_arabic_table.rs
@@ -16,7 +16,7 @@
 #![allow(unused_parens)]
 #![allow(clippy::unnecessary_cast, clippy::unreadable_literal, clippy::double_parens)]
 
-use crate::hb::unicode::Codepoint;
+use crate::unicode::Codepoint;
 
 use super::ot_shaper_arabic::hb_arabic_joining_type_t::{self, D, GroupAlaph, GroupDalathRish, L, R, T, U, X};
 
diff --git a/harfrust/src/hb/ot_shaper_hangul.rs b/harfrust/src/hb/ot_shaper_hangul.rs
index b747d061..cf0d8d14 100644
--- a/harfrust/src/hb/ot_shaper_hangul.rs
+++ b/harfrust/src/hb/ot_shaper_hangul.rs
@@ -8,7 +8,7 @@ use super::ot_shape_normalize::HB_OT_SHAPE_NORMALIZATION_MODE_NONE;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
 use super::*;
-use crate::hb::unicode::Codepoint;
+use crate::unicode::Codepoint;
 use crate::BufferFlags;
 
 const LJMO: u8 = 1;
diff --git a/harfrust/src/hb/ot_shaper_hebrew.rs b/harfrust/src/hb/ot_shaper_hebrew.rs
index a4b3a2fc..61136fa0 100644
--- a/harfrust/src/hb/ot_shaper_hebrew.rs
+++ b/harfrust/src/hb/ot_shaper_hebrew.rs
@@ -1,10 +1,9 @@
+use super::hb_tag_t;
 use super::ot_shape_normalize::*;
 use super::ot_shaper::*;
-use super::{hb_tag_t, unicode};
 use crate::hb::buffer::hb_buffer_t;
 use crate::hb::ot_shape_plan::hb_ot_shape_plan_t;
-use crate::hb::unicode::Codepoint;
-use crate::hb::unicode::{combining_class, modified_combining_class};
+use crate::unicode::{self, combining_class, modified_combining_class, Codepoint};
 
 pub const HEBREW_SHAPER: hb_ot_shaper_t = hb_ot_shaper_t {
     collect_features: None,
diff --git a/harfrust/src/hb/ot_shaper_indic.rs b/harfrust/src/hb/ot_shaper_indic.rs
index 3a759077..0738befb 100644
--- a/harfrust/src/hb/ot_shaper_indic.rs
+++ b/harfrust/src/hb/ot_shaper_indic.rs
@@ -4,7 +4,6 @@ use core::ops::Range;
 
 use read_fonts::types::GlyphId;
 
-use super::algs::*;
 use super::buffer::*;
 use super::font_funcs::FontFuncsDispatch;
 use super::ot_layout::*;
@@ -15,9 +14,10 @@ use super::ot_shape_normalize::*;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
 use super::ot_shaper_syllabic::*;
-use super::unicode::Codepoint;
-use super::unicode::{hb_gc, CharExt};
 use super::{hb_font_t, hb_mask_t, hb_tag_t, script, GlyphInfo, Script};
+use crate::algs::*;
+use crate::unicode::GeneralCategory;
+use crate::unicode::{CharExt, Codepoint};
 
 pub const INDIC_SHAPER: hb_ot_shaper_t = hb_ot_shaper_t {
     collect_features: Some(collect_features),
@@ -591,7 +591,7 @@ fn decompose(_: &hb_ot_shape_normalize_context_t, ab: Codepoint) -> Option<(Code
         _ => {}
     }
 
-    crate::hb::unicode::decompose(ab)
+    crate::unicode::decompose(ab)
 }
 
 fn compose(_: &hb_ot_shape_normalize_context_t, a: Codepoint, b: Codepoint) -> Option<Codepoint> {
@@ -605,7 +605,7 @@ fn compose(_: &hb_ot_shape_normalize_context_t, a: Codepoint, b: Codepoint) -> O
         return Some(0x09DF);
     }
 
-    crate::hb::unicode::compose(a, b)
+    crate::unicode::compose(a, b)
 }
 
 fn setup_masks(_: &hb_ot_shape_plan_t, _: &mut FontFuncsDispatch, buffer: &mut hb_buffer_t) {
@@ -1833,12 +1833,9 @@ fn final_reordering_impl(
     // Apply 'init' to the Left Matra if it's a word start.
     if buffer.info[start].indic_position() == ot_position_t::POS_PRE_M {
         if start == 0
-            || (rb_flag_unsafe(buffer.info[start - 1].general_category().to_u8() as u32)
-                & rb_flag_range(
-                    hb_gc::HB_UNICODE_GENERAL_CATEGORY_FORMAT,
-                    hb_gc::HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK,
-                ))
-                == 0
+            || !buffer.info[start - 1]
+                .general_category()
+                .in_range_inclusive(GeneralCategory::FORMAT, GeneralCategory::NON_SPACING_MARK)
         {
             buffer.info[start].mask |= indic_plan.mask_array[indic_feature::INIT];
         } else {
diff --git a/harfrust/src/hb/ot_shaper_khmer.rs b/harfrust/src/hb/ot_shaper_khmer.rs
index 3679f530..bcf2b9ab 100644
--- a/harfrust/src/hb/ot_shaper_khmer.rs
+++ b/harfrust/src/hb/ot_shaper_khmer.rs
@@ -7,9 +7,8 @@ use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
 use super::ot_shaper_indic::ot_category_t;
 use super::ot_shaper_syllabic::*;
-use super::unicode::CharExt;
-use super::unicode::Codepoint;
 use super::{hb_mask_t, hb_tag_t, GlyphInfo};
+use crate::unicode::{CharExt, Codepoint};
 use alloc::boxed::Box;
 
 pub const KHMER_SHAPER: hb_ot_shaper_t = hb_ot_shaper_t {
@@ -295,7 +294,7 @@ fn decompose(_: &hb_ot_shape_normalize_context_t, ab: Codepoint) -> Option<(Code
     // Decompose split matras that don't have Unicode decompositions.
     match ab {
         0x17BE | 0x17BF | 0x17C0 | 0x17C4 | 0x17C5 => Some((0x17C1, ab)),
-        _ => crate::hb::unicode::decompose(ab),
+        _ => crate::unicode::decompose(ab),
     }
 }
 
@@ -305,7 +304,7 @@ fn compose(_: &hb_ot_shape_normalize_context_t, a: Codepoint, b: Codepoint) -> O
         return None;
     }
 
-    crate::hb::unicode::compose(a, b)
+    crate::unicode::compose(a, b)
 }
 
 fn setup_masks(_: &hb_ot_shape_plan_t, _: &mut FontFuncsDispatch, buffer: &mut hb_buffer_t) {
diff --git a/harfrust/src/hb/ot_shaper_thai.rs b/harfrust/src/hb/ot_shaper_thai.rs
index 39a9f3b9..7502ed51 100644
--- a/harfrust/src/hb/ot_shaper_thai.rs
+++ b/harfrust/src/hb/ot_shaper_thai.rs
@@ -5,7 +5,7 @@ use super::ot_shape_normalize::HB_OT_SHAPE_NORMALIZATION_MODE_AUTO;
 use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
 use super::script;
-use super::unicode::GeneralCategory;
+use crate::unicode::GeneralCategory;
 
 pub const THAI_SHAPER: hb_ot_shaper_t = hb_ot_shaper_t {
     collect_features: None,
diff --git a/harfrust/src/hb/ot_shaper_use.rs b/harfrust/src/hb/ot_shaper_use.rs
index b340bf1d..0c654a83 100644
--- a/harfrust/src/hb/ot_shaper_use.rs
+++ b/harfrust/src/hb/ot_shaper_use.rs
@@ -1,4 +1,3 @@
-use super::algs::*;
 use super::buffer::*;
 use super::font_funcs::FontFuncsDispatch;
 use super::ot_layout::*;
@@ -9,8 +8,9 @@ use super::ot_shape_plan::hb_ot_shape_plan_t;
 use super::ot_shaper::*;
 use super::ot_shaper_arabic::arabic_shape_plan_t;
 use super::ot_shaper_syllabic::*;
-use super::unicode::{CharExt, Codepoint};
 use super::{hb_mask_t, hb_tag_t, script, GlyphInfo, Script};
+use crate::algs::*;
+use crate::unicode::{CharExt, Codepoint};
 use alloc::boxed::Box;
 
 pub const UNIVERSAL_SHAPER: hb_ot_shaper_t = hb_ot_shaper_t {
@@ -567,7 +567,7 @@ fn compose(_: &hb_ot_shape_normalize_context_t, a: Codepoint, b: Codepoint) -> O
         return None;
     }
 
-    crate::hb::unicode::compose(a, b)
+    crate::unicode::compose(a, b)
 }
 
 fn setup_masks(plan: &hb_ot_shape_plan_t, _: &mut FontFuncsDispatch, buffer: &mut hb_buffer_t) {
diff --git a/harfrust/src/hb/unicode.rs b/harfrust/src/hb/unicode.rs
deleted file mode 100644
index 9ca0928c..00000000
--- a/harfrust/src/hb/unicode.rs
+++ /dev/null
@@ -1,751 +0,0 @@
-use super::ucd_table::ucd::*;
-use crate::hb::algs::*;
-use crate::Script;
-
-pub type Codepoint = u32;
-
-// Space estimates based on:
-// https://unicode.org/charts/PDF/U2000.pdf
-// https://docs.microsoft.com/en-us/typography/develop/character-design-standards/whitespace
-pub mod hb_unicode_funcs_t {
-    pub type space_t = u8;
-    pub const NOT_SPACE: u8 = 0;
-    pub const SPACE_EM: u8 = 1;
-    pub const SPACE_EM_2: u8 = 2;
-    pub const SPACE_EM_3: u8 = 3;
-    pub const SPACE_EM_4: u8 = 4;
-    pub const SPACE_EM_5: u8 = 5;
-    pub const SPACE_EM_6: u8 = 6;
-    pub const SPACE_EM_16: u8 = 16;
-    pub const SPACE_4_EM_18: u8 = 17; // 4/18th of an EM!
-    pub const SPACE: u8 = 18;
-    pub const SPACE_FIGURE: u8 = 19;
-    pub const SPACE_PUNCTUATION: u8 = 20;
-    pub const SPACE_NARROW: u8 = 21;
-}
-
-#[derive(Copy, Clone, PartialEq, Eq, Debug)]
-pub struct GeneralCategory(pub u8);
-
-#[allow(unused)]
-impl GeneralCategory {
-    pub const CONTROL: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_CONTROL as _);
-    pub const FORMAT: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_FORMAT as _);
-    pub const UNASSIGNED: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED as _);
-    pub const PRIVATE_USE: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE as _);
-    pub const SURROGATE: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_SURROGATE as _);
-    pub const LOWERCASE_LETTER: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER as _);
-    pub const MODIFIER_LETTER: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER as _);
-    pub const OTHER_LETTER: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER as _);
-    pub const TITLECASE_LETTER: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER as _);
-    pub const UPPERCASE_LETTER: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER as _);
-    pub const SPACING_MARK: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK as _);
-    pub const ENCLOSING_MARK: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK as _);
-    pub const NON_SPACING_MARK: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK as _);
-    pub const DECIMAL_NUMBER: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER as _);
-    pub const LETTER_NUMBER: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER as _);
-    pub const OTHER_NUMBER: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER as _);
-    pub const CONNECT_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION as _);
-    pub const DASH_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION as _);
-    pub const CLOSE_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION as _);
-    pub const FINAL_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION as _);
-    pub const INITIAL_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION as _);
-    pub const OTHER_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION as _);
-    pub const OPEN_PUNCTUATION: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION as _);
-    pub const CURRENCY_SYMBOL: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL as _);
-    pub const MODIFIER_SYMBOL: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL as _);
-    pub const MATH_SYMBOL: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL as _);
-    pub const OTHER_SYMBOL: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL as _);
-    pub const LINE_SEPARATOR: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR as _);
-    pub const PARAGRAPH_SEPARATOR: Self =
-        Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR as _);
-    pub const SPACE_SEPARATOR: Self = Self(hb_gc::HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR as _);
-}
-
-impl GeneralCategory {
-    pub fn to_u8(self) -> u8 {
-        self.0
-    }
-
-    pub fn is_mark(&self) -> bool {
-        matches!(
-            *self,
-            Self::SPACING_MARK | Self::ENCLOSING_MARK | Self::NON_SPACING_MARK
-        )
-    }
-
-    pub fn is_letter(&self) -> bool {
-        matches!(
-            *self,
-            Self::LOWERCASE_LETTER
-                | Self::MODIFIER_LETTER
-                | Self::OTHER_LETTER
-                | Self::TITLECASE_LETTER
-                | Self::UPPERCASE_LETTER
-        )
-    }
-}
-
-#[allow(dead_code)]
-pub mod combining_class {
-    pub const NotReordered: u8 = 0;
-    pub const Overlay: u8 = 1;
-    pub const Nukta: u8 = 7;
-    pub const KanaVoicing: u8 = 8;
-    pub const Virama: u8 = 9;
-
-    /* Hebrew */
-    pub const CCC10: u8 = 10;
-    pub const CCC11: u8 = 11;
-    pub const CCC12: u8 = 12;
-    pub const CCC13: u8 = 13;
-    pub const CCC14: u8 = 14;
-    pub const CCC15: u8 = 15;
-    pub const CCC16: u8 = 16;
-    pub const CCC17: u8 = 17;
-    pub const CCC18: u8 = 18;
-    pub const CCC19: u8 = 19;
-    pub const CCC20: u8 = 20;
-    pub const CCC21: u8 = 21;
-    pub const CCC22: u8 = 22;
-    pub const CCC23: u8 = 23;
-    pub const CCC24: u8 = 24;
-    pub const CCC25: u8 = 25;
-    pub const CCC26: u8 = 26;
-
-    /* Arabic */
-    pub const CCC27: u8 = 27;
-    pub const CCC28: u8 = 28;
-    pub const CCC29: u8 = 29;
-    pub const CCC30: u8 = 30;
-    pub const CCC31: u8 = 31;
-    pub const CCC32: u8 = 32;
-    pub const CCC33: u8 = 33;
-    pub const CCC34: u8 = 34;
-    pub const CCC35: u8 = 35;
-
-    /* Syriac */
-    pub const CCC36: u8 = 36;
-
-    /* Telugu */
-    pub const CCC84: u8 = 84;
-    pub const CCC91: u8 = 91;
-
-    /* Thai */
-    pub const CCC103: u8 = 103;
-    pub const CCC107: u8 = 107;
-
-    /* Lao */
-    pub const CCC118: u8 = 118;
-    pub const CCC122: u8 = 122;
-
-    /* Tibetan */
-    pub const CCC129: u8 = 129;
-    pub const CCC130: u8 = 130;
-    pub const CCC132: u8 = 132;
-
-    pub const AttachedBelowLeft: u8 = 200;
-    pub const AttachedBelow: u8 = 202;
-    pub const AttachedAbove: u8 = 214;
-    pub const AttachedAboveRight: u8 = 216;
-    pub const BelowLeft: u8 = 218;
-    pub const Below: u8 = 220;
-    pub const BelowRight: u8 = 222;
-    pub const Left: u8 = 224;
-    pub const Right: u8 = 226;
-    pub const AboveLeft: u8 = 228;
-    pub const Above: u8 = 230;
-    pub const AboveRight: u8 = 232;
-    pub const DoubleBelow: u8 = 233;
-    pub const DoubleAbove: u8 = 234;
-
-    pub const IotaSubscript: u8 = 240;
-
-    pub const Invalid: u8 = 255;
-}
-
-#[allow(dead_code)]
-pub mod modified_combining_class {
-    // Hebrew
-    //
-    // We permute the "fixed-position" classes 10-26 into the order
-    // described in the SBL Hebrew manual:
-    //
-    // https://www.sbl-site.org/Fonts/SBLHebrewUserManual1.5x.pdf
-    //
-    // (as recommended by:
-    //  https://forum.fontlab.com/archive-old-microsoft-volt-group/vista-and-diacritic-ordering/msg22823/)
-    //
-    // More details here:
-    // https://bugzilla.mozilla.org/show_bug.cgi?id=662055
-    pub const CCC10: u8 = 22; // sheva
-    pub const CCC11: u8 = 15; // hataf segol
-    pub const CCC12: u8 = 16; // hataf patah
-    pub const CCC13: u8 = 17; // hataf qamats
-    pub const CCC14: u8 = 23; // hiriq
-    pub const CCC15: u8 = 18; // tsere
-    pub const CCC16: u8 = 19; // segol
-    pub const CCC17: u8 = 20; // patah
-    pub const CCC18: u8 = 21; // qamats & qamats qatan
-    pub const CCC19: u8 = 14; // holam & holam haser for vav
-    pub const CCC20: u8 = 24; // qubuts
-    pub const CCC21: u8 = 12; // dagesh
-    pub const CCC22: u8 = 25; // meteg
-    pub const CCC23: u8 = 13; // rafe
-    pub const CCC24: u8 = 10; // shin dot
-    pub const CCC25: u8 = 11; // sin dot
-    pub const CCC26: u8 = 26; // point varika
-
-    // Arabic
-    //
-    // Modify to move Shadda (ccc=33) before other marks.  See:
-    // https://unicode.org/faq/normalization.html#8
-    // https://unicode.org/faq/normalization.html#9
-    pub const CCC27: u8 = 28; // fathatan
-    pub const CCC28: u8 = 29; // dammatan
-    pub const CCC29: u8 = 30; // kasratan
-    pub const CCC30: u8 = 31; // fatha
-    pub const CCC31: u8 = 32; // damma
-    pub const CCC32: u8 = 33; // kasra
-    pub const CCC33: u8 = 27; // shadda
-    pub const CCC34: u8 = 34; // sukun
-    pub const CCC35: u8 = 35; // superscript alef
-
-    // Syriac
-    pub const CCC36: u8 = 36; // superscript alaph
-
-    // Telugu
-    //
-    // Modify Telugu length marks (ccc=84, ccc=91).
-    // These are the only matras in the main Indic scripts range that have
-    // a non-zero ccc.  That makes them reorder with the Halant that is
-    // ccc=9.  Just zero them, we don't need them in our Indic shaper.
-    pub const CCC84: u8 = 0; // length mark
-    pub const CCC91: u8 = 0; // ai length mark
-
-    // Thai
-    //
-    // Modify U+0E38 and U+0E39 (ccc=103) to be reordered before U+0E3A (ccc=9).
-    // Assign 3, which is unassigned otherwise.
-    // Uniscribe does this reordering too.
-    pub const CCC103: u8 = 3; // sara u / sara uu
-    pub const CCC107: u8 = 107; // mai *
-
-    // Lao
-    pub const CCC118: u8 = 118; // sign u / sign uu
-    pub const CCC122: u8 = 122; // mai *
-
-    // Tibetan
-    //
-    // In case of multiple vowel-signs, use u first (but after achung)
-    // this allows Dzongkha multi-vowel shortcuts to render correctly
-    pub const CCC129: u8 = 129; // sign aa
-    pub const CCC130: u8 = 132; // sign i
-    pub const CCC132: u8 = 131; // sign u
-}
-
-#[rustfmt::skip]
-static MODIFIED_COMBINING_CLASS: &[u8; 256] = &[
-    combining_class::NotReordered,
-    combining_class::Overlay,
-    2, 3, 4, 5, 6,
-    combining_class::Nukta,
-    combining_class::KanaVoicing,
-    combining_class::Virama,
-
-    // Hebrew
-    modified_combining_class::CCC10,
-    modified_combining_class::CCC11,
-    modified_combining_class::CCC12,
-    modified_combining_class::CCC13,
-    modified_combining_class::CCC14,
-    modified_combining_class::CCC15,
-    modified_combining_class::CCC16,
-    modified_combining_class::CCC17,
-    modified_combining_class::CCC18,
-    modified_combining_class::CCC19,
-    modified_combining_class::CCC20,
-    modified_combining_class::CCC21,
-    modified_combining_class::CCC22,
-    modified_combining_class::CCC23,
-    modified_combining_class::CCC24,
-    modified_combining_class::CCC25,
-    modified_combining_class::CCC26,
-
-    // Arabic
-    modified_combining_class::CCC27,
-    modified_combining_class::CCC28,
-    modified_combining_class::CCC29,
-    modified_combining_class::CCC30,
-    modified_combining_class::CCC31,
-    modified_combining_class::CCC32,
-    modified_combining_class::CCC33,
-    modified_combining_class::CCC34,
-    modified_combining_class::CCC35,
-
-    // Syriac
-    modified_combining_class::CCC36,
-
-    37, 38, 39,
-    40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59,
-    60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79,
-    80, 81, 82, 83,
-
-    // Telugu
-    modified_combining_class::CCC84,
-    85, 86, 87, 88, 89, 90,
-    modified_combining_class::CCC91,
-    92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102,
-
-    // Thai
-    modified_combining_class::CCC103,
-    104, 105, 106,
-    modified_combining_class::CCC107,
-    108, 109, 110, 111, 112, 113, 114, 115, 116, 117,
-
-    // Lao
-    modified_combining_class::CCC118,
-    119, 120, 121,
-    modified_combining_class::CCC122,
-    123, 124, 125, 126, 127, 128,
-
-    // Tibetan
-    modified_combining_class::CCC129,
-    modified_combining_class::CCC130,
-    131,
-    modified_combining_class::CCC132,
-    133, 134, 135, 136, 137, 138, 139,
-
-
-    140, 141, 142, 143, 144, 145, 146, 147, 148, 149,
-    150, 151, 152, 153, 154, 155, 156, 157, 158, 159,
-    160, 161, 162, 163, 164, 165, 166, 167, 168, 169,
-    170, 171, 172, 173, 174, 175, 176, 177, 178, 179,
-    180, 181, 182, 183, 184, 185, 186, 187, 188, 189,
-    190, 191, 192, 193, 194, 195, 196, 197, 198, 199,
-
-    combining_class::AttachedBelowLeft,
-    201,
-    combining_class::AttachedBelow,
-    203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213,
-    combining_class::AttachedAbove,
-    215,
-    combining_class::AttachedAboveRight,
-    217,
-    combining_class::BelowLeft,
-    219,
-    combining_class::Below,
-    221,
-    combining_class::BelowRight,
-    223,
-    combining_class::Left,
-    225,
-    combining_class::Right,
-    227,
-    combining_class::AboveLeft,
-    229,
-    combining_class::Above,
-    231,
-    combining_class::AboveRight,
-    combining_class::DoubleBelow,
-    combining_class::DoubleAbove,
-    235, 236, 237, 238, 239,
-    combining_class::IotaSubscript,
-    241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254,
-    combining_class::Invalid,
-];
-
-pub trait CharExt {
-    fn script(self) -> Script;
-    fn general_category(self) -> GeneralCategory;
-    fn space_fallback(self) -> hb_unicode_funcs_t::space_t;
-    fn combining_class(self) -> u8;
-    fn modified_combining_class(self) -> u8;
-    fn mirrored(self) -> Option<Codepoint>;
-    fn is_emoji_extended_pictographic(self) -> bool;
-    fn is_default_ignorable(self) -> bool;
-    fn is_variation_selector(self) -> bool;
-    fn vertical(self) -> Option<Codepoint>;
-}
-
-impl CharExt for Codepoint {
-    fn script(self) -> Script {
-        _hb_ucd_sc_map[_hb_ucd_sc(self as usize) as usize]
-    }
-
-    fn general_category(self) -> GeneralCategory {
-        GeneralCategory(_hb_ucd_gc(self as usize))
-    }
-
-    fn space_fallback(self) -> hb_unicode_funcs_t::space_t {
-        use hb_unicode_funcs_t::*;
-
-        // All GC=Zs chars that can use a fallback.
-        match self {
-            0x0020 => SPACE,             // SPACE
-            0x00A0 => SPACE,             // NO-BREAK SPACE
-            0x2000 => SPACE_EM_2,        // EN QUAD
-            0x2001 => SPACE_EM,          // EM QUAD
-            0x2002 => SPACE_EM_2,        // EN SPACE
-            0x2003 => SPACE_EM,          // EM SPACE
-            0x2004 => SPACE_EM_3,        // THREE-PER-EM SPACE
-            0x2005 => SPACE_EM_4,        // FOUR-PER-EM SPACE
-            0x2006 => SPACE_EM_6,        // SIX-PER-EM SPACE
-            0x2007 => SPACE_FIGURE,      // FIGURE SPACE
-            0x2008 => SPACE_PUNCTUATION, // PUNCTUATION SPACE
-            0x2009 => SPACE_EM_5,        // THIN SPACE
-            0x200A => SPACE_EM_16,       // HAIR SPACE
-            0x202F => SPACE_NARROW,      // NARROW NO-BREAK SPACE
-            0x205F => SPACE_4_EM_18,     // MEDIUM MATHEMATICAL SPACE
-            0x3000 => SPACE_EM,          // IDEOGRAPHIC SPACE
-            _ => NOT_SPACE,              // OGHAM SPACE MARK
-        }
-    }
-
-    fn combining_class(self) -> u8 {
-        _hb_ucd_ccc(self as usize)
-    }
-
-    fn modified_combining_class(self) -> u8 {
-        let u = self;
-
-        // Reorder SAKOT to ensure it comes after any tone marks.
-        if u == 0x1A60 {
-            return 254;
-        }
-
-        // Reorder PADMA to ensure it comes after any vowel marks.
-        if u == 0x0FC6 {
-            return 254;
-        }
-
-        // Reorder TSA -PHRU to reorder before U+0F74
-        if u == 0x0F39 {
-            return 127;
-        }
-
-        let k = u.combining_class();
-
-        MODIFIED_COMBINING_CLASS[k as usize]
-    }
-
-    fn mirrored(self) -> Option<Codepoint> {
-        let delta = _hb_ucd_bmg(self as usize);
-        if delta == 0 {
-            None
-        } else {
-            Some(((self as i32).wrapping_add(delta as i32)) as u32)
-        }
-    }
-
-    fn is_emoji_extended_pictographic(self) -> bool {
-        super::unicode_emoji_table::is_Extended_Pictographic(self)
-    }
-
-    /// Default_Ignorable codepoints:
-    ///
-    /// Note: While U+115F, U+1160, U+3164 and U+FFA0 are Default_Ignorable,
-    /// we do NOT want to hide them, as the way Uniscribe has implemented them
-    /// is with regular spacing glyphs, and that's the way fonts are made to work.
-    /// As such, we make exceptions for those four.
-    /// Also ignoring U+1BCA0..1BCA3. https://github.com/harfbuzz/harfbuzz/issues/503
-    ///
-    /// Unicode 14.0:
-    /// $ grep '; Default_Ignorable_Code_Point ' DerivedCoreProperties.txt | sed 's/;.*#/#/'
-    /// 00AD          # Cf       SOFT HYPHEN
-    /// 034F          # Mn       COMBINING GRAPHEME JOINER
-    /// 061C          # Cf       ARABIC LETTER MARK
-    /// 115F..1160    # Lo   [2] HANGUL CHOSEONG FILLER..HANGUL JUNGSEONG FILLER
-    /// 17B4..17B5    # Mn   [2] KHMER VOWEL INHERENT AQ..KHMER VOWEL INHERENT AA
-    /// 180B..180D    # Mn   [3] MONGOLIAN FREE VARIATION SELECTOR ONE..MONGOLIAN FREE VARIATION SELECTOR THREE
-    /// 180E          # Cf       MONGOLIAN VOWEL SEPARATOR
-    /// 180F          # Mn       MONGOLIAN FREE VARIATION SELECTOR FOUR
-    /// 200B..200F    # Cf   [5] ZERO WIDTH SPACE..RIGHT-TO-LEFT MARK
-    /// 202A..202E    # Cf   [5] LEFT-TO-RIGHT EMBEDDING..RIGHT-TO-LEFT OVERRIDE
-    /// 2060..2064    # Cf   [5] WORD JOINER..INVISIBLE PLUS
-    /// 2065          # Cn       <reserved-2065>
-    /// 2066..206F    # Cf  [10] LEFT-TO-RIGHT ISOLATE..NOMINAL DIGIT SHAPES
-    /// 3164          # Lo       HANGUL FILLER
-    /// FE00..FE0F    # Mn  [16] VARIATION SELECTOR-1..VARIATION SELECTOR-16
-    /// FEFF          # Cf       ZERO WIDTH NO-BREAK SPACE
-    /// FFA0          # Lo       HALFWIDTH HANGUL FILLER
-    /// FFF0..FFF8    # Cn   [9] <reserved-FFF0>..<reserved-FFF8>
-    /// 1BCA0..1BCA3  # Cf   [4] SHORTHAND FORMAT LETTER OVERLAP..SHORTHAND FORMAT UP STEP
-    /// 1D173..1D17A  # Cf   [8] MUSICAL SYMBOL BEGIN BEAM..MUSICAL SYMBOL END PHRASE
-    /// E0000         # Cn       <reserved-E0000>
-    /// E0001         # Cf       LANGUAGE TAG
-    /// E0002..E001F  # Cn  [30] <reserved-E0002>..<reserved-E001F>
-    /// E0020..E007F  # Cf  [96] TAG SPACE..CANCEL TAG
-    /// E0080..E00FF  # Cn [128] <reserved-E0080>..<reserved-E00FF>
-    /// E0100..E01EF  # Mn [240] VARIATION SELECTOR-17..VARIATION SELECTOR-256
-    /// E01F0..E0FFF  # Cn [3600] <reserved-E01F0>..<reserved-E0FFF>
-    fn is_default_ignorable(self) -> bool {
-        let ch = self;
-        let plane = ch >> 16;
-        if plane == 0 {
-            // BMP
-            let page = ch >> 8;
-            match page {
-                0x00 => ch == 0x00AD,
-                0x03 => ch == 0x034F,
-                0x06 => ch == 0x061C,
-                0x17 => (0x17B4..=0x17B5).contains(&ch),
-                0x18 => (0x180B..=0x180E).contains(&ch),
-                0x20 => {
-                    (0x200B..=0x200F).contains(&ch)
-                        || (0x202A..=0x202E).contains(&ch)
-                        || (0x2060..=0x206F).contains(&ch)
-                }
-                0xFE => (0xFE00..=0xFE0F).contains(&ch) || ch == 0xFEFF,
-                0xFF => (0xFFF0..=0xFFF8).contains(&ch),
-                _ => false,
-            }
-        } else {
-            // Other planes
-            match plane {
-                0x01 => (0x1D173..=0x1D17A).contains(&ch),
-                0x0E => (0xE0000..=0xE0FFF).contains(&ch),
-                _ => false,
-            }
-        }
-    }
-
-    fn is_variation_selector(self) -> bool {
-        // U+180B..180D, U+180F MONGOLIAN FREE VARIATION SELECTORs are handled in the
-        //Arabic shaper. No need to match them here.
-        (0x0FE00..=0x0FE0F).contains(&self) || // VARIATION SELECTOR - 1..16
-        (0xE0100..=0xE01EF).contains(&self) // VARIATION SELECTOR - 17..256
-    }
-
-    fn vertical(self) -> Option<Codepoint> {
-        Some(match self >> 8 {
-            0x20 => match self {
-                0x2013 => 0xfe32, // EN DASH
-                0x2014 => 0xfe31, // EM DASH
-                0x2025 => 0xfe30, // TWO DOT LEADER
-                0x2026 => 0xfe19, // HORIZONTAL ELLIPSIS
-                _ => return None,
-            },
-            0x30 => match self {
-                0x3001 => 0xfe11, // IDEOGRAPHIC COMMA
-                0x3002 => 0xfe12, // IDEOGRAPHIC FULL STOP
-                0x3008 => 0xfe3f, // LEFT ANGLE BRACKET
-                0x3009 => 0xfe40, // RIGHT ANGLE BRACKET
-                0x300a => 0xfe3d, // LEFT DOUBLE ANGLE BRACKET
-                0x300b => 0xfe3e, // RIGHT DOUBLE ANGLE BRACKET
-                0x300c => 0xfe41, // LEFT CORNER BRACKET
-                0x300d => 0xfe42, // RIGHT CORNER BRACKET
-                0x300e => 0xfe43, // LEFT WHITE CORNER BRACKET
-                0x300f => 0xfe44, // RIGHT WHITE CORNER BRACKET
-                0x3010 => 0xfe3b, // LEFT BLACK LENTICULAR BRACKET
-                0x3011 => 0xfe3c, // RIGHT BLACK LENTICULAR BRACKET
-                0x3014 => 0xfe39, // LEFT TORTOISE SHELL BRACKET
-                0x3015 => 0xfe3a, // RIGHT TORTOISE SHELL BRACKET
-                0x3016 => 0xfe17, // LEFT WHITE LENTICULAR BRACKET
-                0x3017 => 0xfe18, // RIGHT WHITE LENTICULAR BRACKET
-                _ => return None,
-            },
-            0xfe => match self {
-                0xfe4f => 0xfe34, // WAVY LOW LINE
-                _ => return None,
-            },
-            0xff => match self {
-                0xff01 => 0xfe15, // FULLWIDTH EXCLAMATION MARK
-                0xff08 => 0xfe35, // FULLWIDTH LEFT PARENTHESIS
-                0xff09 => 0xfe36, // FULLWIDTH RIGHT PARENTHESIS
-                0xff0c => 0xfe10, // FULLWIDTH COMMA
-                0xff1a => 0xfe13, // FULLWIDTH COLON
-                0xff1b => 0xfe14, // FULLWIDTH SEMICOLON
-                0xff1f => 0xfe16, // FULLWIDTH QUESTION MARK
-                0xff3b => 0xfe47, // FULLWIDTH LEFT SQUARE BRACKET
-                0xff3d => 0xfe48, // FULLWIDTH RIGHT SQUARE BRACKET
-                0xff3f => 0xfe33, // FULLWIDTH LOW LINE
-                0xff5b => 0xfe37, // FULLWIDTH LEFT CURLY BRACKET
-                0xff5d => 0xfe38, // FULLWIDTH RIGHT CURLY BRACKET
-                _ => return None,
-            },
-            _ => return None,
-        })
-    }
-}
-
-const S_BASE: u32 = 0xAC00;
-const L_BASE: u32 = 0x1100;
-const V_BASE: u32 = 0x1161;
-const T_BASE: u32 = 0x11A7;
-const L_COUNT: u32 = 19;
-const V_COUNT: u32 = 21;
-const T_COUNT: u32 = 28;
-const N_COUNT: u32 = V_COUNT * T_COUNT;
-const S_COUNT: u32 = L_COUNT * N_COUNT;
-
-pub fn compose(a: Codepoint, b: Codepoint) -> Option<Codepoint> {
-    // Hangul is handled algorithmically.
-    if let Some(ab) = compose_hangul(a, b) {
-        return Some(ab);
-    }
-
-    let u: u32;
-
-    if (a & 0xFFFF_F800) == 0x0000 && (b & 0xFFFF_FF80) == 0x0300 {
-        /* If "a" is small enough and "b" is in the U+0300 range,
-         * the composition data is encoded in a 32bit array sorted
-         * by "a,b" pair. */
-        let k = HB_CODEPOINT_ENCODE3_11_7_14(a, b, 0);
-        let v = _hb_ucd_dm2_u32_map
-            .binary_search_by(|probe| {
-                let key = probe & HB_CODEPOINT_ENCODE3_11_7_14(0x001F_FFFF, 0x001F_FFFF, 0);
-                key.cmp(&k)
-            })
-            .ok()
-            .map(|index| _hb_ucd_dm2_u32_map[index]);
-
-        if let Some(value) = v {
-            u = HB_CODEPOINT_DECODE3_11_7_14_3(value);
-        } else {
-            return None;
-        }
-    } else {
-        /* Otherwise it is stored in a 64bit array sorted by
-         * "a,b" pair. */
-        let k = HB_CODEPOINT_ENCODE3(a, b, 0);
-        let v = _hb_ucd_dm2_u64_map
-            .binary_search_by(|probe| {
-                let key = probe & HB_CODEPOINT_ENCODE3(0x001F_FFFF, 0x001F_FFFF, 0);
-                key.cmp(&k)
-            })
-            .ok()
-            .map(|index| _hb_ucd_dm2_u64_map[index]);
-
-        if let Some(value) = v {
-            u = HB_CODEPOINT_DECODE3_3(value);
-        } else {
-            return None;
-        }
-    }
-
-    if u == 0 {
-        None
-    } else {
-        Some(u)
-    }
-}
-
-fn compose_hangul(a: Codepoint, b: Codepoint) -> Option<Codepoint> {
-    let l = a;
-    let v = b;
-    if L_BASE <= l && l < (L_BASE + L_COUNT) && V_BASE <= v && v < (V_BASE + V_COUNT) {
-        let r = S_BASE + (l - L_BASE) * N_COUNT + (v - V_BASE) * T_COUNT;
-        Some(r)
-    } else if S_BASE <= l
-        && l <= (S_BASE + S_COUNT - T_COUNT)
-        && T_BASE <= v
-        && v < (T_BASE + T_COUNT)
-        && (l - S_BASE) % T_COUNT == 0
-    {
-        let r = l + (v - T_BASE);
-        Some(r)
-    } else {
-        None
-    }
-}
-
-pub fn decompose(ab: Codepoint) -> Option<(Codepoint, Codepoint)> {
-    if let Some((a, b)) = decompose_hangul(ab) {
-        return Some((a, b));
-    }
-
-    let mut i = _hb_ucd_dm(ab as usize) as usize;
-
-    // If no data, there's no decomposition.
-    if i == 0 {
-        return None;
-    }
-    i -= 1;
-
-    if i < _hb_ucd_dm1_p0_map.len() + _hb_ucd_dm1_p2_map.len() {
-        let a = if i < _hb_ucd_dm1_p0_map.len() {
-            _hb_ucd_dm1_p0_map[i] as u32
-        } else {
-            let j = i - _hb_ucd_dm1_p0_map.len();
-            0x20000 | _hb_ucd_dm1_p2_map[j] as u32
-        };
-        return Some((a, 0));
-    }
-
-    i -= _hb_ucd_dm1_p0_map.len() + _hb_ucd_dm1_p2_map.len();
-
-    if i < _hb_ucd_dm2_u32_map.len() {
-        let v = _hb_ucd_dm2_u32_map[i];
-        let a = HB_CODEPOINT_DECODE3_11_7_14_1(v);
-        let b = HB_CODEPOINT_DECODE3_11_7_14_2(v);
-        return Some((a, b));
-    }
-
-    i -= _hb_ucd_dm2_u32_map.len();
-
-    let v = _hb_ucd_dm2_u64_map[i];
-    let a = HB_CODEPOINT_DECODE3_1(v);
-    let b = HB_CODEPOINT_DECODE3_2(v);
-    Some((a, b))
-}
-
-pub fn decompose_hangul(ab: Codepoint) -> Option<(Codepoint, Codepoint)> {
-    let si = ab.wrapping_sub(S_BASE);
-    if si >= S_COUNT {
-        return None;
-    }
-
-    let (a, b) = if si % T_COUNT != 0 {
-        // LV,T
-        (S_BASE + (si / T_COUNT) * T_COUNT, T_BASE + (si % T_COUNT))
-    } else {
-        // L,V
-        (L_BASE + (si / N_COUNT), V_BASE + (si % N_COUNT) / T_COUNT)
-    };
-    Some((a, b))
-}
-
-pub mod hb_gc {
-    pub const HB_UNICODE_GENERAL_CATEGORY_CONTROL: u32 = 0;
-    pub const HB_UNICODE_GENERAL_CATEGORY_FORMAT: u32 = 1;
-    pub const HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED: u32 = 2;
-    pub const HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE: u32 = 3;
-    pub const HB_UNICODE_GENERAL_CATEGORY_SURROGATE: u32 = 4;
-    pub const HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER: u32 = 5;
-    pub const HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER: u32 = 6;
-    pub const HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER: u32 = 7;
-    pub const HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER: u32 = 8;
-    pub const HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER: u32 = 9;
-    pub const HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK: u32 = 10;
-    pub const HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK: u32 = 11;
-    pub const HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK: u32 = 12;
-    pub const HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER: u32 = 13;
-    pub const HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER: u32 = 14;
-    pub const HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER: u32 = 15;
-    pub const HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION: u32 = 16;
-    pub const HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION: u32 = 17;
-    pub const HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION: u32 = 18;
-    pub const HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION: u32 = 19;
-    pub const HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION: u32 = 20;
-    pub const HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION: u32 = 21;
-    pub const HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION: u32 = 22;
-    pub const HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL: u32 = 23;
-    pub const HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL: u32 = 24;
-    pub const HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL: u32 = 25;
-    pub const HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL: u32 = 26;
-    pub const HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR: u32 = 27;
-    pub const HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR: u32 = 28;
-    pub const HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR: u32 = 29;
-}
diff --git a/harfrust/src/lib.rs b/harfrust/src/lib.rs
index ed748f38..2b5f8ddf 100644
--- a/harfrust/src/lib.rs
+++ b/harfrust/src/lib.rs
@@ -10,7 +10,9 @@ A complete [harfbuzz](https://github.com/harfbuzz/harfbuzz) shaping algorithm po
 
 extern crate alloc;
 
+mod algs;
 mod hb;
+mod unicode;
 
 #[cfg(feature = "std")]
 pub(crate) type U32Set = read_fonts::collections::int_set::U32Set;
diff --git a/harfrust/src/hb/unicode_emoji_table.rs b/harfrust/src/unicode/emoji_table.rs
similarity index 99%
rename from harfrust/src/hb/unicode_emoji_table.rs
rename to harfrust/src/unicode/emoji_table.rs
index 4e194bb1..2f25c3e9 100644
--- a/harfrust/src/hb/unicode_emoji_table.rs
+++ b/harfrust/src/unicode/emoji_table.rs
@@ -21,7 +21,7 @@
 #![allow(unused_parens)]
 #![allow(clippy::unnecessary_cast, clippy::unreadable_literal, clippy::double_parens)]
 
-use crate::hb::unicode::Codepoint;
+use crate::unicode::Codepoint;
 
 #[allow(dead_code, non_upper_case_globals, clippy::allow_attributes_without_reason, missing_docs)]
 pub(crate) static _hb_emoji_u8: [u8; 624]=
diff --git a/harfrust/src/unicode/mod.rs b/harfrust/src/unicode/mod.rs
new file mode 100644
index 00000000..da818e96
--- /dev/null
+++ b/harfrust/src/unicode/mod.rs
@@ -0,0 +1,1176 @@
+#![allow(
+    non_camel_case_types,
+    non_snake_case,
+    non_upper_case_globals,
+    clippy::inline_always,
+    clippy::wrong_self_convention,
+    clippy::enum_glob_use
+)]
+
+#[expect(clippy::identity_op)]
+#[rustfmt::skip]
+mod emoji_table;
+#[cfg(not(feature = "icu"))]
+#[rustfmt::skip]
+mod ucd_table;
+
+use crate::Script;
+
+pub type Codepoint = u32;
+
+// Space estimates based on:
+// https://unicode.org/charts/PDF/U2000.pdf
+// https://docs.microsoft.com/en-us/typography/develop/character-design-standards/whitespace
+pub mod hb_unicode_funcs_t {
+    pub type space_t = u8;
+    pub const NOT_SPACE: u8 = 0;
+    pub const SPACE_EM: u8 = 1;
+    pub const SPACE_EM_2: u8 = 2;
+    pub const SPACE_EM_3: u8 = 3;
+    pub const SPACE_EM_4: u8 = 4;
+    pub const SPACE_EM_5: u8 = 5;
+    pub const SPACE_EM_6: u8 = 6;
+    pub const SPACE_EM_16: u8 = 16;
+    pub const SPACE_4_EM_18: u8 = 17; // 4/18th of an EM!
+    pub const SPACE: u8 = 18;
+    pub const SPACE_FIGURE: u8 = 19;
+    pub const SPACE_PUNCTUATION: u8 = 20;
+    pub const SPACE_NARROW: u8 = 21;
+}
+
+/// Data type for the "General_Category" (gc) property from the Unicode
+/// Character Database.
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub struct GeneralCategory(pub u8);
+
+#[allow(unused)]
+impl GeneralCategory {
+    /// Control (`Cc`).
+    pub const CONTROL: Self = Self(0);
+    /// Format (`Cf`).
+    pub const FORMAT: Self = Self(1);
+    /// Unassigned (`Cn`).
+    pub const UNASSIGNED: Self = Self(2);
+    /// Private Use (`Co`).
+    pub const PRIVATE_USE: Self = Self(3);
+    /// Surrogate (`Cs`).
+    pub const SURROGATE: Self = Self(4);
+    /// Lowercase Letter (`Ll`).
+    pub const LOWERCASE_LETTER: Self = Self(5);
+    /// Modifier Letter (`Lm`).
+    pub const MODIFIER_LETTER: Self = Self(6);
+    /// Other Letter (`Lo`).
+    pub const OTHER_LETTER: Self = Self(7);
+    /// Titlecase Letter (`Lt`).
+    pub const TITLECASE_LETTER: Self = Self(8);
+    /// Uppercase Letter (`Lu`).
+    pub const UPPERCASE_LETTER: Self = Self(9);
+    /// Spacing Mark (`Mc`).
+    pub const SPACING_MARK: Self = Self(10);
+    /// Enclosing Mark (`Me`).
+    pub const ENCLOSING_MARK: Self = Self(11);
+    /// Nonspacing Mark (`Mn`).
+    pub const NON_SPACING_MARK: Self = Self(12);
+    /// Decimal Number (`Nd`).
+    pub const DECIMAL_NUMBER: Self = Self(13);
+    /// Letter Number (`Nl`).
+    pub const LETTER_NUMBER: Self = Self(14);
+    /// Other Number (`No`).
+    pub const OTHER_NUMBER: Self = Self(15);
+    /// Connector Punctuation (`Pc`).
+    pub const CONNECT_PUNCTUATION: Self = Self(16);
+    /// Dash Punctuation (`Pd`).
+    pub const DASH_PUNCTUATION: Self = Self(17);
+    /// Close Punctuation (`Pe`).
+    pub const CLOSE_PUNCTUATION: Self = Self(18);
+    /// Final Punctuation (`Pf`).
+    pub const FINAL_PUNCTUATION: Self = Self(19);
+    /// Initial Punctuation (`Pi`).
+    pub const INITIAL_PUNCTUATION: Self = Self(20);
+    /// Other Punctuation (`Po`).
+    pub const OTHER_PUNCTUATION: Self = Self(21);
+    /// Open Punctuation (`Ps`).
+    pub const OPEN_PUNCTUATION: Self = Self(22);
+    /// Currency Symbol (`Sc`).
+    pub const CURRENCY_SYMBOL: Self = Self(23);
+    /// Modifier Symbol (`Sk`).
+    pub const MODIFIER_SYMBOL: Self = Self(24);
+    /// Math Symbol (`Sm`).
+    pub const MATH_SYMBOL: Self = Self(25);
+    /// Other Symbol (`So`).
+    pub const OTHER_SYMBOL: Self = Self(26);
+    /// Line Separator (`Zl`).
+    pub const LINE_SEPARATOR: Self = Self(27);
+    /// Paragraph Separator (`Zp`).
+    pub const PARAGRAPH_SEPARATOR: Self = Self(28);
+    /// Space Separator (`Zs`).
+    pub const SPACE_SEPARATOR: Self = Self(29);
+}
+
+impl GeneralCategory {
+    pub fn is_mark(&self) -> bool {
+        matches!(
+            *self,
+            Self::SPACING_MARK | Self::ENCLOSING_MARK | Self::NON_SPACING_MARK
+        )
+    }
+
+    pub fn is_letter(&self) -> bool {
+        matches!(
+            *self,
+            Self::LOWERCASE_LETTER
+                | Self::MODIFIER_LETTER
+                | Self::OTHER_LETTER
+                | Self::TITLECASE_LETTER
+                | Self::UPPERCASE_LETTER
+        )
+    }
+
+    #[inline(always)]
+    pub(crate) const fn flag(self) -> u32 {
+        1 << self.0
+    }
+
+    #[inline(always)]
+    pub(crate) const fn flag_unsafe(self) -> u32 {
+        if self.0 < 32 {
+            1 << self.0
+        } else {
+            0
+        }
+    }
+
+    #[inline(always)]
+    pub(crate) const fn in_range_inclusive(self, start: Self, end: Self) -> bool {
+        debug_assert!(start.0 < end.0);
+        let range_mask = (1 << (end.0 + 1)) - (1 << start.0);
+        self.flag_unsafe() & range_mask != 0
+    }
+}
+
+#[allow(dead_code)]
+pub mod combining_class {
+    pub const NotReordered: u8 = 0;
+    pub const Overlay: u8 = 1;
+    pub const Nukta: u8 = 7;
+    pub const KanaVoicing: u8 = 8;
+    pub const Virama: u8 = 9;
+
+    /* Hebrew */
+    pub const CCC10: u8 = 10;
+    pub const CCC11: u8 = 11;
+    pub const CCC12: u8 = 12;
+    pub const CCC13: u8 = 13;
+    pub const CCC14: u8 = 14;
+    pub const CCC15: u8 = 15;
+    pub const CCC16: u8 = 16;
+    pub const CCC17: u8 = 17;
+    pub const CCC18: u8 = 18;
+    pub const CCC19: u8 = 19;
+    pub const CCC20: u8 = 20;
+    pub const CCC21: u8 = 21;
+    pub const CCC22: u8 = 22;
+    pub const CCC23: u8 = 23;
+    pub const CCC24: u8 = 24;
+    pub const CCC25: u8 = 25;
+    pub const CCC26: u8 = 26;
+
+    /* Arabic */
+    pub const CCC27: u8 = 27;
+    pub const CCC28: u8 = 28;
+    pub const CCC29: u8 = 29;
+    pub const CCC30: u8 = 30;
+    pub const CCC31: u8 = 31;
+    pub const CCC32: u8 = 32;
+    pub const CCC33: u8 = 33;
+    pub const CCC34: u8 = 34;
+    pub const CCC35: u8 = 35;
+
+    /* Syriac */
+    pub const CCC36: u8 = 36;
+
+    /* Telugu */
+    pub const CCC84: u8 = 84;
+    pub const CCC91: u8 = 91;
+
+    /* Thai */
+    pub const CCC103: u8 = 103;
+    pub const CCC107: u8 = 107;
+
+    /* Lao */
+    pub const CCC118: u8 = 118;
+    pub const CCC122: u8 = 122;
+
+    /* Tibetan */
+    pub const CCC129: u8 = 129;
+    pub const CCC130: u8 = 130;
+    pub const CCC132: u8 = 132;
+
+    pub const AttachedBelowLeft: u8 = 200;
+    pub const AttachedBelow: u8 = 202;
+    pub const AttachedAbove: u8 = 214;
+    pub const AttachedAboveRight: u8 = 216;
+    pub const BelowLeft: u8 = 218;
+    pub const Below: u8 = 220;
+    pub const BelowRight: u8 = 222;
+    pub const Left: u8 = 224;
+    pub const Right: u8 = 226;
+    pub const AboveLeft: u8 = 228;
+    pub const Above: u8 = 230;
+    pub const AboveRight: u8 = 232;
+    pub const DoubleBelow: u8 = 233;
+    pub const DoubleAbove: u8 = 234;
+
+    pub const IotaSubscript: u8 = 240;
+
+    pub const Invalid: u8 = 255;
+}
+
+#[allow(dead_code)]
+pub mod modified_combining_class {
+    // Hebrew
+    //
+    // We permute the "fixed-position" classes 10-26 into the order
+    // described in the SBL Hebrew manual:
+    //
+    // https://www.sbl-site.org/Fonts/SBLHebrewUserManual1.5x.pdf
+    //
+    // (as recommended by:
+    //  https://forum.fontlab.com/archive-old-microsoft-volt-group/vista-and-diacritic-ordering/msg22823/)
+    //
+    // More details here:
+    // https://bugzilla.mozilla.org/show_bug.cgi?id=662055
+    pub const CCC10: u8 = 22; // sheva
+    pub const CCC11: u8 = 15; // hataf segol
+    pub const CCC12: u8 = 16; // hataf patah
+    pub const CCC13: u8 = 17; // hataf qamats
+    pub const CCC14: u8 = 23; // hiriq
+    pub const CCC15: u8 = 18; // tsere
+    pub const CCC16: u8 = 19; // segol
+    pub const CCC17: u8 = 20; // patah
+    pub const CCC18: u8 = 21; // qamats & qamats qatan
+    pub const CCC19: u8 = 14; // holam & holam haser for vav
+    pub const CCC20: u8 = 24; // qubuts
+    pub const CCC21: u8 = 12; // dagesh
+    pub const CCC22: u8 = 25; // meteg
+    pub const CCC23: u8 = 13; // rafe
+    pub const CCC24: u8 = 10; // shin dot
+    pub const CCC25: u8 = 11; // sin dot
+    pub const CCC26: u8 = 26; // point varika
+
+    // Arabic
+    //
+    // Modify to move Shadda (ccc=33) before other marks.  See:
+    // https://unicode.org/faq/normalization.html#8
+    // https://unicode.org/faq/normalization.html#9
+    pub const CCC27: u8 = 28; // fathatan
+    pub const CCC28: u8 = 29; // dammatan
+    pub const CCC29: u8 = 30; // kasratan
+    pub const CCC30: u8 = 31; // fatha
+    pub const CCC31: u8 = 32; // damma
+    pub const CCC32: u8 = 33; // kasra
+    pub const CCC33: u8 = 27; // shadda
+    pub const CCC34: u8 = 34; // sukun
+    pub const CCC35: u8 = 35; // superscript alef
+
+    // Syriac
+    pub const CCC36: u8 = 36; // superscript alaph
+
+    // Telugu
+    //
+    // Modify Telugu length marks (ccc=84, ccc=91).
+    // These are the only matras in the main Indic scripts range that have
+    // a non-zero ccc.  That makes them reorder with the Halant that is
+    // ccc=9.  Just zero them, we don't need them in our Indic shaper.
+    pub const CCC84: u8 = 0; // length mark
+    pub const CCC91: u8 = 0; // ai length mark
+
+    // Thai
+    //
+    // Modify U+0E38 and U+0E39 (ccc=103) to be reordered before U+0E3A (ccc=9).
+    // Assign 3, which is unassigned otherwise.
+    // Uniscribe does this reordering too.
+    pub const CCC103: u8 = 3; // sara u / sara uu
+    pub const CCC107: u8 = 107; // mai *
+
+    // Lao
+    pub const CCC118: u8 = 118; // sign u / sign uu
+    pub const CCC122: u8 = 122; // mai *
+
+    // Tibetan
+    //
+    // In case of multiple vowel-signs, use u first (but after achung)
+    // this allows Dzongkha multi-vowel shortcuts to render correctly
+    pub const CCC129: u8 = 129; // sign aa
+    pub const CCC130: u8 = 132; // sign i
+    pub const CCC132: u8 = 131; // sign u
+}
+
+#[rustfmt::skip]
+static MODIFIED_COMBINING_CLASS: &[u8; 256] = &[
+    combining_class::NotReordered,
+    combining_class::Overlay,
+    2, 3, 4, 5, 6,
+    combining_class::Nukta,
+    combining_class::KanaVoicing,
+    combining_class::Virama,
+
+    // Hebrew
+    modified_combining_class::CCC10,
+    modified_combining_class::CCC11,
+    modified_combining_class::CCC12,
+    modified_combining_class::CCC13,
+    modified_combining_class::CCC14,
+    modified_combining_class::CCC15,
+    modified_combining_class::CCC16,
+    modified_combining_class::CCC17,
+    modified_combining_class::CCC18,
+    modified_combining_class::CCC19,
+    modified_combining_class::CCC20,
+    modified_combining_class::CCC21,
+    modified_combining_class::CCC22,
+    modified_combining_class::CCC23,
+    modified_combining_class::CCC24,
+    modified_combining_class::CCC25,
+    modified_combining_class::CCC26,
+
+    // Arabic
+    modified_combining_class::CCC27,
+    modified_combining_class::CCC28,
+    modified_combining_class::CCC29,
+    modified_combining_class::CCC30,
+    modified_combining_class::CCC31,
+    modified_combining_class::CCC32,
+    modified_combining_class::CCC33,
+    modified_combining_class::CCC34,
+    modified_combining_class::CCC35,
+
+    // Syriac
+    modified_combining_class::CCC36,
+
+    37, 38, 39,
+    40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59,
+    60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79,
+    80, 81, 82, 83,
+
+    // Telugu
+    modified_combining_class::CCC84,
+    85, 86, 87, 88, 89, 90,
+    modified_combining_class::CCC91,
+    92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102,
+
+    // Thai
+    modified_combining_class::CCC103,
+    104, 105, 106,
+    modified_combining_class::CCC107,
+    108, 109, 110, 111, 112, 113, 114, 115, 116, 117,
+
+    // Lao
+    modified_combining_class::CCC118,
+    119, 120, 121,
+    modified_combining_class::CCC122,
+    123, 124, 125, 126, 127, 128,
+
+    // Tibetan
+    modified_combining_class::CCC129,
+    modified_combining_class::CCC130,
+    131,
+    modified_combining_class::CCC132,
+    133, 134, 135, 136, 137, 138, 139,
+
+
+    140, 141, 142, 143, 144, 145, 146, 147, 148, 149,
+    150, 151, 152, 153, 154, 155, 156, 157, 158, 159,
+    160, 161, 162, 163, 164, 165, 166, 167, 168, 169,
+    170, 171, 172, 173, 174, 175, 176, 177, 178, 179,
+    180, 181, 182, 183, 184, 185, 186, 187, 188, 189,
+    190, 191, 192, 193, 194, 195, 196, 197, 198, 199,
+
+    combining_class::AttachedBelowLeft,
+    201,
+    combining_class::AttachedBelow,
+    203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213,
+    combining_class::AttachedAbove,
+    215,
+    combining_class::AttachedAboveRight,
+    217,
+    combining_class::BelowLeft,
+    219,
+    combining_class::Below,
+    221,
+    combining_class::BelowRight,
+    223,
+    combining_class::Left,
+    225,
+    combining_class::Right,
+    227,
+    combining_class::AboveLeft,
+    229,
+    combining_class::Above,
+    231,
+    combining_class::AboveRight,
+    combining_class::DoubleBelow,
+    combining_class::DoubleAbove,
+    235, 236, 237, 238, 239,
+    combining_class::IotaSubscript,
+    241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254,
+    combining_class::Invalid,
+];
+
+pub trait CharExt {
+    fn script(self) -> Script;
+    fn general_category(self) -> GeneralCategory;
+    fn space_fallback(self) -> hb_unicode_funcs_t::space_t;
+    fn combining_class(self) -> u8;
+    fn modified_combining_class(self) -> u8;
+    fn mirroring(self) -> Option<Codepoint>;
+    fn is_emoji_extended_pictographic(self) -> bool;
+    fn is_default_ignorable(self) -> bool;
+    fn is_variation_selector(self) -> bool;
+    fn vertical(self) -> Option<Codepoint>;
+}
+
+impl CharExt for Codepoint {
+    fn script(self) -> Script {
+        script_for(self)
+    }
+
+    fn general_category(self) -> GeneralCategory {
+        general_category_for(self)
+    }
+
+    fn combining_class(self) -> u8 {
+        combining_class_for(self)
+    }
+
+    fn mirroring(self) -> Option<Codepoint> {
+        mirroring_for(self)
+    }
+
+    fn space_fallback(self) -> hb_unicode_funcs_t::space_t {
+        use hb_unicode_funcs_t::*;
+
+        // All GC=Zs chars that can use a fallback.
+        match self {
+            0x0020 => SPACE,             // SPACE
+            0x00A0 => SPACE,             // NO-BREAK SPACE
+            0x2000 => SPACE_EM_2,        // EN QUAD
+            0x2001 => SPACE_EM,          // EM QUAD
+            0x2002 => SPACE_EM_2,        // EN SPACE
+            0x2003 => SPACE_EM,          // EM SPACE
+            0x2004 => SPACE_EM_3,        // THREE-PER-EM SPACE
+            0x2005 => SPACE_EM_4,        // FOUR-PER-EM SPACE
+            0x2006 => SPACE_EM_6,        // SIX-PER-EM SPACE
+            0x2007 => SPACE_FIGURE,      // FIGURE SPACE
+            0x2008 => SPACE_PUNCTUATION, // PUNCTUATION SPACE
+            0x2009 => SPACE_EM_5,        // THIN SPACE
+            0x200A => SPACE_EM_16,       // HAIR SPACE
+            0x202F => SPACE_NARROW,      // NARROW NO-BREAK SPACE
+            0x205F => SPACE_4_EM_18,     // MEDIUM MATHEMATICAL SPACE
+            0x3000 => SPACE_EM,          // IDEOGRAPHIC SPACE
+            _ => NOT_SPACE,              // OGHAM SPACE MARK
+        }
+    }
+
+    fn modified_combining_class(self) -> u8 {
+        let u = self;
+
+        // Reorder SAKOT to ensure it comes after any tone marks.
+        if u == 0x1A60 {
+            return 254;
+        }
+
+        // Reorder PADMA to ensure it comes after any vowel marks.
+        if u == 0x0FC6 {
+            return 254;
+        }
+
+        // Reorder TSA -PHRU to reorder before U+0F74
+        if u == 0x0F39 {
+            return 127;
+        }
+
+        let k = u.combining_class();
+
+        MODIFIED_COMBINING_CLASS[k as usize]
+    }
+
+    fn is_emoji_extended_pictographic(self) -> bool {
+        emoji_table::is_Extended_Pictographic(self)
+    }
+
+    /// Default_Ignorable codepoints:
+    ///
+    /// Note: While U+115F, U+1160, U+3164 and U+FFA0 are Default_Ignorable,
+    /// we do NOT want to hide them, as the way Uniscribe has implemented them
+    /// is with regular spacing glyphs, and that's the way fonts are made to work.
+    /// As such, we make exceptions for those four.
+    /// Also ignoring U+1BCA0..1BCA3. https://github.com/harfbuzz/harfbuzz/issues/503
+    ///
+    /// Unicode 14.0:
+    /// $ grep '; Default_Ignorable_Code_Point ' DerivedCoreProperties.txt | sed 's/;.*#/#/'
+    /// 00AD          # Cf       SOFT HYPHEN
+    /// 034F          # Mn       COMBINING GRAPHEME JOINER
+    /// 061C          # Cf       ARABIC LETTER MARK
+    /// 115F..1160    # Lo   [2] HANGUL CHOSEONG FILLER..HANGUL JUNGSEONG FILLER
+    /// 17B4..17B5    # Mn   [2] KHMER VOWEL INHERENT AQ..KHMER VOWEL INHERENT AA
+    /// 180B..180D    # Mn   [3] MONGOLIAN FREE VARIATION SELECTOR ONE..MONGOLIAN FREE VARIATION SELECTOR THREE
+    /// 180E          # Cf       MONGOLIAN VOWEL SEPARATOR
+    /// 180F          # Mn       MONGOLIAN FREE VARIATION SELECTOR FOUR
+    /// 200B..200F    # Cf   [5] ZERO WIDTH SPACE..RIGHT-TO-LEFT MARK
+    /// 202A..202E    # Cf   [5] LEFT-TO-RIGHT EMBEDDING..RIGHT-TO-LEFT OVERRIDE
+    /// 2060..2064    # Cf   [5] WORD JOINER..INVISIBLE PLUS
+    /// 2065          # Cn       <reserved-2065>
+    /// 2066..206F    # Cf  [10] LEFT-TO-RIGHT ISOLATE..NOMINAL DIGIT SHAPES
+    /// 3164          # Lo       HANGUL FILLER
+    /// FE00..FE0F    # Mn  [16] VARIATION SELECTOR-1..VARIATION SELECTOR-16
+    /// FEFF          # Cf       ZERO WIDTH NO-BREAK SPACE
+    /// FFA0          # Lo       HALFWIDTH HANGUL FILLER
+    /// FFF0..FFF8    # Cn   [9] <reserved-FFF0>..<reserved-FFF8>
+    /// 1BCA0..1BCA3  # Cf   [4] SHORTHAND FORMAT LETTER OVERLAP..SHORTHAND FORMAT UP STEP
+    /// 1D173..1D17A  # Cf   [8] MUSICAL SYMBOL BEGIN BEAM..MUSICAL SYMBOL END PHRASE
+    /// E0000         # Cn       <reserved-E0000>
+    /// E0001         # Cf       LANGUAGE TAG
+    /// E0002..E001F  # Cn  [30] <reserved-E0002>..<reserved-E001F>
+    /// E0020..E007F  # Cf  [96] TAG SPACE..CANCEL TAG
+    /// E0080..E00FF  # Cn [128] <reserved-E0080>..<reserved-E00FF>
+    /// E0100..E01EF  # Mn [240] VARIATION SELECTOR-17..VARIATION SELECTOR-256
+    /// E01F0..E0FFF  # Cn [3600] <reserved-E01F0>..<reserved-E0FFF>
+    fn is_default_ignorable(self) -> bool {
+        let ch = self;
+        let plane = ch >> 16;
+        if plane == 0 {
+            // BMP
+            let page = ch >> 8;
+            match page {
+                0x00 => ch == 0x00AD,
+                0x03 => ch == 0x034F,
+                0x06 => ch == 0x061C,
+                0x17 => (0x17B4..=0x17B5).contains(&ch),
+                0x18 => (0x180B..=0x180E).contains(&ch),
+                0x20 => {
+                    (0x200B..=0x200F).contains(&ch)
+                        || (0x202A..=0x202E).contains(&ch)
+                        || (0x2060..=0x206F).contains(&ch)
+                }
+                0xFE => (0xFE00..=0xFE0F).contains(&ch) || ch == 0xFEFF,
+                0xFF => (0xFFF0..=0xFFF8).contains(&ch),
+                _ => false,
+            }
+        } else {
+            // Other planes
+            match plane {
+                0x01 => (0x1D173..=0x1D17A).contains(&ch),
+                0x0E => (0xE0000..=0xE0FFF).contains(&ch),
+                _ => false,
+            }
+        }
+    }
+
+    fn is_variation_selector(self) -> bool {
+        // U+180B..180D, U+180F MONGOLIAN FREE VARIATION SELECTORs are handled in the
+        //Arabic shaper. No need to match them here.
+        (0x0FE00..=0x0FE0F).contains(&self) || // VARIATION SELECTOR - 1..16
+        (0xE0100..=0xE01EF).contains(&self) // VARIATION SELECTOR - 17..256
+    }
+
+    fn vertical(self) -> Option<Codepoint> {
+        Some(match self >> 8 {
+            0x20 => match self {
+                0x2013 => 0xfe32, // EN DASH
+                0x2014 => 0xfe31, // EM DASH
+                0x2025 => 0xfe30, // TWO DOT LEADER
+                0x2026 => 0xfe19, // HORIZONTAL ELLIPSIS
+                _ => return None,
+            },
+            0x30 => match self {
+                0x3001 => 0xfe11, // IDEOGRAPHIC COMMA
+                0x3002 => 0xfe12, // IDEOGRAPHIC FULL STOP
+                0x3008 => 0xfe3f, // LEFT ANGLE BRACKET
+                0x3009 => 0xfe40, // RIGHT ANGLE BRACKET
+                0x300a => 0xfe3d, // LEFT DOUBLE ANGLE BRACKET
+                0x300b => 0xfe3e, // RIGHT DOUBLE ANGLE BRACKET
+                0x300c => 0xfe41, // LEFT CORNER BRACKET
+                0x300d => 0xfe42, // RIGHT CORNER BRACKET
+                0x300e => 0xfe43, // LEFT WHITE CORNER BRACKET
+                0x300f => 0xfe44, // RIGHT WHITE CORNER BRACKET
+                0x3010 => 0xfe3b, // LEFT BLACK LENTICULAR BRACKET
+                0x3011 => 0xfe3c, // RIGHT BLACK LENTICULAR BRACKET
+                0x3014 => 0xfe39, // LEFT TORTOISE SHELL BRACKET
+                0x3015 => 0xfe3a, // RIGHT TORTOISE SHELL BRACKET
+                0x3016 => 0xfe17, // LEFT WHITE LENTICULAR BRACKET
+                0x3017 => 0xfe18, // RIGHT WHITE LENTICULAR BRACKET
+                _ => return None,
+            },
+            0xfe => match self {
+                0xfe4f => 0xfe34, // WAVY LOW LINE
+                _ => return None,
+            },
+            0xff => match self {
+                0xff01 => 0xfe15, // FULLWIDTH EXCLAMATION MARK
+                0xff08 => 0xfe35, // FULLWIDTH LEFT PARENTHESIS
+                0xff09 => 0xfe36, // FULLWIDTH RIGHT PARENTHESIS
+                0xff0c => 0xfe10, // FULLWIDTH COMMA
+                0xff1a => 0xfe13, // FULLWIDTH COLON
+                0xff1b => 0xfe14, // FULLWIDTH SEMICOLON
+                0xff1f => 0xfe16, // FULLWIDTH QUESTION MARK
+                0xff3b => 0xfe47, // FULLWIDTH LEFT SQUARE BRACKET
+                0xff3d => 0xfe48, // FULLWIDTH RIGHT SQUARE BRACKET
+                0xff3f => 0xfe33, // FULLWIDTH LOW LINE
+                0xff5b => 0xfe37, // FULLWIDTH LEFT CURLY BRACKET
+                0xff5d => 0xfe38, // FULLWIDTH RIGHT CURLY BRACKET
+                _ => return None,
+            },
+            _ => return None,
+        })
+    }
+}
+
+#[cfg(feature = "icu")]
+pub(crate) use icu::*;
+
+#[cfg(not(feature = "icu"))]
+pub(crate) use builtin::*;
+
+#[cfg(feature = "icu")]
+mod icu {
+    use super::{Codepoint, GeneralCategory, Script};
+    use icu_normalizer::properties::{
+        CanonicalCompositionBorrowed, CanonicalDecompositionBorrowed, Decomposed,
+    };
+    use icu_properties::{
+        props::{
+            BidiMirroringGlyph, CanonicalCombiningClass as IcuCanonicalCombiningClass,
+            GeneralCategory as IcuGeneralCategory, Script as IcuScript,
+        },
+        CodePointMapData,
+    };
+
+    pub(crate) fn script_for(c: u32) -> Script {
+        let icu_script = CodePointMapData::<IcuScript>::new().get32(c);
+        SCRIPT_MAP
+            .get(icu_script.to_icu4c_value() as usize)
+            .copied()
+            .unwrap_or(crate::script::UNKNOWN)
+    }
+
+    pub(crate) fn general_category_for(c: u32) -> GeneralCategory {
+        GeneralCategory::from_icu(CodePointMapData::<IcuGeneralCategory>::new().get32(c))
+    }
+
+    pub(crate) fn combining_class_for(c: u32) -> u8 {
+        CodePointMapData::<IcuCanonicalCombiningClass>::new()
+            .get32(c)
+            .to_icu4c_value()
+    }
+
+    pub(crate) fn mirroring_for(c: u32) -> Option<Codepoint> {
+        CodePointMapData::<BidiMirroringGlyph>::new()
+            .get32(c)
+            .mirroring_glyph
+            .map(Codepoint::from)
+    }
+
+    pub(crate) fn compose(a: Codepoint, b: Codepoint) -> Option<Codepoint> {
+        let a = char::from_u32(a)?;
+        let b = char::from_u32(b)?;
+        CanonicalCompositionBorrowed::new()
+            .compose(a, b)
+            .map(Codepoint::from)
+    }
+
+    pub(crate) fn decompose(ab: Codepoint) -> Option<(Codepoint, Codepoint)> {
+        let ch = char::from_u32(ab)?;
+        match CanonicalDecompositionBorrowed::new().decompose(ch) {
+            Decomposed::Default => None,
+            Decomposed::Singleton(a) => Some((Codepoint::from(a), 0)),
+            Decomposed::Expansion(a, b) => Some((Codepoint::from(a), Codepoint::from(b))),
+        }
+    }
+
+    impl GeneralCategory {
+        fn from_icu(category: IcuGeneralCategory) -> Self {
+            use IcuGeneralCategory::*;
+            match category {
+                Control => Self::CONTROL,
+                Format => Self::FORMAT,
+                Unassigned => Self::UNASSIGNED,
+                PrivateUse => Self::PRIVATE_USE,
+                Surrogate => Self::SURROGATE,
+                LowercaseLetter => Self::LOWERCASE_LETTER,
+                ModifierLetter => Self::MODIFIER_LETTER,
+                OtherLetter => Self::OTHER_LETTER,
+                TitlecaseLetter => Self::TITLECASE_LETTER,
+                UppercaseLetter => Self::UPPERCASE_LETTER,
+                SpacingMark => Self::SPACING_MARK,
+                EnclosingMark => Self::ENCLOSING_MARK,
+                NonspacingMark => Self::NON_SPACING_MARK,
+                DecimalNumber => Self::DECIMAL_NUMBER,
+                LetterNumber => Self::LETTER_NUMBER,
+                OtherNumber => Self::OTHER_NUMBER,
+                ConnectorPunctuation => Self::CONNECT_PUNCTUATION,
+                DashPunctuation => Self::DASH_PUNCTUATION,
+                ClosePunctuation => Self::CLOSE_PUNCTUATION,
+                FinalPunctuation => Self::FINAL_PUNCTUATION,
+                InitialPunctuation => Self::INITIAL_PUNCTUATION,
+                OtherPunctuation => Self::OTHER_PUNCTUATION,
+                OpenPunctuation => Self::OPEN_PUNCTUATION,
+                CurrencySymbol => Self::CURRENCY_SYMBOL,
+                ModifierSymbol => Self::MODIFIER_SYMBOL,
+                MathSymbol => Self::MATH_SYMBOL,
+                OtherSymbol => Self::OTHER_SYMBOL,
+                LineSeparator => Self::LINE_SEPARATOR,
+                ParagraphSeparator => Self::PARAGRAPH_SEPARATOR,
+                SpaceSeparator => Self::SPACE_SEPARATOR,
+            }
+        }
+    }
+
+    pub(crate) static SCRIPT_MAP: [Script; 212] = [
+        crate::script::COMMON,                 // 0
+        crate::script::INHERITED,              // 1
+        crate::script::ARABIC,                 // 2
+        crate::script::ARMENIAN,               // 3
+        crate::script::BENGALI,                // 4
+        crate::script::BOPOMOFO,               // 5
+        crate::script::CHEROKEE,               // 6
+        crate::script::COPTIC,                 // 7
+        crate::script::CYRILLIC,               // 8
+        crate::script::DESERET,                // 9
+        crate::script::DEVANAGARI,             // 10
+        crate::script::ETHIOPIC,               // 11
+        crate::script::GEORGIAN,               // 12
+        crate::script::GOTHIC,                 // 13
+        crate::script::GREEK,                  // 14
+        crate::script::GUJARATI,               // 15
+        crate::script::GURMUKHI,               // 16
+        crate::script::HAN,                    // 17
+        crate::script::HANGUL,                 // 18
+        crate::script::HEBREW,                 // 19
+        crate::script::HIRAGANA,               // 20
+        crate::script::KANNADA,                // 21
+        crate::script::KATAKANA,               // 22
+        crate::script::KHMER,                  // 23
+        crate::script::LAO,                    // 24
+        crate::script::LATIN,                  // 25
+        crate::script::MALAYALAM,              // 26
+        crate::script::MONGOLIAN,              // 27
+        crate::script::MYANMAR,                // 28
+        crate::script::OGHAM,                  // 29
+        crate::script::OLD_ITALIC,             // 30
+        crate::script::ORIYA,                  // 31
+        crate::script::RUNIC,                  // 32
+        crate::script::SINHALA,                // 33
+        crate::script::SYRIAC,                 // 34
+        crate::script::TAMIL,                  // 35
+        crate::script::TELUGU,                 // 36
+        crate::script::THAANA,                 // 37
+        crate::script::THAI,                   // 38
+        crate::script::TIBETAN,                // 39
+        crate::script::CANADIAN_SYLLABICS,     // 40
+        crate::script::YI,                     // 41
+        crate::script::TAGALOG,                // 42
+        crate::script::HANUNOO,                // 43
+        crate::script::BUHID,                  // 44
+        crate::script::TAGBANWA,               // 45
+        crate::script::BRAILLE,                // 46
+        crate::script::CYPRIOT,                // 47
+        crate::script::LIMBU,                  // 48
+        crate::script::LINEAR_B,               // 49
+        crate::script::OSMANYA,                // 50
+        crate::script::SHAVIAN,                // 51
+        crate::script::TAI_LE,                 // 52
+        crate::script::UGARITIC,               // 53
+        crate::script::UNKNOWN,                // 54
+        crate::script::BUGINESE,               // 55
+        crate::script::GLAGOLITIC,             // 56
+        crate::script::KHAROSHTHI,             // 57
+        crate::script::SYLOTI_NAGRI,           // 58
+        crate::script::NEW_TAI_LUE,            // 59
+        crate::script::TIFINAGH,               // 60
+        crate::script::OLD_PERSIAN,            // 61
+        crate::script::BALINESE,               // 62
+        crate::script::BATAK,                  // 63
+        crate::script::UNKNOWN,                // 64
+        crate::script::BRAHMI,                 // 65
+        crate::script::CHAM,                   // 66
+        crate::script::UNKNOWN,                // 67
+        crate::script::CYRILLIC,               // 68 (Cyrs)
+        crate::script::UNKNOWN,                // 69
+        crate::script::UNKNOWN,                // 70
+        crate::script::EGYPTIAN_HIEROGLYPHS,   // 71
+        crate::script::GEORGIAN,               // 72 (Geok)
+        crate::script::HAN,                    // 73 (Hans)
+        crate::script::HAN,                    // 74 (Hant)
+        crate::script::PAHAWH_HMONG,           // 75
+        crate::script::OLD_HUNGARIAN,          // 76
+        crate::script::UNKNOWN,                // 77
+        crate::script::JAVANESE,               // 78
+        crate::script::KAYAH_LI,               // 79
+        crate::script::LATIN,                  // 80 (Latf)
+        crate::script::LATIN,                  // 81 (Latg)
+        crate::script::LEPCHA,                 // 82
+        crate::script::LINEAR_A,               // 83
+        crate::script::MANDAIC,                // 84
+        crate::script::UNKNOWN,                // 85
+        crate::script::MEROITIC_HIEROGLYPHS,   // 86
+        crate::script::NKO,                    // 87
+        crate::script::OLD_TURKIC,             // 88
+        crate::script::OLD_PERMIC,             // 89
+        crate::script::PHAGS_PA,               // 90
+        crate::script::PHOENICIAN,             // 91
+        crate::script::MIAO,                   // 92
+        crate::script::UNKNOWN,                // 93
+        crate::script::UNKNOWN,                // 94
+        crate::script::SYRIAC,                 // 95 (Syre)
+        crate::script::SYRIAC,                 // 96 (Syrj)
+        crate::script::SYRIAC,                 // 97 (Syrn)
+        crate::script::UNKNOWN,                // 98
+        crate::script::VAI,                    // 99
+        crate::script::UNKNOWN,                // 100
+        crate::script::CUNEIFORM,              // 101
+        crate::script::UNKNOWN,                // 102
+        crate::script::UNKNOWN,                // 103
+        crate::script::CARIAN,                 // 104
+        crate::script::UNKNOWN,                // 105
+        crate::script::TAI_THAM,               // 106
+        crate::script::LYCIAN,                 // 107
+        crate::script::LYDIAN,                 // 108
+        crate::script::OL_CHIKI,               // 109
+        crate::script::REJANG,                 // 110
+        crate::script::SAURASHTRA,             // 111
+        crate::script::SIGNWRITING,            // 112
+        crate::script::SUNDANESE,              // 113
+        crate::script::UNKNOWN,                // 114
+        crate::script::MEETEI_MAYEK,           // 115
+        crate::script::IMPERIAL_ARAMAIC,       // 116
+        crate::script::AVESTAN,                // 117
+        crate::script::CHAKMA,                 // 118
+        crate::script::UNKNOWN,                // 119
+        crate::script::KAITHI,                 // 120
+        crate::script::MANICHAEAN,             // 121
+        crate::script::INSCRIPTIONAL_PAHLAVI,  // 122
+        crate::script::PSALTER_PAHLAVI,        // 123
+        crate::script::UNKNOWN,                // 124
+        crate::script::INSCRIPTIONAL_PARTHIAN, // 125
+        crate::script::SAMARITAN,              // 126
+        crate::script::TAI_VIET,               // 127
+        crate::script::UNKNOWN,                // 128
+        crate::script::UNKNOWN,                // 129
+        crate::script::BAMUM,                  // 130
+        crate::script::LISU,                   // 131
+        crate::script::UNKNOWN,                // 132
+        crate::script::OLD_SOUTH_ARABIAN,      // 133
+        crate::script::BASSA_VAH,              // 134
+        crate::script::DUPLOYAN,               // 135
+        crate::script::ELBASAN,                // 136
+        crate::script::GRANTHA,                // 137
+        crate::script::UNKNOWN,                // 138
+        crate::script::UNKNOWN,                // 139
+        crate::script::MENDE_KIKAKUI,          // 140
+        crate::script::MEROITIC_CURSIVE,       // 141
+        crate::script::OLD_NORTH_ARABIAN,      // 142
+        crate::script::NABATAEAN,              // 143
+        crate::script::PALMYRENE,              // 144
+        crate::script::KHUDAWADI,              // 145
+        crate::script::WARANG_CITI,            // 146
+        crate::script::UNKNOWN,                // 147
+        crate::script::UNKNOWN,                // 148
+        crate::script::MRO,                    // 149
+        crate::script::NUSHU,                  // 150
+        crate::script::SHARADA,                // 151
+        crate::script::SORA_SOMPENG,           // 152
+        crate::script::TAKRI,                  // 153
+        crate::script::TANGUT,                 // 154
+        crate::script::UNKNOWN,                // 155
+        crate::script::ANATOLIAN_HIEROGLYPHS,  // 156
+        crate::script::KHOJKI,                 // 157
+        crate::script::TIRHUTA,                // 158
+        crate::script::CAUCASIAN_ALBANIAN,     // 159
+        crate::script::MAHAJANI,               // 160
+        crate::script::AHOM,                   // 161
+        crate::script::HATRAN,                 // 162
+        crate::script::MODI,                   // 163
+        crate::script::MULTANI,                // 164
+        crate::script::PAU_CIN_HAU,            // 165
+        crate::script::SIDDHAM,                // 166
+        crate::script::ADLAM,                  // 167
+        crate::script::BHAIKSUKI,              // 168
+        crate::script::MARCHEN,                // 169
+        crate::script::NEWA,                   // 170
+        crate::script::OSAGE,                  // 171
+        crate::script::UNKNOWN,                // 172
+        crate::script::HANGUL,                 // 173 (Jamo)
+        crate::script::UNKNOWN,                // 174
+        crate::script::MASARAM_GONDI,          // 175
+        crate::script::SOYOMBO,                // 176
+        crate::script::ZANABAZAR_SQUARE,       // 177
+        crate::script::DOGRA,                  // 178
+        crate::script::GUNJALA_GONDI,          // 179
+        crate::script::MAKASAR,                // 180
+        crate::script::MEDEFAIDRIN,            // 181
+        crate::script::HANIFI_ROHINGYA,        // 182
+        crate::script::SOGDIAN,                // 183
+        crate::script::OLD_SOGDIAN,            // 184
+        crate::script::ELYMAIC,                // 185
+        crate::script::NYIAKENG_PUACHUE_HMONG, // 186
+        crate::script::NANDINAGARI,            // 187
+        crate::script::WANCHO,                 // 188
+        crate::script::CHORASMIAN,             // 189
+        crate::script::DIVES_AKURU,            // 190
+        crate::script::KHITAN_SMALL_SCRIPT,    // 191
+        crate::script::YEZIDI,                 // 192
+        crate::script::CYPRO_MINOAN,           // 193
+        crate::script::OLD_UYGHUR,             // 194
+        crate::script::TANGSA,                 // 195
+        crate::script::TOTO,                   // 196
+        crate::script::VITHKUQI,               // 197
+        crate::script::KAWI,                   // 198
+        crate::script::NAG_MUNDARI,            // 199
+        crate::script::ARABIC,                 // 200
+        crate::script::GARAY,                  // 201
+        crate::script::GURUNG_KHEMA,           // 202
+        crate::script::KIRAT_RAI,              // 203
+        crate::script::OL_ONAL,                // 204
+        crate::script::SUNUWAR,                // 205
+        crate::script::TODHRI,                 // 206
+        crate::script::TULU_TIGALARI,          // 207
+        crate::script::BERIA_ERFE,             // 208
+        crate::script::SIDETIC,                // 209
+        crate::script::TAI_YO,                 // 210
+        crate::script::TOLONG_SIKI,            // 211
+    ];
+}
+
+#[cfg(not(feature = "icu"))]
+#[expect(clippy::manual_range_contains)]
+mod builtin {
+    use super::{super::algs::*, ucd_table::ucd::*, Codepoint, GeneralCategory, Script};
+
+    pub(crate) fn script_for(c: u32) -> Script {
+        _hb_ucd_sc_map[_hb_ucd_sc(c as usize) as usize]
+    }
+
+    pub(crate) fn general_category_for(c: u32) -> GeneralCategory {
+        GeneralCategory(_hb_ucd_gc(c as usize))
+    }
+
+    pub(crate) fn combining_class_for(c: u32) -> u8 {
+        _hb_ucd_ccc(c as usize)
+    }
+
+    pub(crate) fn mirroring_for(c: u32) -> Option<Codepoint> {
+        let delta = _hb_ucd_bmg(c as usize);
+        if delta == 0 {
+            None
+        } else {
+            Some(((c as i32).wrapping_add(delta as i32)) as u32)
+        }
+    }
+
+    pub(crate) fn compose(a: Codepoint, b: Codepoint) -> Option<Codepoint> {
+        // Hangul is handled algorithmically.
+        if let Some(ab) = compose_hangul(a, b) {
+            return Some(ab);
+        }
+
+        let u: u32;
+
+        if (a & 0xFFFF_F800) == 0x0000 && (b & 0xFFFF_FF80) == 0x0300 {
+            /* If "a" is small enough and "b" is in the U+0300 range,
+             * the composition data is encoded in a 32bit array sorted
+             * by "a,b" pair. */
+            let k = HB_CODEPOINT_ENCODE3_11_7_14(a, b, 0);
+            let v = _hb_ucd_dm2_u32_map
+                .binary_search_by(|probe| {
+                    let key = probe & HB_CODEPOINT_ENCODE3_11_7_14(0x001F_FFFF, 0x001F_FFFF, 0);
+                    key.cmp(&k)
+                })
+                .ok()
+                .map(|index| _hb_ucd_dm2_u32_map[index]);
+
+            if let Some(value) = v {
+                u = HB_CODEPOINT_DECODE3_11_7_14_3(value);
+            } else {
+                return None;
+            }
+        } else {
+            /* Otherwise it is stored in a 64bit array sorted by
+             * "a,b" pair. */
+            let k = HB_CODEPOINT_ENCODE3(a, b, 0);
+            let v = _hb_ucd_dm2_u64_map
+                .binary_search_by(|probe| {
+                    let key = probe & HB_CODEPOINT_ENCODE3(0x001F_FFFF, 0x001F_FFFF, 0);
+                    key.cmp(&k)
+                })
+                .ok()
+                .map(|index| _hb_ucd_dm2_u64_map[index]);
+
+            if let Some(value) = v {
+                u = HB_CODEPOINT_DECODE3_3(value);
+            } else {
+                return None;
+            }
+        }
+
+        if u == 0 {
+            None
+        } else {
+            Some(u)
+        }
+    }
+
+    pub(crate) fn decompose(ab: Codepoint) -> Option<(Codepoint, Codepoint)> {
+        if let Some((a, b)) = decompose_hangul(ab) {
+            return Some((a, b));
+        }
+
+        let mut i = _hb_ucd_dm(ab as usize) as usize;
+
+        // If no data, there's no decomposition.
+        if i == 0 {
+            return None;
+        }
+        i -= 1;
+
+        if i < _hb_ucd_dm1_p0_map.len() + _hb_ucd_dm1_p2_map.len() {
+            let a = if i < _hb_ucd_dm1_p0_map.len() {
+                _hb_ucd_dm1_p0_map[i] as u32
+            } else {
+                let j = i - _hb_ucd_dm1_p0_map.len();
+                0x20000 | _hb_ucd_dm1_p2_map[j] as u32
+            };
+            return Some((a, 0));
+        }
+
+        i -= _hb_ucd_dm1_p0_map.len() + _hb_ucd_dm1_p2_map.len();
+
+        if i < _hb_ucd_dm2_u32_map.len() {
+            let v = _hb_ucd_dm2_u32_map[i];
+            let a = HB_CODEPOINT_DECODE3_11_7_14_1(v);
+            let b = HB_CODEPOINT_DECODE3_11_7_14_2(v);
+            return Some((a, b));
+        }
+
+        i -= _hb_ucd_dm2_u32_map.len();
+
+        let v = _hb_ucd_dm2_u64_map[i];
+        let a = HB_CODEPOINT_DECODE3_1(v);
+        let b = HB_CODEPOINT_DECODE3_2(v);
+        Some((a, b))
+    }
+
+    const S_BASE: u32 = 0xAC00;
+    const L_BASE: u32 = 0x1100;
+    const V_BASE: u32 = 0x1161;
+    const T_BASE: u32 = 0x11A7;
+    const L_COUNT: u32 = 19;
+    const V_COUNT: u32 = 21;
+    const T_COUNT: u32 = 28;
+    const N_COUNT: u32 = V_COUNT * T_COUNT;
+    const S_COUNT: u32 = L_COUNT * N_COUNT;
+
+    fn compose_hangul(a: Codepoint, b: Codepoint) -> Option<Codepoint> {
+        let l = a;
+        let v = b;
+        if L_BASE <= l && l < (L_BASE + L_COUNT) && V_BASE <= v && v < (V_BASE + V_COUNT) {
+            let r = S_BASE + (l - L_BASE) * N_COUNT + (v - V_BASE) * T_COUNT;
+            Some(r)
+        } else if S_BASE <= l
+            && l <= (S_BASE + S_COUNT - T_COUNT)
+            && T_BASE <= v
+            && v < (T_BASE + T_COUNT)
+            && (l - S_BASE) % T_COUNT == 0
+        {
+            let r = l + (v - T_BASE);
+            Some(r)
+        } else {
+            None
+        }
+    }
+
+    fn decompose_hangul(ab: Codepoint) -> Option<(Codepoint, Codepoint)> {
+        let si = ab.wrapping_sub(S_BASE);
+        if si >= S_COUNT {
+            return None;
+        }
+        let (a, b) = if si % T_COUNT != 0 {
+            // LV,T
+            (S_BASE + (si / T_COUNT) * T_COUNT, T_BASE + (si % T_COUNT))
+        } else {
+            // L,V
+            (L_BASE + (si / N_COUNT), V_BASE + (si % N_COUNT) / T_COUNT)
+        };
+        Some((a, b))
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    #[test]
+    fn general_category_ranges() {
+        use super::GeneralCategory as Gc;
+        assert!(!Gc::CONTROL.in_range_inclusive(Gc::FORMAT, Gc::NON_SPACING_MARK));
+        assert!(Gc::LOWERCASE_LETTER.in_range_inclusive(Gc::LOWERCASE_LETTER, Gc::UPPERCASE_LETTER));
+        assert!(Gc::UPPERCASE_LETTER.in_range_inclusive(Gc::LOWERCASE_LETTER, Gc::UPPERCASE_LETTER));
+        assert!(!Gc::SPACING_MARK.in_range_inclusive(Gc::LOWERCASE_LETTER, Gc::UPPERCASE_LETTER));
+    }
+
+    #[cfg(feature = "icu")]
+    #[test]
+    fn script_mapping() {
+        use icu_properties::props::Script as IcuScript;
+        use icu_properties::PropertyNamesShort;
+        // These ICU script codes are aliases, placeholders, or meta scripts.
+        // No Unicode code points map directly to them via the Script property,
+        // so our direct-script table intentionally keeps them as UNKNOWN.
+        const ALLOWED_UNKNOWN_SCRIPT_VALUES: &[usize] = &[
+            54, 64, 67, 69, 70, 77, 85, 93, 94, 98, 100, 102, 105, 114, 119, 124, 128, 129, 132,
+            138, 139, 147, 148, 155, 172, 174,
+        ];
+        let short_names = PropertyNamesShort::<IcuScript>::new();
+        let mut mismatches = Vec::new();
+        for (icu4c_value, script) in super::icu::SCRIPT_MAP.iter().enumerate() {
+            let icu_script = IcuScript::from_icu4c_value(icu4c_value as u16);
+            match short_names.get(icu_script) {
+                Some(short_name) => {
+                    let expected = if short_name.len() == 4 {
+                        let bytes = short_name.as_bytes();
+                        let tag =
+                            crate::Tag::from_be_bytes([bytes[0], bytes[1], bytes[2], bytes[3]]);
+                        crate::Script::from_iso15924_tag(tag).unwrap_or(crate::script::UNKNOWN)
+                    } else {
+                        crate::script::UNKNOWN
+                    };
+                    if *script == crate::script::UNKNOWN
+                        && ALLOWED_UNKNOWN_SCRIPT_VALUES.contains(&icu4c_value)
+                    {
+                        continue;
+                    }
+                    if expected != *script {
+                        let expected_tag = expected.tag().to_be_bytes();
+                        let actual_tag = script.tag().to_be_bytes();
+                        mismatches.push(format!(
+                            "value {icu4c_value}: Script::from_iso15924_tag({short_name}) -> {}, mapped tag {}",
+                            core::str::from_utf8(&expected_tag).unwrap_or("????"),
+                            core::str::from_utf8(&actual_tag).unwrap_or("????")
+                        ));
+                    }
+                }
+                None => {
+                    if *script != crate::script::UNKNOWN {
+                        mismatches.push(format!(
+                            "value {icu4c_value}: no ICU short name but mapped tag {}",
+                            core::str::from_utf8(&script.tag().to_be_bytes()).unwrap_or("????")
+                        ));
+                    }
+                }
+            }
+        }
+        assert!(
+            mismatches.is_empty(),
+            "script mapping mismatches:\n{}",
+            mismatches.join("\n")
+        );
+    }
+}
diff --git a/harfrust/src/hb/ucd_table.rs b/harfrust/src/unicode/ucd_table.rs
similarity index 99%
rename from harfrust/src/hb/ucd_table.rs
rename to harfrust/src/unicode/ucd_table.rs
index 980b32a2..c3d2e66d 100644
--- a/harfrust/src/hb/ucd_table.rs
+++ b/harfrust/src/unicode/ucd_table.rs
@@ -12,9 +12,7 @@ pub(crate) mod ucd {
 #![allow(unused_parens)]
 #![allow(clippy::unnecessary_cast, clippy::unreadable_literal, clippy::double_parens)]
 
-use crate::hb::algs::{HB_CODEPOINT_ENCODE3, HB_CODEPOINT_ENCODE3_11_7_14};
-use crate::hb::common::script;
-use crate::hb::common::Script as hb_script_t;
+use crate::{algs::{HB_CODEPOINT_ENCODE3, HB_CODEPOINT_ENCODE3_11_7_14}, script, Script as hb_script_t};
 
 #[allow(dead_code, non_upper_case_globals, clippy::allow_attributes_without_reason, missing_docs)]
 pub(crate) static _hb_ucd_sc_map: [hb_script_t; 176]=