youknowone · Oct 29, 2026 · Oct 29, 2026
diff --git a/.cspell.json b/.cspell.json
@@ -97,6 +97,8 @@
     "implicits",
     "inity",
     "interps",
+    "isxidcontinue",
+    "isxidstart",
     "jitted",
     "jitting",
     "kwonly",

diff --git a/Lib/test/test_unicodedata.py b/Lib/test/test_unicodedata.py
@@ -769,7 +769,6 @@ def parse_cp(s):
         for cp in range(0, sys.maxunicode + 1):
             self.assertEqual(self.db.name(chr(cp), None), names_ref.get(cp))
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON; AttributeError: module 'unicodedata' has no attribute 'isxidstart'
     def test_isxidstart(self):
         self.assertTrue(self.db.isxidstart('S'))
         self.assertTrue(self.db.isxidstart('\u0AD0'))  # GUJARATI OM
@@ -783,7 +782,6 @@ def test_isxidstart(self):
         self.assertRaises(TypeError, self.db.isxidstart)
         self.assertRaises(TypeError, self.db.isxidstart, 'xx')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON; AttributeError: module 'unicodedata' has no attribute 'isxidcontinue'
     def test_isxidcontinue(self):
         self.assertTrue(self.db.isxidcontinue('S'))
         self.assertTrue(self.db.isxidcontinue('_'))
@@ -1013,7 +1011,6 @@ def graphemes(*args):
             'a\U0001F1FA\U0001F1E6\U0001F1FA\U0001F1F3'),
             ['a', '\U0001F1FA\U0001F1E6', '\U0001F1FA\U0001F1F3'])
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON; AttributeError: module 'unicodedata' has no attribute 'block'
     def test_block(self):
         self.assertEqual(self.db.block('\u0000'), 'Basic Latin')
         self.assertEqual(self.db.block('\u0041'), 'Basic Latin')
@@ -1098,7 +1095,6 @@ def test_block(self):
         self.assertEqual(self.db.block('\U00100000'), 'Supplementary Private Use Area-B')
         self.assertEqual(self.db.block('\U0010FFFF'), 'Supplementary Private Use Area-B')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON; AttributeError: module 'unicodedata' has no attribute 'block'
     def test_block_invalid_input(self):
         self.assertRaises(TypeError, self.db.block)
         self.assertRaises(TypeError, self.db.block, b'x')

diff --git a/crates/stdlib/src/unicodedata.rs b/crates/stdlib/src/unicodedata.rs
@@ -33,7 +33,7 @@ impl<'a> TryFromBorrowedObject<'a> for NormalizeFormArg {
 mod unicodedata {
     use super::{NormalizeFormArg, unicode_core};
     use crate::vm::{
-        Py, PyObjectRef, PyPayload, PyResult, VirtualMachine,
+        Py, PyObject, PyObjectRef, PyPayload, PyResult, VirtualMachine,
         builtins::{PyModule, PyStr, PyStrRef},
         common::lock::PyMutex,
         function::OptionalArg,
@@ -52,6 +52,56 @@ mod unicodedata {
             .map_err(|_| vm.new_type_error("argument must be a unicode character, not str"))
     }
 
+    fn property_char(
+        character: &PyObject,
+        function: &str,
+        vm: &VirtualMachine,
+    ) -> PyResult<CodePoint> {
+        let Some(character_str) = character.downcast_ref::<PyStr>() else {
+            let kind = if vm.is_none(character) {
+                "None".to_owned()
+            } else {
+                character.class().slot_name().to_string()
+            };
+            let kind = &kind[..kind.floor_char_boundary(50.min(kind.len()))];
+            return Err(vm.new_type_error(format!(
+                "{function}() argument must be a unicode character, not {kind}"
+            )));
+        };
+        character_str
+            .as_wtf8()
+            .code_points()
+            .exactly_one()
+            .map_err(|_| {
+                vm.new_type_error(format!(
+                    "{function}(): argument must be a unicode character, not a string of length {}",
+                    character_str.char_len()
+                ))
+            })
+    }
+
+    /// Return True if the character has the XID_Start property, else False.
+    #[pyfunction]
+    fn isxidstart(chr: PyObjectRef, vm: &VirtualMachine) -> PyResult<bool> {
+        Ok(property_char(&chr, "isxidstart", vm)?
+            .to_char()
+            .is_some_and(unicode_core::identifier::is_xid_start))
+    }
+
+    /// Return True if the character has the XID_Continue property, else False.
+    #[pyfunction]
+    fn isxidcontinue(chr: PyObjectRef, vm: &VirtualMachine) -> PyResult<bool> {
+        Ok(property_char(&chr, "isxidcontinue", vm)?
+            .to_char()
+            .is_some_and(unicode_core::identifier::is_xid_continue))
+    }
+
+    /// Return block assigned to the character chr.
+    #[pyfunction]
+    fn block(chr: PyObjectRef, vm: &VirtualMachine) -> PyResult<&'static str> {
+        property_char(&chr, "block", vm).map(unicode_core::data::block)
+    }
+
     #[pyfunction]
     fn grapheme_cluster_break(chr: PyStrRef, vm: &VirtualMachine) -> PyResult<&'static str> {
         grapheme_char(&chr, vm).map(grapheme::grapheme_cluster_break)

diff --git a/crates/unicode/build.rs b/crates/unicode/build.rs
@@ -443,6 +443,17 @@ fn generate_algo_names(names: Vec<(u32, Box<str>)>) {
     );
 }
 
+/// Block names and ranges from the same Unicode release as the other latest tables.
+fn generate_blocks() {
+    let mut writer = open_writer("blocks.rs");
+    write_slice_pre(&mut writer, "BLOCKS", "(u32, u32, &str)");
+    for line in UnicodeLineReader::from_file_name("Blocks.txt", true) {
+        let name = line.field(NonZeroUsize::new(1).unwrap(), None);
+        writeln!(writer, "({}, {}, {name:?}),", line.start, line.end).unwrap();
+    }
+    write_slice_post(&mut writer);
+}
+
 /// Formal name aliases (`NameAliases.txt`) and named sequences
 /// (`NamedSequences.txt`).  `unicodedata.lookup` consults both; `\N{}`
 /// accepts only a single character, so named sequences stay out of
@@ -892,6 +903,7 @@ fn drive_parsers() {
         binary_props_3_2,
         combining_class_3_2,
         generate_name_lookups,
+        generate_blocks,
     ];
 
     let mut handles = Vec::with_capacity(parsers.len() + 2);

diff --git a/crates/unicode/src/data.rs b/crates/unicode/src/data.rs
@@ -20,6 +20,7 @@ use icu_properties::props::{
 use rustpython_wtf8::CodePoint;
 
 // Latest Unicode
+include!(concat!(env!("OUT_DIR"), "/generated/blocks.rs"));
 include!(concat!(env!("OUT_DIR"), "/generated/algo_names.rs"));
 include!(concat!(env!("OUT_DIR"), "/generated/name_lookups.rs"));
 include!(concat!(env!("OUT_DIR"), "/generated/decomp.rs"));
@@ -268,6 +269,23 @@ pub fn character_name(ch: char) -> Option<String> {
         .or_else(|| lookup_table(ALGO_NAMES, ch).map(|v| v.to_name(ch)))
 }
 
+/// Unicode block name, including ranges containing surrogates or unassigned points.
+#[must_use]
+pub fn block(c: CodePoint) -> &'static str {
+    let c = c.to_u32();
+    BLOCKS
+        .binary_search_by(|&(start, end, _)| {
+            if c > end {
+                Ordering::Less
+            } else if c < start {
+                Ordering::Greater
+            } else {
+                Ordering::Equal
+            }
+        })
+        .map_or("No_Block", |i| BLOCKS[i].2)
+}
+
 /// A view over the Unicode character database at a fixed version.
 ///
 /// `modern` selects the latest bundled UCD; otherwise the Unicode 3.2.0 tables

diff --git a/crates/unicode/unicode/README.md b/crates/unicode/unicode/README.md
@@ -7,3 +7,5 @@ The files in `latest` need to be periodically bumped to match `icu4x` and Rust.
 `NameAliases.txt` and `NamedSequences.txt` feed `unicodedata.lookup` (formal aliases and named sequences). They are not used by `unicodedata.name`, which still reports the official character name.
 
 RustPython vendors [Unicode 3.2.0](https://www.unicode.org/reports/tr28/tr28-3.html) to match CPython. CPython uses 3.2.0 to ensure backwards compatibility with a few older modules. Unicode 3.2.0 was released in 2002 and does not need periodic refreshes.
+
+`Blocks.txt` supplies the ranges and original block names for `unicodedata.block`. The current copy is from [Unicode 17.0.0](https://www.unicode.org/Public/17.0.0/ucd/Blocks.txt).
Original file line number	Diff line number	Diff line change
Expand Up		@@ -7,3 +7,5 @@ The files in `latest` need to be periodically bumped to match `icu4x` and Rust.
		`NameAliases.txt` and `NamedSequences.txt` feed `unicodedata.lookup` (formal aliases and named sequences). They are not used by `unicodedata.name`, which still reports the official character name.

		RustPython vendors [Unicode 3.2.0](https://www.unicode.org/reports/tr28/tr28-3.html) to match CPython. CPython uses 3.2.0 to ensure backwards compatibility with a few older modules. Unicode 3.2.0 was released in 2002 and does not need periodic refreshes.

		`Blocks.txt` supplies the ranges and original block names for `unicodedata.block`. The current copy is from [Unicode 17.0.0](https://www.unicode.org/Public/17.0.0/ucd/Blocks.txt).