[mypyc] Add str.isspace primitive (python#20842)

Author: VaggelisD

mypyc/lib-rt/CPy.h

@@ -770,7 +770,8 @@ PyObject *CPy_Encode(PyObject *obj, PyObject *encoding, PyObject *errors);
 Py_ssize_t CPyStr_Count(PyObject *unicode, PyObject *substring, CPyTagged start);
 Py_ssize_t CPyStr_CountFull(PyObject *unicode, PyObject *substring, CPyTagged start, CPyTagged end);
 CPyTagged CPyStr_Ord(PyObject *obj);
-
+PyObject *CPyStr_Multiply(PyObject *str, CPyTagged count);
+bool CPyStr_IsSpace(PyObject *str);
 
 // Bytes operations
 

mypyc/lib-rt/str_ops.c

@@ -621,3 +621,36 @@ CPyTagged CPyStr_Ord(PyObject *obj) {
         PyExc_TypeError, "ord() expected a character, but a string of length %zd found", s);
     return CPY_INT_TAG;
 }
+
+PyObject *CPyStr_Multiply(PyObject *str, CPyTagged count) {
+    Py_ssize_t temp_count = CPyTagged_AsSsize_t(count);
+    if (temp_count == -1 && PyErr_Occurred()) {
+        PyErr_SetString(PyExc_OverflowError, CPYTHON_LARGE_INT_ERRMSG);
+        return NULL;
+    }
+    return PySequence_Repeat(str, temp_count);
+}
+
+
+bool CPyStr_IsSpace(PyObject *str) {
+    Py_ssize_t len = PyUnicode_GET_LENGTH(str);
+    if (len == 0) return false;
+
+    if (PyUnicode_IS_ASCII(str)) {
+        const Py_UCS1 *data = PyUnicode_1BYTE_DATA(str);
+        for (Py_ssize_t i = 0; i < len; i++) {
+            if (!_Py_ascii_whitespace[data[i]])
+                return false;
+        }
+        return true;
+    }
+
+    int kind = PyUnicode_KIND(str);
+    const void *data = PyUnicode_DATA(str);
+    for (Py_ssize_t i = 0; i < len; i++) {
+        Py_UCS4 ch = PyUnicode_READ(kind, data, i);
+        if (!Py_UNICODE_ISSPACE(ch))
+            return false;
+    }
+    return true;
+}

mypyc/primitives/str_ops.py

@@ -375,6 +375,14 @@
     error_kind=ERR_NEG_INT,
 )
 
+method_op(
+    name="isspace",
+    arg_types=[str_rprimitive],
+    return_type=bool_rprimitive,
+    c_function_name="CPyStr_IsSpace",
+    error_kind=ERR_NEVER,
+)
+
 # obj.decode()
 method_op(
     name="decode",

mypyc/test-data/fixtures/ir.py

@@ -129,6 +129,7 @@ def removeprefix(self, prefix: str, /) -> str: ...
     def removesuffix(self, suffix: str, /) -> str: ...
     def islower(self) -> bool: ...
     def count(self, substr: str, start: Optional[int] = None, end: Optional[int] = None) -> int: pass
+    def isspace(self) -> bool: ...
 
 class float:
     def __init__(self, x: object) -> None: pass

mypyc/test-data/irbuild-str.test

@@ -771,3 +771,35 @@ L0:
     r0 = 'literal'
     r1 = 'literal'
     return 1
+
+[case testStrMultiply]
+def s_times_i(s: str, n: int) -> str:
+    return s * n
+def i_times_s(s: str, n: int) -> str:
+    return n * s
+[out]
+def s_times_i(s, n):
+    s :: str
+    n :: int
+    r0 :: str
+L0:
+    r0 = CPyStr_Multiply(s, n)
+    return r0
+def i_times_s(s, n):
+    s :: str
+    n :: int
+    r0 :: str
+L0:
+    r0 = CPyStr_Multiply(s, n)
+    return r0
+
+[case testStrIsSpace]
+def is_space(x: str) -> bool:
+    return x.isspace()
+[out]
+def is_space(x):
+    x :: str
+    r0 :: bool
+L0:
+    r0 = CPyStr_IsSpace(x)
+    return r0

mypyc/test-data/run-strings.test

@@ -1103,3 +1103,23 @@ def test_optional_ne() -> None:
     assert ne_s_opt_s_opt('y', 'x')
     assert ne_s_opt_s_opt(None, 'x')
     assert ne_s_opt_s_opt('x', None)
+
+[case testConstantFoldFormatArgs]
+from typing import Final
+
+FMT: Final = "{} {}"
+
+def test_format() -> None:
+    assert FMT.format(400 + 20, "roll" + "up") == "420 rollup"
+
+[case testIsSpace]
+from typing import Any
+
+def test_isspace() -> None:
+    # Verify correctness across all Unicode codepoints.
+    # Exercises UCS-1 (0x00-0xFF), UCS-2 (0x100-0xFFFF), and UCS-4 (0x10000-x10FFFF inclusive) string kinds.
+    # Any forces generic dispatch so we compare our primitive against stdlib's
+    for i in range(0x110000):
+        c = chr(i)
+        a: Any = c
+        assert c.isspace() == a.isspace()