[Python-checkins] bpo-39500: Document PyUnicode_IsIdentifier() function (GH-18397)

[Victor Stinner]

https://github.com/python/cpython/commit/f3e7ea5b8c220cd63101e419d529c8563f9c6115
commit: f3e7ea5b8c220cd63101e419d529c8563f9c6115
branch: master
author: Victor Stinner <vstinner at python.org>
committer: GitHub <noreply at github.com>
date: 2020-02-11T14:29:33+01:00
summary:

bpo-39500: Document PyUnicode_IsIdentifier() function (GH-18397)

PyUnicode_IsIdentifier() does not call Py_FatalError() anymore if the
string is not ready.

files:
A Misc/NEWS.d/next/C API/2020-02-07-09-35-43.bpo-39500.xRAEgX.rst
M Doc/c-api/unicode.rst
M Objects/unicodeobject.c
M Parser/tokenizer.c

diff --git a/Doc/c-api/unicode.rst b/Doc/c-api/unicode.rst
index 96d77c4084132..b1787ed1ce89c 100644
--- a/Doc/c-api/unicode.rst
+++ b/Doc/c-api/unicode.rst
@@ -240,6 +240,16 @@ access internal read-only data of Unicode objects:
       :c:func:`PyUnicode_nBYTE_DATA` family of macros.
 
 
+.. c:function:: int PyUnicode_IsIdentifier(PyObject *o)
+
+   Return ``1`` if the string is a valid identifier according to the language
+   definition, section :ref:`identifiers`. Return ``0`` otherwise.
+
+   .. versionchanged:: 3.9
+      The function does not call :c:func:`Py_FatalError` anymore if the string
+      is not ready.
+
+
 Unicode Character Properties
 """"""""""""""""""""""""""""
 
diff --git a/Misc/NEWS.d/next/C API/2020-02-07-09-35-43.bpo-39500.xRAEgX.rst b/Misc/NEWS.d/next/C API/2020-02-07-09-35-43.bpo-39500.xRAEgX.rst
new file mode 100644
index 0000000000000..2ca359f0ec11a
--- /dev/null
+++ b/Misc/NEWS.d/next/C API/2020-02-07-09-35-43.bpo-39500.xRAEgX.rst	
@@ -0,0 +1,2 @@
+:c:func:`PyUnicode_IsIdentifier` does not call :c:func:`Py_FatalError`
+anymore if the string is not ready.
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index fd08ddbf57434..aa874f2a12d29 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -12198,22 +12198,33 @@ unicode_isnumeric_impl(PyObject *self)
 int
 PyUnicode_IsIdentifier(PyObject *self)
 {
-    int kind;
-    void *data;
     Py_ssize_t i;
-    Py_UCS4 first;
+    int ready = PyUnicode_IS_READY(self);
 
-    if (PyUnicode_READY(self) == -1) {
-        Py_FatalError("identifier not ready");
+    Py_ssize_t len = ready ? PyUnicode_GET_LENGTH(self) : PyUnicode_GET_SIZE(self);
+    if (len == 0) {
+        /* an empty string is not a valid identifier */
         return 0;
     }
 
-    /* Special case for empty strings */
-    if (PyUnicode_GET_LENGTH(self) == 0)
-        return 0;
-    kind = PyUnicode_KIND(self);
-    data = PyUnicode_DATA(self);
+    int kind;
+    void *data;
+    wchar_t *wstr;
+    if (ready) {
+        kind = PyUnicode_KIND(self);
+        data = PyUnicode_DATA(self);
+    }
+    else {
+        wstr = _PyUnicode_WSTR(self);
+    }
 
+    Py_UCS4 ch;
+    if (ready) {
+        ch = PyUnicode_READ(kind, data, 0);
+    }
+    else {
+        ch = wstr[0];
+    }
     /* PEP 3131 says that the first character must be in
        XID_Start and subsequent characters in XID_Continue,
        and for the ASCII range, the 2.x rules apply (i.e
@@ -12222,13 +12233,21 @@ PyUnicode_IsIdentifier(PyObject *self)
        definition of XID_Start and XID_Continue, it is sufficient
        to check just for these, except that _ must be allowed
        as starting an identifier.  */
-    first = PyUnicode_READ(kind, data, 0);
-    if (!_PyUnicode_IsXidStart(first) && first != 0x5F /* LOW LINE */)
+    if (!_PyUnicode_IsXidStart(ch) && ch != 0x5F /* LOW LINE */) {
         return 0;
+    }
 
-    for (i = 1; i < PyUnicode_GET_LENGTH(self); i++)
-        if (!_PyUnicode_IsXidContinue(PyUnicode_READ(kind, data, i)))
+    for (i = 1; i < len; i++) {
+        if (ready) {
+            ch = PyUnicode_READ(kind, data, i);
+        }
+        else {
+            ch = wstr[i];
+        }
+        if (!_PyUnicode_IsXidContinue(ch)) {
             return 0;
+        }
+    }
     return 1;
 }
 
diff --git a/Parser/tokenizer.c b/Parser/tokenizer.c
index f73c32684c7b7..c37cd927df5a4 100644
--- a/Parser/tokenizer.c
+++ b/Parser/tokenizer.c
@@ -1079,8 +1079,9 @@ verify_identifier(struct tok_state *tok)
     }
     result = PyUnicode_IsIdentifier(s);
     Py_DECREF(s);
-    if (result == 0)
+    if (result == 0) {
         tok->done = E_IDENTIFIER;
+    }
     return result;
 }