[3.13] gh-130077: Properly match full soft keywords in the parser (GH-135317)

Lib/test/test_syntax.py

@@ -352,6 +352,13 @@
 Traceback (most recent call last):
 SyntaxError: invalid syntax
 
+# But prefixes of soft keywords should
+# still raise specialized errors
+
+>>> (mat x)
+Traceback (most recent call last):
+SyntaxError: invalid syntax. Perhaps you forgot a comma?
+
 From compiler_complex_args():
 
 >>> def f(None=1):

Misc/NEWS.d/next/Core_and_Builtins/2025-06-09-23-57-37.gh-issue-130077.MHknDB.rst

@@ -0,0 +1,2 @@
+Properly raise custom syntax errors when incorrect syntax containing names
+that are prefixes of soft keywords is encountered.  Patch by Pablo Galindo.

Parser/pegen.c

@@ -609,6 +609,8 @@
     Py_ssize_t size;
     PyBytes_AsStringAndSize(t->bytes, &the_token, &size);
     for (char **keyword = p->soft_keywords; *keyword != NULL; keyword++) {
+        if (strlen(*keyword) == (size_t)size &&
+            strncmp(*keyword, the_token, (size_t)size) == 0) {
         if (strncmp(*keyword, the_token, size) == 0) {
             return _PyPegen_name_from_token(p, t);
         }
@@ -617,30 +619,30 @@
 }
 
 static PyObject *
 parsenumber_raw(const char *s)
 {
     const char *end;
     long x;
     double dx;
     Py_complex compl;
     int imflag;
 
     assert(s != NULL);
     errno = 0;
     end = s + strlen(s) - 1;
     imflag = *end == 'j' || *end == 'J';
     if (s[0] == '0') {
         x = (long)PyOS_strtoul(s, (char **)&end, 0);
         if (x < 0 && errno == 0) {
             return PyLong_FromString(s, (char **)0, 0);
         }
     }
     else {
         x = PyOS_strtol(s, (char **)&end, 0);
     }
     if (*end == '\0') {
         if (errno != 0) {
             return PyLong_FromString(s, (char **)0, 0);
         }
         return PyLong_FromLong(x);
     }
@@ -661,7 +663,7 @@
 }
 
 static PyObject *
 parsenumber(const char *s)
 {
     char *dup;
     char *end;
@@ -690,7 +692,7 @@
 }
 
 expr_ty
 _PyPegen_number_token(Parser *p)
 {
     Token *t = _PyPegen_expect_token(p, NUMBER);
     if (t == NULL) {
@@ -749,7 +751,7 @@
    statement by looking at what is left in the buffer after parsing.
    Trailing whitespace and comments are OK. */
 static int // bool
 bad_single_statement(Parser *p)
 {
     char *cur = p->tok->cur;
     char c = *cur;
@@ -775,7 +777,7 @@
 }
 
 static int
 compute_parser_flags(PyCompilerFlags *flags)
 {
     int parser_flags = 0;
     if (!flags) {
@@ -870,7 +872,7 @@
 }
 
 static void
 reset_parser_state_for_error_pass(Parser *p)
 {
     for (int i = 0; i < p->fill; i++) {
         p->tokens[i]->memo = NULL;
@@ -883,7 +885,7 @@
 }
 
 static inline int
 _is_end_of_source(Parser *p) {
     int err = p->tok->done;
     return err == E_EOF || err == E_EOFS || err == E_EOLS;
 }
@@ -934,7 +936,7 @@
 }
 
 mod_ty
 _PyPegen_run_parser_from_file_pointer(FILE *fp, int start_rule, PyObject *filename_ob,
                              const char *enc, const char *ps1, const char *ps2,
                              PyCompilerFlags *flags, int *errcode,
                              PyObject **interactive_src, PyArena *arena)
@@ -982,7 +984,7 @@
 }
 
 mod_ty
 _PyPegen_run_parser_from_string(const char *str, int start_rule, PyObject *filename_ob,
                        PyCompilerFlags *flags, PyArena *arena)
 {
     int exec_input = start_rule == Py_file_input;
@@ -1020,4 +1022,4 @@
 error:
     _PyTokenizer_Free(tok);
     return result;
 }