py/parsenum: Throw an exception for invalid int literals like "01".

This includes making int("01") parse in base 10 like standard Python.
When a base of 0 is specified it means auto-detect based on the prefix, and
literals begining with 0 (except when the literal is all 0's) like "01" are
then invalid and now throw an exception.

The new error message is different from CPython. It says e.g.,
`SyntaxError: invalid syntax for integer with base 0: '09'`

Additional test cases were added to cover the changed & added code.

Co-authored-by: Damien George <[email protected]>
Signed-off-by: Jeff Epler <[email protected]>

Author: jepler

py/objint.c

@@ -55,7 +55,7 @@ static mp_obj_t mp_obj_int_make_new(const mp_obj_type_t *type_in, size_t n_args,
                 return o;
             } else if (mp_get_buffer(args[0], &bufinfo, MP_BUFFER_READ)) {
                 // a textual representation, parse it
-                return mp_parse_num_integer(bufinfo.buf, bufinfo.len, 0, NULL);
+                return mp_parse_num_integer(bufinfo.buf, bufinfo.len, 10, NULL);
             #if MICROPY_PY_BUILTINS_FLOAT
             } else if (mp_obj_is_float(args[0])) {
                 return mp_obj_new_int_from_float(mp_obj_float_get(args[0]));

py/parsenum.c

@@ -151,13 +151,13 @@ mp_obj_t mp_parse_num_integer(const char *restrict str_, size_t len, int base, m
         raise_exc(exc, lex);
         #elif MICROPY_ERROR_REPORTING == MICROPY_ERROR_REPORTING_NORMAL
         mp_obj_t exc = mp_obj_new_exception_msg_varg(&mp_type_ValueError,
-            MP_ERROR_TEXT("invalid syntax for integer with base %d"), base);
+            MP_ERROR_TEXT("invalid syntax for integer with base %d"), base == 1 ? 0 : base);
         raise_exc(exc, lex);
         #else
         vstr_t vstr;
         mp_print_t print;
         vstr_init_print(&vstr, 50, &print);
-        mp_printf(&print, "invalid syntax for integer with base %d: ", base);
+        mp_printf(&print, "invalid syntax for integer with base %d: ", base == 1 ? 0 : base);
         mp_str_print_quoted(&print, str_val_start, top - str_val_start, true);
         mp_obj_t exc = mp_obj_new_exception_arg1(&mp_type_ValueError,
             mp_obj_new_str_from_utf8_vstr(&vstr));

py/parsenumbase.c

@@ -30,35 +30,28 @@
 
 // find real radix base, and strip preceding '0x', '0o' and '0b'
 // puts base in *base, and returns number of bytes to skip the prefix
+// in base-0, puts 1 in *base to indicate a number that starts with 0, to provoke a
+// ValueError if it's not all-digits-zero.
 size_t mp_parse_num_base(const char *str, size_t len, int *base) {
     const byte *p = (const byte *)str;
     if (len <= 1) {
         goto no_prefix;
     }
     unichar c = *(p++);
-    if ((*base == 0 || *base == 16) && c == '0') {
-        c = *(p++);
-        if ((c | 32) == 'x') {
+    if (c == '0') {
+        c = *(p++) | 32;
+        int b = *base;
+        if (c == 'x' && !(b & ~16)) {
             *base = 16;
-        } else if (*base == 0 && (c | 32) == 'o') {
+        } else if (c == 'o' && !(b & ~8)) {
             *base = 8;
-        } else if (*base == 0 && (c | 32) == 'b') {
+        } else if (c == 'b' && !(b & ~2)) {
             *base = 2;
         } else {
-            if (*base == 0) {
-                *base = 10;
-            }
-            p -= 2;
-        }
-    } else if (*base == 8 && c == '0') {
-        c = *(p++);
-        if ((c | 32) != 'o') {
-            p -= 2;
-        }
-    } else if (*base == 2 && c == '0') {
-        c = *(p++);
-        if ((c | 32) != 'b') {
             p -= 2;
+            if (b == 0) {
+                *base = 1;
+            }
         }
     } else {
         p--;

tests/basics/int1.py

@@ -13,6 +13,7 @@
 print(int('+1'))
 print(int('-1'))
 print(int('01'))
+print(int('00'))
 print(int('9'))
 print(int('10'))
 print(int('+10'))
@@ -31,6 +32,7 @@
 print(int('0', 10))
 print(int('1', 10))
 print(int(' \t 1 \t ', 10))
+print(int(' \t 00 \t ', 10))
 print(int('11', 10))
 print(int('11', 16))
 print(int('11', 8))
@@ -52,6 +54,17 @@
 print(int('0o12  \t  ', 8))
 print(int(b"12", 10))
 print(int(b"12"))
+print(int('000 ', 0))
+print(int('000 ', 2))
+print(int('000 ', 8))
+print(int('000 ', 10))
+print(int('000 ', 16))
+print(int('000 ', 36))
+print(int('010 ', 2))
+print(int('010 ', 8))
+print(int('010 ', 10))
+print(int('010 ', 16))
+print(int('010 ', 36))
 
 
 def test(value, base):
@@ -79,6 +92,8 @@ def test(value, base):
 test('0xg', 16)
 test('1 1', 16)
 test('123', 37)
+test('01', 0)
+test('01 ', 0)
 
 # check that we don't parse this as a floating point number
 print(0x1e+1)

tests/basics/lexer.py

@@ -83,3 +83,11 @@ def a(x):
     exec(r"'\U0000000'")
 except SyntaxError:
     print("SyntaxError")
+
+# Properly formed integer literals
+print(eval("00"))
+# badly formed integer literals
+try:
+    eval("01")
+except SyntaxError:
+    print("SyntaxError")