gh-98740: Fix validation of conditional expressions in RE (GH-98764)

In very rare circumstances the JUMP opcode could be confused with the
argument of the opcode in the "then" part which doesn't end with the
JUMP opcode. This led to incorrect detection of the final JUMP opcode
and incorrect calculation of the size of the subexpression.

NOTE: Changed return value of functions _validate_inner() and
_validate_charset() in Modules/_sre/sre.c.  Now they return 0 on success,
-1 on failure, and 1 if the last op is JUMP (which usually is a failure).
Previously they returned 1 on success and 0 on failure.

diff --git a/Doc/library/re.rst b/Doc/library/re.rst
index 5b304f717b07fa38880788378cc0987736672027..0e7dda04b1d68d522eb88d3ab613375fa428fbea 100644 (file)
--- a/Doc/library/re.rst
+++ b/Doc/library/re.rst
@@ -483,6 +483,9 @@ The special characters are:
    some fixed length.  Patterns which start with negative lookbehind assertions may
    match at the beginning of the string being searched.
 
+.. _re-conditional-expression:
+.. index:: single: (?(; in regular expressions
+
 ``(?(id/name)yes-pattern|no-pattern)``
    Will try to match with ``yes-pattern`` if the group with given *id* or
    *name* exists, and with ``no-pattern`` if it doesn't. ``no-pattern`` is

diff --git a/Lib/test/test_re.py b/Lib/test/test_re.py
index 3f0f84ea8cee6fad52cedfba574a8f2fd59a993f..11628a236ade9a3b72989420dc7d7945df99e62b 100644 (file)
--- a/Lib/test/test_re.py
+++ b/Lib/test/test_re.py
@@ -630,6 +630,11 @@ class ReTests(unittest.TestCase):
         self.checkPatternError(r'()(?(2)a)',
                                "invalid group reference 2", 5)
 
+    def test_re_groupref_exists_validation_bug(self):
+        for i in range(256):
+            with self.subTest(code=i):
+                re.compile(r'()(?(1)\x%02x?)' % i)
+
     def test_re_groupref_overflow(self):
         from re._constants import MAXGROUPS
         self.checkTemplateError('()', r'\g<%s>' % MAXGROUPS, 'xx',

diff --git a/Misc/NEWS.d/next/Library/2022-10-27-12-56-38.gh-issue-98740.ZoqqGM.rst b/Misc/NEWS.d/next/Library/2022-10-27-12-56-38.gh-issue-98740.ZoqqGM.rst
new file mode 100644 (file)
index 0000000..887d506
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2022-10-27-12-56-38.gh-issue-98740.ZoqqGM.rst
@@ -0,0 +1,3 @@
+Fix internal error in the :mod:`re` module which in very rare circumstances
+prevented compilation of a regular expression containing a :ref:`conditional
+expression <re-conditional-expression>` without the "else" branch.

diff --git a/Modules/_sre/sre.c b/Modules/_sre/sre.c
index aae02652664f72b1397232007e9b0de42ac8aee0..44039ce25fb32a131546afbb8b90009d05c12e03 100644 (file)
--- a/Modules/_sre/sre.c
+++ b/Modules/_sre/sre.c
@@ -1623,7 +1623,7 @@ bad_template:
 #endif
 
 /* Report failure */
-#define FAIL do { VTRACE(("FAIL: %d\n", __LINE__)); return 0; } while (0)
+#define FAIL do { VTRACE(("FAIL: %d\n", __LINE__)); return -1; } while (0)
 
 /* Extract opcode, argument, or skip count from code array */
 #define GET_OP                                          \
@@ -1647,7 +1647,7 @@ bad_template:
         skip = *code;                                   \
         VTRACE(("%lu (skip to %p)\n",                   \
                (unsigned long)skip, code+skip));        \
-        if (skip-adj > (uintptr_t)(end - code))      \
+        if (skip-adj > (uintptr_t)(end - code))         \
             FAIL;                                       \
         code++;                                         \
     } while (0)
@@ -1736,9 +1736,10 @@ _validate_charset(SRE_CODE *code, SRE_CODE *end)
         }
     }
 
-    return 1;
+    return 0;
 }
 
+/* Returns 0 on success, -1 on failure, and 1 if the last op is JUMP. */
 static int
 _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
 {
@@ -1816,7 +1817,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
         case SRE_OP_IN_LOC_IGNORE:
             GET_SKIP;
             /* Stop 1 before the end; we check the FAILURE below */
-            if (!_validate_charset(code, code+skip-2))
+            if (_validate_charset(code, code+skip-2))
                 FAIL;
             if (code[skip-2] != SRE_OP_FAILURE)
                 FAIL;
@@ -1870,7 +1871,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                 }
                 /* Validate the charset */
                 if (flags & SRE_INFO_CHARSET) {
-                    if (!_validate_charset(code, newcode-1))
+                    if (_validate_charset(code, newcode-1))
                         FAIL;
                     if (newcode[-1] != SRE_OP_FAILURE)
                         FAIL;
@@ -1891,7 +1892,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                     if (skip == 0)
                         break;
                     /* Stop 2 before the end; we check the JUMP below */
-                    if (!_validate_inner(code, code+skip-3, groups))
+                    if (_validate_inner(code, code+skip-3, groups))
                         FAIL;
                     code += skip-3;
                     /* Check that it ends with a JUMP, and that each JUMP
@@ -1905,6 +1906,8 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                     else if (code+skip-1 != target)
                         FAIL;
                 }
+                if (code != target)
+                    FAIL;
             }
             break;
 
@@ -1920,7 +1923,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                     FAIL;
                 if (max > SRE_MAXREPEAT)
                     FAIL;
-                if (!_validate_inner(code, code+skip-4, groups))
+                if (_validate_inner(code, code+skip-4, groups))
                     FAIL;
                 code += skip-4;
                 GET_OP;
@@ -1940,7 +1943,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                     FAIL;
                 if (max > SRE_MAXREPEAT)
                     FAIL;
-                if (!_validate_inner(code, code+skip-3, groups))
+                if (_validate_inner(code, code+skip-3, groups))
                     FAIL;
                 code += skip-3;
                 GET_OP;
@@ -1958,7 +1961,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
         case SRE_OP_ATOMIC_GROUP:
             {
                 GET_SKIP;
-                if (!_validate_inner(code, code+skip-2, groups))
+                if (_validate_inner(code, code+skip-2, groups))
                     FAIL;
                 code += skip-2;
                 GET_OP;
@@ -2010,24 +2013,17 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                to allow arbitrary jumps anywhere in the code; so we just look
                for a JUMP opcode preceding our skip target.
             */
-            if (skip >= 3 && skip-3 < (uintptr_t)(end - code) &&
-                code[skip-3] == SRE_OP_JUMP)
-            {
-                VTRACE(("both then and else parts present\n"));
-                if (!_validate_inner(code+1, code+skip-3, groups))
-                    FAIL;
+            VTRACE(("then part:\n"));
+            int rc = _validate_inner(code+1, code+skip-1, groups);
+            if (rc == 1) {
+                VTRACE(("else part:\n"));
                 code += skip-2; /* Position after JUMP, at <skipno> */
                 GET_SKIP;
-                if (!_validate_inner(code, code+skip-1, groups))
-                    FAIL;
-                code += skip-1;
-            }
-            else {
-                VTRACE(("only a then part present\n"));
-                if (!_validate_inner(code+1, code+skip-1, groups))
-                    FAIL;
-                code += skip-1;
+                rc = _validate_inner(code, code+skip-1, groups);
             }
+            if (rc)
+                FAIL;
+            code += skip-1;
             break;
 
         case SRE_OP_ASSERT:
@@ -2038,7 +2034,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
             if (arg & 0x80000000)
                 FAIL; /* Width too large */
             /* Stop 1 before the end; we check the SUCCESS below */
-            if (!_validate_inner(code+1, code+skip-2, groups))
+            if (_validate_inner(code+1, code+skip-2, groups))
                 FAIL;
             code += skip-2;
             GET_OP;
@@ -2046,6 +2042,12 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
                 FAIL;
             break;
 
+        case SRE_OP_JUMP:
+            if (code + 1 != end)
+                FAIL;
+            VTRACE(("JUMP: %d\n", __LINE__));
+            return 1;
+
         default:
             FAIL;
 
@@ -2053,7 +2055,7 @@ _validate_inner(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
     }
 
     VTRACE(("okay\n"));
-    return 1;
+    return 0;
 }
 
 static int
@@ -2068,7 +2070,7 @@ _validate_outer(SRE_CODE *code, SRE_CODE *end, Py_ssize_t groups)
 static int
 _validate(PatternObject *self)
 {
-    if (!_validate_outer(self->code, self->code+self->codesize, self->groups))
+    if (_validate_outer(self->code, self->code+self->codesize, self->groups))
     {
         PyErr_SetString(PyExc_RuntimeError, "invalid SRE code");
         return 0;