Patchwork [CREW] parsers: fail fast if Python has wrong minor version (issue4110)

diff --git a/mercurial/parsers.c b/mercurial/parsers.c
--- a/mercurial/parsers.c
+++ b/mercurial/parsers.c
@@ -14,6 +14,8 @@ 
 
 #include "util.h"
 
+static char *versionerrortext = "Python minor version mismatch";
+
 static int8_t hextable[256] = {
 	-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
 	-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
@@ -1924,6 +1926,16 @@ 
 
 static void module_init(PyObject *mod)
 {
+	/* This module constant has two purposes.  First, it lets us unit test
+	 * the ImportError raised without hard-coding any error text.  This
+	 * means we can change the text in the future without breaking tests,
+	 * even across changesets without a recompile.  Second, its presence
+	 * can be used to determine whether the version-checking logic is
+	 * present, which also helps in testing across changesets without a
+	 * recompile.  Note that this means the pure-Python version of parsers
+	 * should not have this module constant. */
+	PyModule_AddStringConstant(mod, "versionerrortext", versionerrortext);
+
 	dirs_module_init(mod);
 
 	indexType.tp_new = PyType_GenericNew;
@@ -1941,6 +1953,24 @@ 
 	dirstate_unset = Py_BuildValue("ciii", 'n', 0, -1, -1);
 }
 
+static int check_python_version()
+{
+	PyObject *sys = PyImport_ImportModule("sys");
+	long hexversion = PyInt_AsLong(PyObject_GetAttrString(sys, "hexversion"));
+	/* sys.hexversion is a 32-bit number by default, so the -1 case
+	 * should only occur in unusual circumstances (e.g. if sys.hexversion
+	 * is manually set to an invalid value). */
+	if ((hexversion == -1) || (hexversion >> 16 != PY_VERSION_HEX >> 16)) {
+		PyErr_Format(PyExc_ImportError, "%s: The Mercurial extension "
+			"modules were compiled with Python " PY_VERSION ", but "
+			"Mercurial is currently using Python with sys.hexversion=%ld: "
+			"Python %s\n at: %s", versionerrortext, hexversion,
+			Py_GetVersion(), Py_GetProgramFullPath());
+		return -1;
+	}
+	return 0;
+}
+
 #ifdef IS_PY3K
 static struct PyModuleDef parsers_module = {
 	PyModuleDef_HEAD_INIT,
@@ -1952,6 +1982,8 @@ 
 
 PyMODINIT_FUNC PyInit_parsers(void)
 {
+	if (check_python_version() == -1)
+		return;
 	PyObject *mod = PyModule_Create(&parsers_module);
 	module_init(mod);
 	return mod;
@@ -1959,6 +1991,8 @@ 
 #else
 PyMODINIT_FUNC initparsers(void)
 {
+	if (check_python_version() == -1)
+		return;
 	PyObject *mod = Py_InitModule3("parsers", methods, parsers_doc);
 	module_init(mod);
 }
diff --git a/tests/test-parseindex2.py b/tests/test-parseindex2.py
--- a/tests/test-parseindex2.py
+++ b/tests/test-parseindex2.py
@@ -1,8 +1,13 @@ 
-"""This unit test tests parsers.parse_index2()."""
+"""This unit test primarily tests parsers.parse_index2().
+
+It also checks certain aspects of the parsers module as a whole.
+"""
 
 from mercurial import parsers
 from mercurial.node import nullid, nullrev
 import struct
+import subprocess
+import sys
 
 # original python implementation
 def gettype(q):
@@ -95,7 +100,70 @@ 
     index, chunkcache = parsers.parse_index2(data, inline)
     return list(index), chunkcache
 
+def importparsers(hexversion):
+    """Import mercurial.parsers with the given sys.hexversion."""
+    # The file parsers.c inspects sys.hexversion to determine the version
+    # of the currently-running Python interpreter, so we monkey-patch
+    # sys.hexversion to simulate using different versions.
+    code = ("import sys; sys.hexversion=%s; "
+            "import mercurial.parsers" % hexversion)
+    cmd = "python -c \"%s\"" % code
+    # We need to do these tests inside a subprocess because parser.c's
+    # version-checking code happens inside the module init function, and
+    # when using reload() to reimport an extension module, "The init function
+    # of extension modules is not called a second time"
+    # (from http://docs.python.org/2/library/functions.html?#reload).
+    p = subprocess.Popen(cmd, shell=True,
+                         stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
+    return p.communicate()  # returns stdout, stderr
+
+def printhexfail(testnumber, hexversion, stdout, expected):
+    try:
+        hexstring = hex(hexversion)
+    except TypeError:
+        hexstring = None
+    print ("FAILED: version test #%s with Python %s and patched "
+           "sys.hexversion %r (%r):\n Expected %s but got:\n-->'%s'\n" %
+           (testnumber, sys.version_info, hexversion, hexstring, expected,
+            stdout))
+
+def testversionokay(testnumber, hexversion):
+    stdout, stderr = importparsers(hexversion)
+    if stdout:
+        printhexfail(testnumber, hexversion, stdout, expected="no stdout")
+
+def testversionfail(testnumber, hexversion):
+    stdout, stderr = importparsers(hexversion)
+    # We include versionerrortext to distinguish from other ImportErrors.
+    errtext = "ImportError: %s" % parsers.versionerrortext
+    if errtext not in stdout:
+        printhexfail(testnumber, hexversion, stdout,
+                     expected="stdout to contain %r" % errtext)
+
+def makehex(major, minor, micro):
+    return int("%x%02x%02x00" % (major, minor, micro), 16)
+
+def runversiontests():
+    """Check the version-detection logic when importing parsers."""
+    info = sys.version_info
+    major, minor, micro = info[0], info[1], info[2]
+    # Test same major-minor versions.
+    testversionokay(1, makehex(major, minor, micro))
+    testversionokay(2, makehex(major, minor, micro + 1))
+    # Test different major-minor versions.
+    testversionfail(3, makehex(major + 1, minor, micro))
+    testversionfail(4, makehex(major, minor + 1, micro))
+    testversionfail(5, "'foo'")
+
 def runtest() :
+    # Only test the version-detection logic if it is present.
+    try:
+        parsers.versionerrortext
+    except AttributeError:
+        pass
+    else:
+        runversiontests()
+
     # Check that parse_index2() raises TypeError on bad arguments.
     try:
         parse_index2(0, True)