From 13b55291ac84d2fb74f93c286a0bd5673f439e38 Mon Sep 17 00:00:00 2001 From: "Gregory P. Smith" Date: Mon, 6 Sep 2010 08:30:23 +0000 Subject: [PATCH] hashlib has two new constant attributes: algorithms_guaranteed and algorithms_avaiable that respectively list the names of hash algorithms guaranteed to exist in all Python implementations and the names of hash algorithms available in the current process. Renames the attribute new in 3.2a0 'algorithms' to 'algorithms_guaranteed'. --- Doc/library/hashlib.rst | 29 ++++++++++++----- Lib/hashlib.py | 17 ++++++---- Lib/test/test_hashlib.py | 8 +++-- Misc/NEWS | 5 +++ Modules/_hashopenssl.c | 70 +++++++++++++++++++++++++++++++++++++++- 5 files changed, 111 insertions(+), 18 deletions(-) diff --git a/Doc/library/hashlib.rst b/Doc/library/hashlib.rst index 6fa01be0b67..fa9582119b1 100644 --- a/Doc/library/hashlib.rst +++ b/Doc/library/hashlib.rst @@ -70,10 +70,13 @@ More condensed: >>> hashlib.sha224(b"Nobody inspects the spammish repetition").hexdigest() 'a4337bc45a8fc544c03f52dc550cd6e1e87021bc896588bd79e901e2' -A generic :func:`new` constructor that takes the string name of the desired -algorithm as its first parameter also exists to allow access to the above listed -hashes as well as any other algorithms that your OpenSSL library may offer. The -named constructors are much faster than :func:`new` and should be preferred. +.. function:: new(name[, data]) + + Is a generic constructor that takes the string name of the desired + algorithm as its first parameter. It also exists to allow access to the + above listed hashes as well as any other algorithms that your OpenSSL + library may offer. The named constructors are much faster than :func:`new` + and should be preferred. Using :func:`new` with an algorithm provided by OpenSSL: @@ -82,12 +85,22 @@ Using :func:`new` with an algorithm provided by OpenSSL: >>> h.hexdigest() 'cc4a5ce1b3df48aec5d22d1f16b894a0b894eccc' -This module provides the following constant attribute: +Hashlib provides the following constant attributes: -.. data:: hashlib.algorithms +.. data:: algorithms_guaranteed - A tuple providing the names of the hash algorithms guaranteed to be - supported by this module. + Contains the names of the hash algorithms guaranteed to be supported + by this module on all platforms. + + .. versionadded:: 3.2 + +.. data:: algorithms_available + + Contains the names of the hash algorithms that are available + in the running Python interpreter. These names will be recognized + when passed to :func:`new`. :attr:`algorithms_guaranteed` + will always be a subset. Duplicate algorithms with different + name formats may appear in this set (thanks to OpenSSL). .. versionadded:: 3.2 diff --git a/Lib/hashlib.py b/Lib/hashlib.py index aacb268bfd4..9d860c481b1 100644 --- a/Lib/hashlib.py +++ b/Lib/hashlib.py @@ -1,6 +1,4 @@ -# $Id$ -# -# Copyright (C) 2005-2007 Gregory P. Smith (greg@krypto.org) +# Copyright (C) 2005-2010 Gregory P. Smith (greg@krypto.org) # Licensed to PSF under a Contributor Agreement. # @@ -15,8 +13,9 @@ than using new(name): md5(), sha1(), sha224(), sha256(), sha384(), and sha512() -More algorithms may be available on your platform but the above are -guaranteed to exist. +More algorithms may be available on your platform but the above are guaranteed +to exist. See the algorithms_guaranteed and algorithms_available attributes +to find out what algorithm names can be passed to new(). NOTE: If you want the adler32 or crc32 hash functions they are available in the zlib module. @@ -57,9 +56,11 @@ More condensed: # always available algorithm is added. __always_supported = ('md5', 'sha1', 'sha224', 'sha256', 'sha384', 'sha512') -algorithms = __always_supported +algorithms_guaranteed = __always_supported +algorithms_available = frozenset(__always_supported) -__all__ = __always_supported + ('new', 'algorithms') +__all__ = __always_supported + ('new', 'algorithms_guaranteed', + 'algorithms_available') def __get_builtin_constructor(name): @@ -124,6 +125,8 @@ try: import _hashlib new = __hash_new __get_hash = __get_openssl_constructor + algorithms_available = algorithms_available.union( + _hashlib.openssl_md_meth_names) except ImportError: new = __py_new __get_hash = __get_builtin_constructor diff --git a/Lib/test/test_hashlib.py b/Lib/test/test_hashlib.py index 4fb693e8faf..c8db6a86acf 100644 --- a/Lib/test/test_hashlib.py +++ b/Lib/test/test_hashlib.py @@ -101,11 +101,15 @@ class HashLibTestCase(unittest.TestCase): c = cons(a) c.hexdigest() - def test_algorithms_attribute(self): - self.assertEqual(hashlib.algorithms, + def test_algorithms_guaranteed(self): + self.assertEqual(hashlib.algorithms_guaranteed, tuple(_algo for _algo in self.supported_hash_names if _algo.islower())) + def test_algorithms_available(self): + self.assertTrue(set(hashlib.algorithms_guaranteed). + issubset(hashlib.algorithms_available)) + def test_unknown_hash(self): try: hashlib.new('spam spam spam spam spam') diff --git a/Misc/NEWS b/Misc/NEWS index 9567a43617c..732c1455894 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -17,6 +17,11 @@ Library unsupported operation is attempted (for example, writing to a file open only for reading). +- hashlib has two new constant attributes: algorithms_guaranteed and + algorithms_avaiable that respectively list the names of hash algorithms + guaranteed to exist in all Python implementations and the names of hash + algorithms available in the current process. + What's New in Python 3.2 Alpha 2? ================================= diff --git a/Modules/_hashopenssl.c b/Modules/_hashopenssl.c index 9f550a6a0dd..dd4317fca04 100644 --- a/Modules/_hashopenssl.c +++ b/Modules/_hashopenssl.c @@ -38,6 +38,8 @@ /* EVP is the preferred interface to hashing in OpenSSL */ #include +/* We use the object interface to discover what hashes OpenSSL supports. */ +#include #define MUNCH_SIZE INT_MAX @@ -488,6 +490,62 @@ EVP_new(PyObject *self, PyObject *args, PyObject *kwdict) return ret_obj; } + +/* State for our callback function so that it can accumulate a result. */ +typedef struct _internal_name_mapper_state { + PyObject *set; + int error; +} _InternalNameMapperState; + + +/* A callback function to pass to OpenSSL's OBJ_NAME_do_all(...) */ +static void +_openssl_hash_name_mapper(const OBJ_NAME *openssl_obj_name, void *arg) +{ + _InternalNameMapperState *state = (_InternalNameMapperState *)arg; + PyObject *py_name; + + assert(state != NULL); + if (openssl_obj_name == NULL) + return; + /* Ignore aliased names, they pollute the list and OpenSSL appears to + * have a its own definition of alias as the resulting list still + * contains duplicate and alternate names for several algorithms. */ + if (openssl_obj_name->alias) + return; + + py_name = PyUnicode_FromString(openssl_obj_name->name); + if (py_name == NULL) { + state->error = 1; + } else { + if (PySet_Add(state->set, py_name) != 0) { + Py_DECREF(py_name); + state->error = 1; + } + } +} + + +/* Ask OpenSSL for a list of supported ciphers, filling in a Python set. */ +static PyObject* +generate_hash_name_list(void) +{ + _InternalNameMapperState state; + state.set = PyFrozenSet_New(NULL); + if (state.set == NULL) + return NULL; + state.error = 0; + + OBJ_NAME_do_all(OBJ_NAME_TYPE_MD_METH, &_openssl_hash_name_mapper, &state); + + if (state.error) { + Py_DECREF(state.set); + return NULL; + } + return state.set; +} + + /* * This macro generates constructor function definitions for specific * hash algorithms. These constructors are much faster than calling @@ -581,7 +639,7 @@ static struct PyModuleDef _hashlibmodule = { PyMODINIT_FUNC PyInit__hashlib(void) { - PyObject *m; + PyObject *m, *openssl_md_meth_names; OpenSSL_add_all_digests(); @@ -598,6 +656,16 @@ PyInit__hashlib(void) if (m == NULL) return NULL; + openssl_md_meth_names = generate_hash_name_list(); + if (openssl_md_meth_names == NULL) { + Py_DECREF(m); + return NULL; + } + if (PyModule_AddObject(m, "openssl_md_meth_names", openssl_md_meth_names)) { + Py_DECREF(m); + return NULL; + } + #if HASH_OBJ_CONSTRUCTOR Py_INCREF(&EVPtype); PyModule_AddObject(m, "HASH", (PyObject *)&EVPtype);