Merge pull request #13 from cppalliance/sha1

Implement, test and fuzz SHA1

.drone.jsonnet

@@ -391,7 +391,15 @@ local windows_pipeline(name, image, environment, arch = "amd64") =
         "cppalliance/droneubuntu2404:1",
         { TOOLSET: 'clang', COMPILER: 'clang++-18', CXXSTD: '03,11,14,17,20,2b' },
         "clang-18",
-        ["deb http://apt.llvm.org/jammy/ llvm-toolchain-jammy-18 main"],
+        ["deb http://apt.llvm.org/noble/ llvm-toolchain-noble-18 main"],
+    ),
+
+    linux_pipeline(
+        "Linux 24.04 Clang 19",
+        "cppalliance/droneubuntu2404:1",
+        { TOOLSET: 'clang', COMPILER: 'clang++-19', CXXSTD: '03,11,14,17,20,2b' },
+        "clang-19",
+        ["deb http://apt.llvm.org/noble/ llvm-toolchain-noble-19 main"],
     ),
 
     macos_pipeline(

.github/workflows/fuzz.yml

@@ -86,6 +86,26 @@ jobs:
               - "deb http://apt.llvm.org/jammy/ llvm-toolchain-jammy-17 main"
             source_keys:
               - "https://apt.llvm.org/llvm-snapshot.gpg.key"
+          - toolset: clang
+            compiler: clang++-18
+            cxxstd: "14,17,20"
+            os: ubuntu-24.04
+            install:
+              - clang-18
+            sources:
+              - "deb http://apt.llvm.org/noble/ llvm-toolchain-noble-18 main"
+            source_keys:
+              - "https://apt.llvm.org/llvm-snapshot.gpg.key"
+          - toolset: clang
+            compiler: clang++-19
+            cxxstd: "14,17,20"
+            os: ubuntu-24.04
+            install:
+              - clang-19
+            sources:
+              - "deb http://apt.llvm.org/noble/ llvm-toolchain-noble-19 main"
+            source_keys:
+              - "https://apt.llvm.org/llvm-snapshot.gpg.key"
 
 
     timeout-minutes: 60

doc/crypt.adoc

@@ -21,6 +21,8 @@ include::crypt/api_reference.adoc[]
 
 include::crypt/md5.adoc[]
 
+include::crypt/sha1.adoc[]
+
 include::crypt/config.adoc[]
 
 include::crypt/reference.adoc[]

doc/crypt/reference.adoc

@@ -13,3 +13,5 @@ The following books, papers and blog posts serve as the basis for the algorithms
 :linkattrs:
 
 - Ronald L. Rivest, https://www.ietf.org/rfc/rfc1321.txt[RFC 1321:  The MD5 Message-Digest Algorithm], 1992
+
+- Donald E. Eastlake and Paul E. Jones, https://datatracker.ietf.org/doc/html/rfc3174[RFC 3174:  US Secure Hash Algorithm 1 (SHA1)], 2001

doc/crypt/sha1.adoc

@@ -0,0 +1,115 @@
+////
+Copyright 2024 Matt Borland
+Distributed under the Boost Software License, Version 1.0.
+https://www.boost.org/LICENSE_1_0.txt
+////
+
+[#sha1]
+:idprefix: sha1_
+
+= SHA1
+
+This library supports SHA1 as described in https://datatracker.ietf.org/doc/html/rfc3174[RFC 3174].
+There is a wide range of acceptable inputs for the base sha1 function:
+
+== Hashing Functions
+
+[source, c++]
+----
+namespace boost {
+namespace crypt {
+
+uisng return_type = boost::crypt::array<uint8_t, 20>;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char* str) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char* str, size_t len) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const unsigned char* str) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const unsigned char* str, size_t len) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char16_t* str) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char16_t* str, size_t len) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char32_t* str) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const char32_t* str, size_t len) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const wchar_t* str) noexcept -> return_type;
+
+BOOST_CRYPT_GPU_ENABLED constexpr auto sha1(const wchar_t* str, size_t len) noexcept -> return_type;
+
+inline auto sha1(const std::string& str) noexcept -> return_type;
+
+inline auto sha1(const std::u16string& str) noexcept -> return_type;
+
+inline auto sha1(const std::u32string& str) noexcept -> return_type;
+
+inline auto sha1(const std::wstring& str) noexcept -> return_type;
+
+#ifdef BOOST_CRYPT_HAS_STRING_VIEW
+
+inline auto sha1(std::string_view str) noexcept -> return_type;
+
+inline auto sha1(std::u16string_view str) noexcept -> return_type;
+
+inline auto sha1(std::u32string_view str) noexcept -> return_type;
+
+inline auto sha1(std::wstring_view str) noexcept -> return_type;
+
+#endif // BOOST_CRYPT_HAS_STRING_VIEW
+
+} //namespace crypt
+} //namespace boost
+----
+
+== File Hashing Functions
+
+We also have the ability to scan files and return the sha1 value:
+
+[source, c++]
+----
+namespace boost {
+namespace crypt {
+
+uisng return_type = boost::crypt::array<uint8_t, 16>;
+
+inline auto sha1_file(const char* filepath) noexcept -> return_type;
+
+inline auto sha1_file(const std::string& filepath) noexcept -> return_type;
+
+inline auto sha1_file(std::string_view filepath) noexcept -> return_type;
+
+} // namespace crypt
+} // namespace boost
+----
+
+== Hashing Object
+
+[#sha1_hasher]
+Lastly, there is also the ability to create a sha1 hashing object and feed it bytes as the user parses them.
+This class does not use any dynamic memory allocation.
+
+[source, c++]
+----
+namespace boost {
+namespace crypt {
+
+class sha1_hasher
+{
+    init();
+
+    template <typename ByteType>
+    BOOST_CRYPT_GPU_ENABLED constexpr auto process_byte(ByteType byte) noexcept -> void;
+
+    template <typename ForwardIter>
+    BOOST_CRYPT_GPU_ENABLED constexpr auto process_bytes(ForwardIter buffer, size_t byte_count) noexcept -> void;
+
+    constexpr auto get_digest() noexcept -> boost::crypt::array<boost::crypt::uint8_t, 20>;
+};
+
+} // namespace crypt
+} // namespace boost
+----

fuzzing/Jamfile

@@ -15,9 +15,7 @@ local all_fuzzers = [ regex.replace-list
 
 for local fuzzer in $(all_fuzzers)
 {
-    # These two fuzzers are the most complex ones. The rest are really
-    # simple, so less time is enough
-    local fuzz_time = 30 ;
+    local fuzz_time = 120 ;
 
     # Create the output corpus directories
     make /tmp/corpus/$(fuzzer) : : common.MkDir ;

fuzzing/fuzz_sha1.cpp

@@ -0,0 +1,33 @@
+// Copyright 2024 Matt Borland
+// Distributed under the Boost Software License, Version 1.0.
+// https://www.boost.org/LICENSE_1_0.txt
+
+#include <boost/crypt/hash/sha1.hpp>
+#include <iostream>
+#include <exception>
+#include <string>
+
+extern "C" int LLVMFuzzerTestOneInput(const std::uint8_t* data, std::size_t size)
+{
+    try
+    {
+        auto c_data = reinterpret_cast<const char*>(data);
+        std::string c_data_str {c_data, size}; // Guarantee null termination since we can't pass the size argument
+
+        boost::crypt::sha1(c_data_str);
+        boost::crypt::sha1(c_data, size);
+        boost::crypt::sha1(data, size);
+
+        #ifdef BOOST_CRYPT_HAS_STRING_VIEW
+        std::string_view view {c_data_str};
+        boost::crypt::sha1(view);
+        #endif
+    }
+    catch(...)
+    {
+        std::cerr << "Error with: " << data << std::endl;
+        std::terminate();
+    }
+
+    return 0;
+}

fuzzing/seedcorpus/fuzz_sha1/sha1.txt

@@ -0,0 +1,18 @@
+"The quick brown fox jumps over the lazy dog"
+"The quick brown fox jumps over the lazy dog."
+""
+"aB3$x9Yz"
+"12345"
+"!@#$%^&*()"
+"FuzzTest123"
+" "
+"Lorem ipsum dolor sit amet"
+"a"
+"9876543210"
+"ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+"ñÑáéíóúÁÉÍÓÚ"
+"\n\r\t"
+"0"
+"ThisIsAVeryLongStringWithNoSpacesOrPunctuationToTestEdgeCases"
+"<?php echo 'test'; ?>"
+"SELECT * FROM users;"

include/boost/crypt/hash/hasher_state.hpp

@@ -0,0 +1,24 @@
+// Copyright 2024 Matt Borland
+// Distributed under the Boost Software License, Version 1.0.
+// https://www.boost.org/LICENSE_1_0.txt
+
+#ifndef BOOST_CRYPT_HASH_HASHER_STATE_HPP
+#define BOOST_CRYPT_HASH_HASHER_STATE_HPP
+
+#include <boost/crypt/utility/cstdint.hpp>
+
+namespace boost {
+namespace crypt {
+
+enum class hasher_state : boost::crypt::uint8_t
+{
+    success,            // no issues
+    null,               // nullptr as parameter
+    input_too_long,     // input data too long (exceeded size_t)
+    state_error         // added more input after get_digest without re-init
+};
+
+} // namespace crypt
+} // namespace boost
+
+#endif // BOOST_CRYPT_HASH_HASHER_STATE_HPP