LibCompress: Implement gzip.

4 files changed, 314 insertions, 1 deletions

diff --git a/Libraries/LibCompress/CMakeLists.txt b/Libraries/LibCompress/CMakeLists.txt
index a61ff5a486..f76ba8d5b5 100644
--- a/Libraries/LibCompress/CMakeLists.txt
+++ b/Libraries/LibCompress/CMakeLists.txt
@@ -1,7 +1,8 @@
 set(SOURCES
     Deflate.cpp
     Zlib.cpp
+    Gzip.cpp
 )
 
 serenity_lib(LibCompress compression)
-target_link_libraries(LibCompress LibC)
+target_link_libraries(LibCompress LibC LibCrypto)
diff --git a/Libraries/LibCompress/Gzip.cpp b/Libraries/LibCompress/Gzip.cpp
new file mode 100644
index 0000000000..3e7170342c
--- /dev/null
+++ b/Libraries/LibCompress/Gzip.cpp
@@ -0,0 +1,186 @@
+/*
+ * Copyright (c) 2020, the SerenityOS developers.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice, this
+ *    list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <LibCompress/Gzip.h>
+
+#include <AK/String.h>
+
+namespace Compress {
+
+bool GzipDecompressor::BlockHeader::valid_magic_number() const
+{
+    return identification_1 == 0x1f && identification_2 == 0x8b;
+}
+
+bool GzipDecompressor::BlockHeader::supported_by_implementation() const
+{
+    if (compression_method != 0x08) {
+        // RFC 1952 does not define any compression methods other than deflate.
+        return false;
+    }
+
+    if (flags > Flags::MAX) {
+        // RFC 1952 does not define any more flags.
+        return false;
+    }
+
+    if (flags & Flags::FHCRC) {
+        TODO();
+    }
+
+    return true;
+}
+
+GzipDecompressor::GzipDecompressor(InputStream& stream)
+    : m_input_stream(stream)
+{
+}
+
+GzipDecompressor::~GzipDecompressor()
+{
+    m_current_member.clear();
+}
+
+// FIXME: Again, there are surely a ton of bugs because the code doesn't check for read errors.
+size_t GzipDecompressor::read(Bytes bytes)
+{
+    if (m_current_member.has_value()) {
+        size_t nread = current_member().m_stream.read(bytes);
+        current_member().m_checksum.update(bytes.trim(nread));
+        current_member().m_nread += nread;
+
+        if (nread < bytes.size()) {
+            LittleEndian<u32> crc32, input_size;
+            m_input_stream >> crc32 >> input_size;
+
+            if (crc32 != current_member().m_checksum.digest()) {
+                m_error = true;
+                return 0;
+            }
+
+            if (input_size != current_member().m_nread) {
+                m_error = true;
+                return 0;
+            }
+
+            m_current_member.clear();
+
+            return nread + read(bytes.slice(nread));
+        }
+
+        return nread;
+    } else {
+        if (m_input_stream.eof())
+            return 0;
+
+        // FIXME: This fails with the new changes?
+        BlockHeader header;
+        m_input_stream >> Bytes { &header, sizeof(header) };
+
+        if (!header.valid_magic_number() || !header.supported_by_implementation()) {
+            m_error = true;
+            return 0;
+        }
+
+        if (header.flags & Flags::FEXTRA) {
+            LittleEndian<u16> subfield_id, length;
+            m_input_stream >> subfield_id >> length;
+            m_input_stream.discard_or_error(length);
+        }
+
+        if (header.flags & Flags::FNAME) {
+            String original_filename;
+            m_input_stream >> original_filename;
+        }
+
+        if (header.flags & Flags::FCOMMENT) {
+            String comment;
+            m_input_stream >> comment;
+        }
+
+        m_current_member.emplace(header, m_input_stream);
+        return read(bytes);
+    }
+}
+
+bool GzipDecompressor::read_or_error(Bytes bytes)
+{
+    if (read(bytes) < bytes.size()) {
+        m_error = true;
+        return false;
+    }
+
+    return true;
+}
+
+bool GzipDecompressor::discard_or_error(size_t count)
+{
+    u8 buffer[4096];
+
+    size_t ndiscarded = 0;
+    while (ndiscarded < count) {
+        if (eof()) {
+            m_error = true;
+            return false;
+        }
+
+        ndiscarded += read({ buffer, min<size_t>(count - ndiscarded, sizeof(buffer)) });
+    }
+
+    return true;
+}
+
+ByteBuffer GzipDecompressor::decompress_all(ReadonlyBytes bytes)
+{
+    InputMemoryStream memory_stream { bytes };
+    GzipDecompressor gzip_stream { memory_stream };
+
+    auto buffer = ByteBuffer::create_uninitialized(4096);
+
+    size_t nread = 0;
+    while (!gzip_stream.eof()) {
+        nread += gzip_stream.read(buffer.bytes().slice(nread));
+
+        if (buffer.size() - nread < 4096)
+            buffer.grow(buffer.size() + 4096);
+    }
+
+    buffer.trim(nread);
+    return buffer;
+}
+
+bool GzipDecompressor::eof() const
+{
+    if (m_current_member.has_value()) {
+        // FIXME: There is an ugly edge case where we read the whole deflate block
+        //        but haven't read CRC32 and ISIZE.
+        return current_member().m_stream.eof() && m_input_stream.eof();
+    } else {
+        return m_input_stream.eof();
+    }
+}
+
+}
diff --git a/Libraries/LibCompress/Gzip.h b/Libraries/LibCompress/Gzip.h
new file mode 100644
index 0000000000..72e5e49aba
--- /dev/null
+++ b/Libraries/LibCompress/Gzip.h
@@ -0,0 +1,92 @@
+/*
+ * Copyright (c) 2020, the SerenityOS developers.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice, this
+ *    list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#pragma once
+
+#include <LibCompress/Deflate.h>
+#include <LibCrypto/Checksum/CRC32.h>
+
+namespace Compress {
+
+class GzipDecompressor final : public InputStream {
+public:
+    GzipDecompressor(InputStream&);
+    ~GzipDecompressor();
+
+    size_t read(Bytes) override;
+    bool read_or_error(Bytes) override;
+    bool discard_or_error(size_t) override;
+    bool eof() const override;
+
+    static ByteBuffer decompress_all(ReadonlyBytes);
+
+private:
+    struct [[gnu::packed]] BlockHeader
+    {
+        u8 identification_1;
+        u8 identification_2;
+        u8 compression_method;
+        u8 flags;
+        LittleEndian<u32> modification_time;
+        u8 extra_flags;
+        u8 operating_system;
+
+        bool valid_magic_number() const;
+        bool supported_by_implementation() const;
+    };
+
+    struct Flags {
+        static constexpr u8 FTEXT = 1 << 0;
+        static constexpr u8 FHCRC = 1 << 1;
+        static constexpr u8 FEXTRA = 1 << 2;
+        static constexpr u8 FNAME = 1 << 3;
+        static constexpr u8 FCOMMENT = 1 << 4;
+
+        static constexpr u8 MAX = FTEXT | FHCRC | FEXTRA | FNAME | FCOMMENT;
+    };
+
+    class Member {
+    public:
+        Member(BlockHeader header, InputStream& stream)
+            : m_header(header)
+            , m_stream(stream)
+        {
+        }
+
+        BlockHeader m_header;
+        DeflateDecompressor m_stream;
+        Crypto::Checksum::CRC32 m_checksum;
+        size_t m_nread { 0 };
+    };
+
+    const Member& current_member() const { return m_current_member.value(); }
+    Member& current_member() { return m_current_member.value(); }
+
+    InputStream& m_input_stream;
+    Optional<Member> m_current_member;
+};
+
+}
diff --git a/Userland/test-compress.cpp b/Userland/test-compress.cpp
index 87473970c6..bf674fc4fd 100644
--- a/Userland/test-compress.cpp
+++ b/Userland/test-compress.cpp
@@ -27,6 +27,7 @@
 #include <AK/TestSuite.h>
 
 #include <LibCompress/Deflate.h>
+#include <LibCompress/Gzip.h>
 #include <LibCompress/Zlib.h>
 
 static bool compare(ReadonlyBytes lhs, ReadonlyBytes rhs)
@@ -119,4 +120,37 @@ TEST_CASE(zlib_decompress_simple)
     EXPECT(compare({ uncompressed, sizeof(uncompressed) - 1 }, decompressed.bytes()));
 }
 
+TEST_CASE(gzip_decompress_simple)
+{
+    const u8 compressed[] = {
+        0x1f, 0x8b, 0x08, 0x00, 0x77, 0xff, 0x47, 0x5f, 0x02, 0xff, 0x2b, 0xcf,
+        0x2f, 0x4a, 0x31, 0x54, 0x48, 0x4c, 0x4a, 0x56, 0x28, 0x07, 0xb2, 0x8c,
+        0x00, 0xc2, 0x1d, 0x22, 0x15, 0x0f, 0x00, 0x00, 0x00
+    };
+
+    const u8 uncompressed[] = "word1 abc word2";
+
+    const auto decompressed = Compress::GzipDecompressor::decompress_all({ compressed, sizeof(compressed) });
+
+    EXPECT(compare({ uncompressed, sizeof(uncompressed) - 1 }, decompressed.bytes()));
+}
+
+TEST_CASE(gzip_multiple_members)
+{
+
+    const u8 compressed[] = {
+        0x1f, 0x8b, 0x08, 0x00, 0xe0, 0x03, 0x48, 0x5f, 0x02, 0xff, 0x4b, 0x4c,
+        0x4a, 0x4e, 0x4c, 0x4a, 0x06, 0x00, 0x4c, 0x99, 0x6e, 0x72, 0x06, 0x00,
+        0x00, 0x00, 0x1f, 0x8b, 0x08, 0x00, 0xe0, 0x03, 0x48, 0x5f, 0x02, 0xff,
+        0x4b, 0x4c, 0x4a, 0x4e, 0x4c, 0x4a, 0x06, 0x00, 0x4c, 0x99, 0x6e, 0x72,
+        0x06, 0x00, 0x00, 0x00
+    };
+
+    const u8 uncompressed[] = "abcabcabcabc";
+
+    const auto decompressed = Compress::GzipDecompressor::decompress_all({ compressed, sizeof(compressed) });
+
+    EXPECT(compare({ uncompressed, sizeof(uncompressed) - 1 }, decompressed.bytes()));
+}
+
 TEST_MAIN(Compress)