Commit f85c01d4c7520d2242626f4c0684ab97e47af373 | zig

author	FnControlOption <fncontroloption@noreply.codeberg.org> 2023-01-21 15:47:31-08:00
committer	Andrew Kelley <andrew@ziglang.org> 2023-01-22 00:33:29-05:00
log	f85c01d4c7520d2242626f4c0684ab97e47af373
tree	9385b5c065a3e90449dc1e02883d579799dca1fe
parent	562d52e23dce703b0eb45a212522588274afbd38

Implement gzip header CRC check.

From RFC 1952:

> If FHCRC is set, a CRC16 for the gzip header is present,
> immediately before the compressed data. The CRC16 consists
> of the two least significant bytes of the CRC32 for all
> bytes of the gzip header up to and not including the CRC16.

2 files changed, 54 insertions(+), 8 deletions(-)

lib
- std
  - compress.zig
  - compress
    gzip.zig

lib/std/compress.zig+30

...	@@ -4,6 +4,36 @@ pub const deflate = @import("compress/deflate.zig");	...	@@ -4,6 +4,36 @@ pub const deflate = @import("compress/deflate.zig");
4	pub const gzip = @import("compress/gzip.zig");	4	pub const gzip = @import("compress/gzip.zig");
5	pub const zlib = @import("compress/zlib.zig");	5	pub const zlib = @import("compress/zlib.zig");
6		6
		7	pub fn HashedReader(
		8	comptime ReaderType: anytype,
		9	comptime HasherType: anytype,
		10	) type {
		11	return struct {
		12	child_reader: ReaderType,
		13	hasher: HasherType,
		14
		15	pub const Error = ReaderType.Error;
		16	pub const Reader = std.io.Reader(*@This(), Error, read);
		17
		18	pub fn read(self: *@This(), buf: []u8) Error!usize {
		19	const amt = try self.child_reader.read(buf);
		20	self.hasher.update(buf);
		21	return amt;
		22	}
		23
		24	pub fn reader(self: *@This()) Reader {
		25	return .{ .context = self };
		26	}
		27	};
		28	}
		29
		30	pub fn hashedReader(
		31	reader: anytype,
		32	hasher: anytype,
		33	) HashedReader(@TypeOf(reader), @TypeOf(hasher)) {
		34	return .{ .child_reader = reader, .hasher = hasher };
		35	}
		36
7	test {	37	test {
8	_ = deflate;	38	_ = deflate;
9	_ = gzip;	39	_ = gzip;

lib/std/compress/gzip.zig+24-8

 @@ -44,8 +44,11 @@ pub fn GzipStream(comptime ReaderType: type) type {
         },
         fn init(allocator: mem.Allocator, source: ReaderType) !Self {
+            var hasher = std.compress.hashedReader(source, std.hash.Crc32.init());
+            const hashed_reader = hasher.reader();
             // gzip header format is specified in RFC1952
-            const header = try source.readBytesNoEof(10);
+            const header = try hashed_reader.readBytesNoEof(10);
             // Check the ID1/ID2 fields
             if (header[0] != 0x1f or header[1] != 0x8b)
 @@ -66,31 +69,31 @@ pub fn GzipStream(comptime ReaderType: type) type {
             _ = XFL;
             const extra = if (FLG & FEXTRA != 0) blk: {
-                const len = try source.readIntLittle(u16);
+                const len = try hashed_reader.readIntLittle(u16);
                 const tmp_buf = try allocator.alloc(u8, len);
                 errdefer allocator.free(tmp_buf);
-                try source.readNoEof(tmp_buf);
+                try hashed_reader.readNoEof(tmp_buf);
                 break :blk tmp_buf;
             } else null;
             errdefer if (extra) |p| allocator.free(p);
             const filename = if (FLG & FNAME != 0)
-                try source.readUntilDelimiterAlloc(allocator, 0, max_string_len)
+                try hashed_reader.readUntilDelimiterAlloc(allocator, 0, max_string_len)
             else
                 null;
             errdefer if (filename) |p| allocator.free(p);
             const comment = if (FLG & FCOMMENT != 0)
-                try source.readUntilDelimiterAlloc(allocator, 0, max_string_len)
+                try hashed_reader.readUntilDelimiterAlloc(allocator, 0, max_string_len)
             else
                 null;
             errdefer if (comment) |p| allocator.free(p);
             if (FLG & FHCRC != 0) {
-                // TODO: Evaluate and check the header checksum. The stdlib has
+                const hash = try source.readIntLittle(u16);
-                // no CRC16 yet :(
+                if (hash != @truncate(u16, hasher.hasher.final()))
-                _ = try source.readIntLittle(u16);
+                    return error.WrongChecksum;
             }
             return Self{
 @@ -230,3 +233,16 @@ test "sanity checks" {
         }, ""),
     );
 }
+test "header checksum" {
+    try testReader(&[_]u8{
+        // GZIP header
+        0x1f, 0x8b, 0x08, 0x12, 0x00, 0x09, 0x6e, 0x88, 0x00, 0xff, 0x48, 0x65, 0x6c, 0x6c, 0x6f, 0x00,
+        // header.FHCRC (should cover entire header)
+        0x99, 0xd6,
+        // GZIP data
+        0x01, 0x00, 0x00, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    }, "");
+}