deblob

commit: 33b5cab3c8d3dc208cec7bfe8da85e790cf9c611
parent 9f519629b20042a02e547c310d95d958776573f0
Author: Haelwenn (lanodan) Monnier <contact@hacktivis.me>
Date:   Mon,  1 Aug 2022 10:12:36 +0200

Add support for Java class files

Changed the way magic headers are stored because "\xCA\xFE\xBA\xBE"
corresponds to [U+00CA, U+00FE, U+00BA, U+00BE].
And the real 0xCAFEBABE isn't valid UTF-8, which it shouldn't even need to be.

FirstRepeatingElement.class is from <https://bae.st/objects/873a9ede-7995-4c38-b64a-0d49d2c0e98c>

Diffstat:
A
LICENSES/WTFPL.txt
11
+++++++++++
M
main.ha
22
++++++++++++----------
A
test/fixtures/FirstRepeatingElement.class
0

A
test/fixtures/FirstRepeatingElement.class.license
2
++

4 files changed, 25 insertions(+), 10 deletions(-)
diff --git a/LICENSES/WTFPL.txt b/LICENSES/WTFPL.txt
@@ -0,0 +1,11 @@
+DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE
+Version 2, December 2004
+
+Copyright (C) 2004 Sam Hocevar <sam@hocevar.net>
+
+Everyone is permitted to copy and distribute verbatim or modified copies of this license document, and changing it is allowed as long as the name is changed.
+
+DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE
+TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. You just DO WHAT THE FUCK YOU WANT TO.
diff --git a/main.ha b/main.ha
@@ -14,12 +14,13 @@ use strings;
 let excludes: []str = [];
 let noop: bool = false;
 
-def ELF: str = "\x7FELF";
-def UNIX_AR: str = "!<arch>\n";
-def PC_BIOS_ROM: str = "\x55\xAA";
-def ERL_BEAM: str = "FOR1";
-
-const magic: [_]str = [ELF, UNIX_AR, PC_BIOS_ROM, ERL_BEAM];
+let magic: [_][]u8 = [
+	[0x7F, 'E', 'L', 'F'], // ELF
+	['!', '<', 'a', 'r', 'c', 'h', '>', '\n'], // Unix ar(1)
+	[0x55, 0xAA], // IBM PC BIOS ROM
+	['F', 'O', 'R', '1'], // Erlang BEAM
+	[0xCA, 0xFE, 0xBA, 0xBE], // Java Class File & Mach-O Executable
+];
 
 fn is_blob(filename: str) (bool | fs::error | io::error) = {
 	static let buffer: [512]u8 = [0...];
@@ -33,8 +34,8 @@ fn is_blob(filename: str) (bool | fs::error | io::error) = {
 	};
 
 	for (let i = 0z; i < len(magic); i += 1) {
-		const magic = strings::toutf8(magic[i]);
-		if (bytes::hasprefix(buffer, magic)) {
+		assert(len(magic[i]) > 0);
+		if (bytes::hasprefix(buffer, magic[i])) {
 			return true;
 		};
 	};
@@ -51,6 +52,7 @@ fn is_blob(filename: str) (bool | fs::error | io::error) = {
 		("test/fixtures/hello.a", true),
 		("test/fixtures/Elixir.Hex.API.Auth.beam", true),
 		//("test/fixtures/option.rom", true),
+		("test/fixtures/FirstRepeatingElement.class", true),
 		("test/fixtures/empty", false),
 	];
 
@@ -152,7 +154,7 @@ fn check_dir(dirname: str) (void | errors::invalid) = {
 	case let e: fs::error =>
 		fmt::fatalf("os::readdir({}): {}", dirname, fs::strerror(e));
 	};
-	assert(len(files_before) == 10);
+	assert(len(files_before) == 12);
 
 	const ret = check_dir(dirname);
 	assert(ret is void);
@@ -163,7 +165,7 @@ fn check_dir(dirname: str) (void | errors::invalid) = {
 	case let e: fs::error =>
 		fmt::fatalf("os::readdir({}): {}", dirname, fs::strerror(e));
 	};
-	assert(len(files_after) == 6);
+	assert(len(files_after) == 7);
 };
 
 export fn main() void = {
diff --git a/test/fixtures/FirstRepeatingElement.class b/test/fixtures/FirstRepeatingElement.class
Binary files differ.
diff --git a/test/fixtures/FirstRepeatingElement.class.license b/test/fixtures/FirstRepeatingElement.class.license
@@ -0,0 +1,2 @@
+SPDX-FileCopyrightText: 2022 xianc78 <https://bae.st/users/xianc78>
+SPDX-License-Identifier: WTFPL