WINTERMUTE: Add a DCP extractor script

engines/wintermute/dcp_extractor.py

@@ -0,0 +1,154 @@
+#! /usr/bin/env python3
+
+# Sources:
+# https://archive.softwareheritage.org/browse/content/sha1_git:0bc8340ae58e9d88a23e52b0723b9a92e14f4e62/?origin_url=https://bitbucket.org/MnemonicWME/wme1&path=src/engine_core/wme_base/dcpackage.h
+# https://archive.softwareheritage.org/browse/content/sha1_git:4fd95c6076f4f2ce5dbbd1eead41b357bd232c66/?origin_url=https://bitbucket.org/MnemonicWME/wme1&path=src/engine_core/wme_base/BFileManager.cpp
+
+import collections
+from datetime import datetime
+import functools
+import pathlib
+import struct
+import zlib
+
+Header = collections.namedtuple('Header', ['magic1', 'magic2', 'pkg_version', 'game_version', 'priority', 'cd', 'master_index', 'creation_time', 'desc', 'num_dirs'])
+DirEntry = collections.namedtuple('DirEntry', ['name', 'cd', 'num_entries', 'files'])
+FileEntry = collections.namedtuple('FileEntry', ['name', 'offset', 'length', 'comp_length', 'flags', 'timedate1', 'timedate2'], defaults=(0, 0))
+
+def read_struct(f, fmt, constructor):
+    if type(fmt) is str:
+        fmt = struct.Struct(fmt)
+
+    buf = f.read(fmt.size)
+    if len(buf) != fmt.size:
+        raise Exception("File too small")
+
+    return constructor(*fmt.unpack(buf))
+
+def read_str(f):
+    sz = f.read(1)
+    if len(sz) != 1:
+        raise Exception("File too small")
+    sz, = struct.unpack('<B', sz)
+    s = f.read(sz)
+    if len(s) != sz:
+        raise Exception("File too small")
+    return s
+
+def read_headers(f, abs_offset = 0):
+    f.seek(abs_offset)
+
+    header = read_struct(f, '<L4sLLBBBxL100sL', Header)
+    if header.magic1 != 0xdec0adde:
+        raise Exception("Invalid magic")
+    if header.magic2 != b'JUNK':
+        raise Exception("Invalid magic")
+    if header.pkg_version > 0x200:
+        raise Exception("Invalid version")
+
+    if header.pkg_version == 0x200:
+        dir_offset, = struct.unpack('<L', f.read(4))
+        dir_offset += abs_offset
+        f.seek(dir_offset)
+
+    dirs = []
+    for pkg in range(header.num_dirs):
+        files = []
+        dir_name = read_str(f)
+        dir_name = dir_name.rstrip(b'\x00')
+        dirent = read_struct(f, '<BL', functools.partial(DirEntry, dir_name, files=files))
+        dirs.append(dirent)
+
+        for i in range(dirent.num_entries):
+            fname = read_str(f)
+            fname = bytes(b ^ 0x44 for b in fname)
+            fname = fname.rstrip(b'\x00')
+
+            if header.pkg_version == 0x200:
+                fmt = '<LLLLLL'
+            else:
+                fmt = '<LLLL'
+            fileent = read_struct(f, fmt, functools.partial(FileEntry, fname))
+            fileent = fileent._replace(offset=fileent.offset + abs_offset)
+            files.append(fileent)
+
+    return header, dirs
+
+def read_file(f, fileent):
+    f.seek(fileent.offset)
+    if fileent.comp_length:
+        buf = f.read(fileent.comp_length)
+        buf = zlib.decompress(buf)
+    else:
+        buf = f.read(fileent.length)
+    if len(buf) != fileent.length:
+        raise Exception("Invalid file size")
+    return buf
+
+def lookup_sig(f):
+    import mmap
+    with mmap.mmap(f.fileno(), 0, access=mmap.ACCESS_READ) as mm:
+        offset = mm.find(b'\xde\xad\xc0\xdeJUNK')
+    if offset == -1:
+        raise Exception("Signature not found")
+    return offset
+
+def dcp_list(options, offset=0):
+    header, dirs = read_headers(options.input, offset)
+    for dirent in dirs:
+        print("Directory {0} from CD {1} with {2} entries".format(dirent.name.decode('utf-8'), dirent.cd, dirent.num_entries))
+        print("{0:<8}\t@{1:<8}\t{2:<8}\t{3:<19}\t{4}".format('sz', 'offset  ', 'compsz', 'date', 'name'))
+        for fl in dirent.files:
+            print("{0:<8}\t@{1:<8}\t{2:<8}\t{3}\t{4}".format(
+                fl.length, fl.offset, fl.length if fl.comp_length == 0 else fl.comp_length,
+                datetime.fromtimestamp(fl.timedate1 | (fl.timedate2 << 64)).isoformat(), fl.name.decode('utf-8')))
+
+def dcp_extract(options, offset=0):
+    header, dirs = read_headers(options.input, offset)
+
+    output_dir = options.output_dir
+    for dirent in dirs:
+        print("Directory {0} from CD {1} with {2} entries".format(dirent.name.decode('utf-8'), dirent.cd, dirent.num_entries))
+        print("{0:<8}\t@{1:<8}\t{2:<8}\t{3:<19}\t{4}".format('sz', 'offset  ', 'compsz', 'date', 'name'))
+        output_int = output_dir / dirent.name.decode('utf-8')
+        for fl in dirent.files:
+            print("{0:<8}\t@{1:<8}\t{2:<8}\t{3}\t{4}".format(
+                fl.length, fl.offset, fl.length if fl.comp_length == 0 else fl.comp_length,
+                datetime.fromtimestamp(fl.timedate1 | (fl.timedate2 << 64)).isoformat(), fl.name.decode('utf-8')))
+            output_file = output_int / pathlib.Path(fl.name.decode('utf-8').replace('\\', '/'))
+            output_file.parent.mkdir(parents=True, exist_ok=True)
+
+            with output_file.open('wb') as output_f:
+                buf = read_file(options.input, fl)
+                output_f.write(buf)
+
+def main():
+    import argparse
+
+    parser = argparse.ArgumentParser(
+        prog='dcp_extractor.py',
+        description='Wintermute DCP archive extractor')
+
+    parser.add_argument('--sfx',
+        action='store_true')
+
+    action_parsers = parser.add_subparsers(required=True)
+    list_parser =  action_parsers.add_parser('list', help='list archive contents')
+    list_parser.add_argument('input', type=argparse.FileType('rb'), metavar='dcp file')
+    list_parser.set_defaults(action=dcp_list)
+    extract_parser =  action_parsers.add_parser('extract', help='extract archive contents')
+    extract_parser.add_argument('input', type=argparse.FileType('rb'), metavar='dcp file')
+    extract_parser.add_argument('output_dir', type=pathlib.Path, metavar='output directory')
+    extract_parser.set_defaults(action=dcp_extract)
+
+    options = parser.parse_args()
+
+    offset = 0
+    if options.sfx:
+        offset = lookup_sig(options.input)
+        print("Found signature at {}".format(offset))
+
+    options.action(options, offset=offset)
+
+if __name__ == "__main__":
+    main()