|
| 1 | +#!/usr/bin/env python |
| 2 | +# -*- coding: UTF-8 -*- |
| 3 | +# pyneofile_light.py - lightweight pack/unpack/list with Py2/3 compatibility. |
| 4 | + |
| 5 | +import os, sys, io, json, time, struct |
| 6 | + |
| 7 | +PY2 = (sys.version_info[0] == 2) |
| 8 | + |
| 9 | +try: |
| 10 | + basestring |
| 11 | +except NameError: |
| 12 | + basestring = (str,) |
| 13 | + |
| 14 | +try: |
| 15 | + unicode # Py2 |
| 16 | +except NameError: |
| 17 | + unicode = str |
| 18 | + |
| 19 | +def to_text(s, encoding="utf-8", errors="strict"): |
| 20 | + if s is None: |
| 21 | + return u"" |
| 22 | + if isinstance(s, unicode): |
| 23 | + return s |
| 24 | + if isinstance(s, (bytes, bytearray)): |
| 25 | + return s.decode(encoding, errors) |
| 26 | + return unicode(s) |
| 27 | + |
| 28 | +def to_bytes(s, encoding="utf-8", errors="strict"): |
| 29 | + if s is None: |
| 30 | + return b"" |
| 31 | + if isinstance(s, (bytes, bytearray)): |
| 32 | + return bytes(s) |
| 33 | + return to_text(s, encoding, errors).encode(encoding, errors) |
| 34 | + |
| 35 | +def normpath(p): |
| 36 | + s = to_text(p) |
| 37 | + s = s.replace("\\\\", "/") |
| 38 | + if "://" not in s: |
| 39 | + while "//" in s: |
| 40 | + s = s.replace("//", "/") |
| 41 | + return s.rstrip("/") or s |
| 42 | + |
| 43 | +def ensure_dir(path): |
| 44 | + if not path: |
| 45 | + return |
| 46 | + if not os.path.isdir(path): |
| 47 | + os.makedirs(path) |
| 48 | + |
| 49 | +_zlib = None |
| 50 | +_gzip = None |
| 51 | +_bz2 = None |
| 52 | +_lzma = None |
| 53 | +HAVE_LZMA = False |
| 54 | + |
| 55 | +def _init_backends(): |
| 56 | + global _zlib, _gzip, _bz2, _lzma, HAVE_LZMA |
| 57 | + if _zlib is None: |
| 58 | + import zlib as _z |
| 59 | + _zlib = _z |
| 60 | + if _gzip is None: |
| 61 | + import gzip as _g |
| 62 | + _gzip = _g |
| 63 | + if _bz2 is None: |
| 64 | + import bz2 as _b |
| 65 | + _bz2 = _b |
| 66 | + if _lzma is None: |
| 67 | + try: |
| 68 | + import lzma as _l |
| 69 | + _lzma = _l |
| 70 | + HAVE_LZMA = True |
| 71 | + except Exception: |
| 72 | + _lzma = None |
| 73 | + HAVE_LZMA = False |
| 74 | + |
| 75 | +COMP_NONE = 0 |
| 76 | +COMP_ZLIB = 1 |
| 77 | +COMP_GZIP = 2 |
| 78 | +COMP_BZ2 = 3 |
| 79 | +COMP_XZ = 4 |
| 80 | + |
| 81 | +def compress(data, algo): |
| 82 | + _init_backends() |
| 83 | + if algo == COMP_NONE: |
| 84 | + return data |
| 85 | + if algo == COMP_ZLIB: |
| 86 | + return _zlib.compress(data, 6) |
| 87 | + if algo == COMP_GZIP: |
| 88 | + buf = io.BytesIO() |
| 89 | + gz = _gzip.GzipFile(fileobj=buf, mode='wb', compresslevel=6) |
| 90 | + try: |
| 91 | + gz.write(data) |
| 92 | + finally: |
| 93 | + gz.close() |
| 94 | + return buf.getvalue() |
| 95 | + if algo == COMP_BZ2: |
| 96 | + return _bz2.compress(data, 6) if hasattr(_bz2, "compress") else _bz2.BZ2Compressor(6).compress(data) |
| 97 | + if algo == COMP_XZ: |
| 98 | + if not HAVE_LZMA: |
| 99 | + raise RuntimeError("xz/lzma not available on this Python") |
| 100 | + return _lzma.compress(data, preset=6) |
| 101 | + raise ValueError("Unknown compression algorithm: %r" % algo) |
| 102 | + |
| 103 | +def decompress(data, algo): |
| 104 | + _init_backends() |
| 105 | + if algo == COMP_NONE: |
| 106 | + return data |
| 107 | + if algo == COMP_ZLIB: |
| 108 | + return _zlib.decompress(data) |
| 109 | + if algo == COMP_GZIP: |
| 110 | + buf = io.BytesIO(data) |
| 111 | + gz = _gzip.GzipFile(fileobj=buf, mode='rb') |
| 112 | + try: |
| 113 | + return gz.read() |
| 114 | + finally: |
| 115 | + gz.close() |
| 116 | + if algo == COMP_BZ2: |
| 117 | + return _bz2.decompress(data) |
| 118 | + if algo == COMP_XZ: |
| 119 | + if not HAVE_LZMA: |
| 120 | + raise RuntimeError("xz/lzma not available on this Python") |
| 121 | + return _lzma.decompress(data) |
| 122 | + raise ValueError("Unknown compression algorithm: %r" % algo) |
| 123 | + |
| 124 | +AUTO_XZ_MIN = 2 * 1024 * 1024 |
| 125 | +AUTO_BZ2_MIN = 256 * 1024 |
| 126 | +AUTO_ZLIB_MIN = 16 * 1024 |
| 127 | + |
| 128 | +def pick_algo(total_size): |
| 129 | + _init_backends() |
| 130 | + if total_size >= AUTO_XZ_MIN and HAVE_LZMA: |
| 131 | + return COMP_XZ |
| 132 | + if total_size >= AUTO_BZ2_MIN: |
| 133 | + return COMP_BZ2 |
| 134 | + if total_size >= AUTO_ZLIB_MIN: |
| 135 | + return COMP_ZLIB |
| 136 | + return COMP_NONE |
| 137 | + |
| 138 | +MAGIC = b"PNF1" |
| 139 | + |
| 140 | +def _iter_files(sources): |
| 141 | + for src in sources: |
| 142 | + src = to_text(src) |
| 143 | + if os.path.isdir(src): |
| 144 | + base = src |
| 145 | + for root, _, files in os.walk(src): |
| 146 | + for f in files: |
| 147 | + p = os.path.join(root, f) |
| 148 | + st = os.stat(p) |
| 149 | + name = normpath(os.path.relpath(p, base)) |
| 150 | + yield name, p, st.st_size, int(st.st_mtime) |
| 151 | + else: |
| 152 | + st = os.stat(src) |
| 153 | + name = normpath(os.path.basename(src)) |
| 154 | + yield name, src, st.st_size, int(st.st_mtime) |
| 155 | + |
| 156 | +def pack(sources, outfile, compression="auto"): |
| 157 | + assert isinstance(sources, (list, tuple)) and sources, "sources must be a non-empty list" |
| 158 | + algo_map = { |
| 159 | + 'none': COMP_NONE, 'zlib': COMP_ZLIB, 'gzip': COMP_GZIP, |
| 160 | + 'bz2': COMP_BZ2, 'xz': COMP_XZ, 'auto': 'auto', |
| 161 | + } |
| 162 | + if compression not in algo_map: |
| 163 | + raise ValueError("Unknown compression: %r" % compression) |
| 164 | + entries = list(_iter_files(sources)) |
| 165 | + total_size = sum(s for _,__,s,___ in entries) |
| 166 | + algo = pick_algo(total_size) if compression == 'auto' else algo_map[compression] |
| 167 | + |
| 168 | + manifest = {"files": [{"name": n, "size": s, "mtime": m} for (n, p, s, m) in entries]} |
| 169 | + hdr = to_bytes(json.dumps(manifest, sort_keys=True, separators=(',', ':'))) |
| 170 | + flags = 0 if algo == COMP_NONE else 1 |
| 171 | + header = MAGIC + struct.pack(">BBI", flags, algo, len(hdr)) + hdr |
| 172 | + |
| 173 | + if outfile == "-" or outfile is None: |
| 174 | + out = sys.stdout if outfile == "-" else sys.stdout |
| 175 | + stream = getattr(out, "buffer", out) |
| 176 | + stream.write(header) |
| 177 | + raw = io.BytesIO() |
| 178 | + for name, path, size, mtime in entries: |
| 179 | + with open(path, "rb") as f: |
| 180 | + while True: |
| 181 | + chunk = f.read(1024 * 1024) |
| 182 | + if not chunk: |
| 183 | + break |
| 184 | + raw.write(chunk) |
| 185 | + payload = raw.getvalue() |
| 186 | + stream.write(compress(payload, algo)) |
| 187 | + stream.flush() |
| 188 | + return |
| 189 | + |
| 190 | + ensure_dir(os.path.dirname(outfile) or ".") |
| 191 | + with open(outfile, "wb") as f: |
| 192 | + f.write(header) |
| 193 | + raw = io.BytesIO() |
| 194 | + for name, path, size, mtime in entries: |
| 195 | + with open(path, "rb") as rf: |
| 196 | + while True: |
| 197 | + chunk = rf.read(1024 * 1024) |
| 198 | + if not chunk: |
| 199 | + break |
| 200 | + raw.write(chunk) |
| 201 | + payload = raw.getvalue() |
| 202 | + f.write(compress(payload, algo)) |
| 203 | + |
| 204 | +def _read_header(stream): |
| 205 | + magic = stream.read(4) |
| 206 | + if magic != MAGIC: |
| 207 | + raise ValueError("Not a PNF archive or corrupt magic: %r" % magic) |
| 208 | + flags, algo, hdrln = struct.unpack(">BBI", stream.read(6)) |
| 209 | + hdr = stream.read(hdrln) |
| 210 | + man = json.loads(to_text(hdr)) |
| 211 | + return flags, algo, man |
| 212 | + |
| 213 | +def list_archive(infile): |
| 214 | + stream = sys.stdin if infile in (None, "-") else open(infile, "rb") |
| 215 | + stream = getattr(stream, "buffer", stream) |
| 216 | + try: |
| 217 | + _, algo, man = _read_header(stream) |
| 218 | + return man |
| 219 | + finally: |
| 220 | + if stream is not sys.stdin and stream is not getattr(sys.stdin, "buffer", sys.stdin): |
| 221 | + stream.close() |
| 222 | + |
| 223 | +def unpack(infile, outdir): |
| 224 | + stream = sys.stdin if infile in (None, "-") else open(infile, "rb") |
| 225 | + stream = getattr(stream, "buffer", stream) |
| 226 | + try: |
| 227 | + flags, algo, man = _read_header(stream) |
| 228 | + payload = stream.read() |
| 229 | + data = decompress(payload, algo) |
| 230 | + cursor = 0 |
| 231 | + ensure_dir(outdir or ".") |
| 232 | + for entry in man.get("files", []): |
| 233 | + name = normpath(entry["name"]) |
| 234 | + size = int(entry["size"]) |
| 235 | + mtime = int(entry.get("mtime", int(time.time()))) |
| 236 | + outpath = os.path.join(outdir or ".", name) |
| 237 | + ensure_dir(os.path.dirname(outpath)) |
| 238 | + with open(outpath, "wb") as f: |
| 239 | + f.write(data[cursor:cursor+size]) |
| 240 | + os.utime(outpath, (mtime, mtime)) |
| 241 | + cursor += size |
| 242 | + if cursor != len(data): |
| 243 | + sys.stderr.write("Warning: payload trailing bytes: %d\n" % (len(data)-cursor)) |
| 244 | + finally: |
| 245 | + if stream is not sys.stdin and stream is not getattr(sys.stdin, "buffer", sys.stdin): |
| 246 | + stream.close() |
| 247 | + |
| 248 | +def _usage(): |
| 249 | + msg = ( |
| 250 | + "pyneofile_light.py - lightweight pack/unpack/list\\n\\n" |
| 251 | + "Usage:\\n" |
| 252 | + " pack OUTFILE [--comp auto|none|zlib|gzip|bz2|xz] SRC [SRC...]\\n" |
| 253 | + " unpack INFILE OUTDIR\\n" |
| 254 | + " list INFILE\\n" |
| 255 | + ) |
| 256 | + sys.stdout.write(msg) |
| 257 | + |
| 258 | +def main(argv=None): |
| 259 | + argv = list(sys.argv[1:] if argv is None else argv) |
| 260 | + if not argv or argv[0] in ("-h", "--help"): |
| 261 | + _usage() |
| 262 | + return 0 |
| 263 | + cmd = argv.pop(0) |
| 264 | + if cmd == "pack": |
| 265 | + if not argv: |
| 266 | + _usage(); return 2 |
| 267 | + outfile = argv.pop(0) |
| 268 | + comp = "auto" |
| 269 | + if argv and argv[0] == "--comp": |
| 270 | + argv.pop(0) |
| 271 | + if not argv: |
| 272 | + _usage(); return 2 |
| 273 | + comp = argv.pop(0) |
| 274 | + sources = argv |
| 275 | + if not sources: |
| 276 | + _usage(); return 2 |
| 277 | + pack(sources, outfile, comp) |
| 278 | + return 0 |
| 279 | + elif cmd == "unpack": |
| 280 | + if len(argv) < 2: |
| 281 | + _usage(); return 2 |
| 282 | + infile, outdir = argv[0], argv[1] |
| 283 | + unpack(infile, outdir) |
| 284 | + return 0 |
| 285 | + elif cmd == "list": |
| 286 | + if len(argv) < 1: |
| 287 | + _usage(); return 2 |
| 288 | + man = list_archive(argv[0]) |
| 289 | + sys.stdout.write(json.dumps(man, indent=2, sort_keys=True) + ("\n" if not PY2 else "")) |
| 290 | + return 0 |
| 291 | + else: |
| 292 | + _usage() |
| 293 | + return 2 |
| 294 | + |
| 295 | +if __name__ == "__main__": |
| 296 | + sys.exit(main()) |
0 commit comments