|
| 1 | +"""Internal classes used by the gzip, lzma and bz2 modules""" |
| 2 | + |
| 3 | +import io |
| 4 | + |
| 5 | + |
| 6 | +BUFFER_SIZE = io.DEFAULT_BUFFER_SIZE # Compressed data read chunk size |
| 7 | + |
| 8 | + |
| 9 | +class BaseStream(io.BufferedIOBase): |
| 10 | + """Mode-checking helper functions.""" |
| 11 | + |
| 12 | + def _check_not_closed(self): |
| 13 | + if self.closed: |
| 14 | + raise ValueError("I/O operation on closed file") |
| 15 | + |
| 16 | + def _check_can_read(self): |
| 17 | + if not self.readable(): |
| 18 | + raise io.UnsupportedOperation("File not open for reading") |
| 19 | + |
| 20 | + def _check_can_write(self): |
| 21 | + if not self.writable(): |
| 22 | + raise io.UnsupportedOperation("File not open for writing") |
| 23 | + |
| 24 | + def _check_can_seek(self): |
| 25 | + if not self.readable(): |
| 26 | + raise io.UnsupportedOperation("Seeking is only supported " |
| 27 | + "on files open for reading") |
| 28 | + if not self.seekable(): |
| 29 | + raise io.UnsupportedOperation("The underlying file object " |
| 30 | + "does not support seeking") |
| 31 | + |
| 32 | + |
| 33 | +class DecompressReader(io.RawIOBase): |
| 34 | + """Adapts the decompressor API to a RawIOBase reader API""" |
| 35 | + |
| 36 | + def readable(self): |
| 37 | + return True |
| 38 | + |
| 39 | + def __init__(self, fp, decomp_factory, trailing_error=(), **decomp_args): |
| 40 | + self._fp = fp |
| 41 | + self._eof = False |
| 42 | + self._pos = 0 # Current offset in decompressed stream |
| 43 | + |
| 44 | + # Set to size of decompressed stream once it is known, for SEEK_END |
| 45 | + self._size = -1 |
| 46 | + |
| 47 | + # Save the decompressor factory and arguments. |
| 48 | + # If the file contains multiple compressed streams, each |
| 49 | + # stream will need a separate decompressor object. A new decompressor |
| 50 | + # object is also needed when implementing a backwards seek(). |
| 51 | + self._decomp_factory = decomp_factory |
| 52 | + self._decomp_args = decomp_args |
| 53 | + self._decompressor = self._decomp_factory(**self._decomp_args) |
| 54 | + |
| 55 | + # Exception class to catch from decompressor signifying invalid |
| 56 | + # trailing data to ignore |
| 57 | + self._trailing_error = trailing_error |
| 58 | + |
| 59 | + def close(self): |
| 60 | + self._decompressor = None |
| 61 | + return super().close() |
| 62 | + |
| 63 | + def seekable(self): |
| 64 | + return self._fp.seekable() |
| 65 | + |
| 66 | + def readinto(self, b): |
| 67 | + with memoryview(b) as view, view.cast("B") as byte_view: |
| 68 | + data = self.read(len(byte_view)) |
| 69 | + byte_view[:len(data)] = data |
| 70 | + return len(data) |
| 71 | + |
| 72 | + def read(self, size=-1): |
| 73 | + if size < 0: |
| 74 | + return self.readall() |
| 75 | + |
| 76 | + if not size or self._eof: |
| 77 | + return b"" |
| 78 | + data = None # Default if EOF is encountered |
| 79 | + # Depending on the input data, our call to the decompressor may not |
| 80 | + # return any data. In this case, try again after reading another block. |
| 81 | + while True: |
| 82 | + if self._decompressor.eof: |
| 83 | + rawblock = (self._decompressor.unused_data or |
| 84 | + self._fp.read(BUFFER_SIZE)) |
| 85 | + if not rawblock: |
| 86 | + break |
| 87 | + # Continue to next stream. |
| 88 | + self._decompressor = self._decomp_factory( |
| 89 | + **self._decomp_args) |
| 90 | + try: |
| 91 | + data = self._decompressor.decompress(rawblock, size) |
| 92 | + except self._trailing_error: |
| 93 | + # Trailing data isn't a valid compressed stream; ignore it. |
| 94 | + break |
| 95 | + else: |
| 96 | + if self._decompressor.needs_input: |
| 97 | + rawblock = self._fp.read(BUFFER_SIZE) |
| 98 | + if not rawblock: |
| 99 | + raise EOFError("Compressed file ended before the " |
| 100 | + "end-of-stream marker was reached") |
| 101 | + else: |
| 102 | + rawblock = b"" |
| 103 | + data = self._decompressor.decompress(rawblock, size) |
| 104 | + if data: |
| 105 | + break |
| 106 | + if not data: |
| 107 | + self._eof = True |
| 108 | + self._size = self._pos |
| 109 | + return b"" |
| 110 | + self._pos += len(data) |
| 111 | + return data |
| 112 | + |
| 113 | + # Rewind the file to the beginning of the data stream. |
| 114 | + def _rewind(self): |
| 115 | + self._fp.seek(0) |
| 116 | + self._eof = False |
| 117 | + self._pos = 0 |
| 118 | + self._decompressor = self._decomp_factory(**self._decomp_args) |
| 119 | + |
| 120 | + def seek(self, offset, whence=io.SEEK_SET): |
| 121 | + # Recalculate offset as an absolute file position. |
| 122 | + if whence == io.SEEK_SET: |
| 123 | + pass |
| 124 | + elif whence == io.SEEK_CUR: |
| 125 | + offset = self._pos + offset |
| 126 | + elif whence == io.SEEK_END: |
| 127 | + # Seeking relative to EOF - we need to know the file's size. |
| 128 | + if self._size < 0: |
| 129 | + while self.read(io.DEFAULT_BUFFER_SIZE): |
| 130 | + pass |
| 131 | + offset = self._size + offset |
| 132 | + else: |
| 133 | + raise ValueError("Invalid value for whence: {}".format(whence)) |
| 134 | + |
| 135 | + # Make it so that offset is the number of bytes to skip forward. |
| 136 | + if offset < self._pos: |
| 137 | + self._rewind() |
| 138 | + else: |
| 139 | + offset -= self._pos |
| 140 | + |
| 141 | + # Read and discard data until we reach the desired position. |
| 142 | + while offset > 0: |
| 143 | + data = self.read(min(io.DEFAULT_BUFFER_SIZE, offset)) |
| 144 | + if not data: |
| 145 | + break |
| 146 | + offset -= len(data) |
| 147 | + |
| 148 | + return self._pos |
| 149 | + |
| 150 | + def tell(self): |
| 151 | + """Return the current file position.""" |
| 152 | + return self._pos |
0 commit comments