# -*- coding: utf-8 - # # This file is part of http-parser released under the MIT license. # See the NOTICE for more information. import array import codecs from UserDict import DictMixin try: from thread import allocate_lock as Lock except ImportError: from dummy_thread import allocate_lock as Lock # open() uses st_blksize whenever we can DEFAULT_BUFFER_SIZE = 8 * 1024 # bytes bytes = str # bytearray is a mutable type that is easily turned into an instance of # bytes class bytearray(list): def __init__(self, i=0): if isinstance(i, int): list.__init__(self, [' '] * i) else: list.__init__(self, i) # for bytes(bytearray()) usage def __str__(self): return ''.join(self) # append automatically converts integers to characters def append(self, item): if isinstance(item, str): list.append(self, item) else: list.append(self, chr(item)) # += def __iadd__(self, other): for byte in other: self.append(byte) return self def endswith(self, c): return self[-1] == c def startswith(self, c): return self[0] == c class IOrderedDict(dict, DictMixin): def __init__(self, *args, **kwds): if len(args) > 1: raise TypeError('expected at most 1 arguments, got %d' % len(args)) try: self.__end except AttributeError: self.clear() self.update(*args, **kwds) self.__lower = {} def get(self, key, default=None): if key in self: return self[key] return default def clear(self): self.__end = end = [] end += [None, end, end] # sentinel node for doubly linked list self.__map = {} # key --> [key, prev, next] self.__lower = {} # key -> key.lower() dict.clear(self) def __setitem__(self, key, value): if key not in self: end = self.__end curr = end[1] curr[2] = end[1] = self.__map[key] = [key, curr, end] self.__lower[key.lower()] = key key = self.__lower[key.lower()] dict.__setitem__(self, key, value) def __delitem__(self, key): if key in self: key = self.__lower.pop(key.lower()) dict.__delitem__(self, key) key, prev, next = self.__map.pop(key) prev[2] = next next[1] = prev def __getitem__(self, key, dict_getitem=dict.__getitem__): if key in self: key = self.__lower.get(key.lower()) return dict_getitem(self, key) def __contains__(self, key): return key.lower() in self.__lower def __iter__(self): end = self.__end curr = end[2] while curr is not end: yield curr[0] curr = curr[2] def __reversed__(self): end = self.__end curr = end[1] while curr is not end: yield curr[0] curr = curr[1] def popitem(self, last=True): if not self: raise KeyError('dictionary is empty') if last: key = reversed(self).next() else: key = iter(self).next() value = self.pop(key) return key, value def __reduce__(self): items = [[k, self[k]] for k in self] tmp = self.__map, self.__end del self.__map, self.__end inst_dict = vars(self).copy() self.__map, self.__end = tmp if inst_dict: return (self.__class__, (items,), inst_dict) return self.__class__, (items,) def keys(self): return list(self) setdefault = DictMixin.setdefault update = DictMixin.update pop = DictMixin.pop values = DictMixin.values items = DictMixin.items iterkeys = DictMixin.iterkeys itervalues = DictMixin.itervalues iteritems = DictMixin.iteritems def __repr__(self): if not self: return '%s()' % (self.__class__.__name__,) return '%s(%r)' % (self.__class__.__name__, self.items()) def copy(self): return self.__class__(self) @classmethod def fromkeys(cls, iterable, value=None): d = cls() for key in iterable: d[key] = value return d def __eq__(self, other): if isinstance(other, OrderedDict): if len(self) != len(other): return False for p, q in zip(self.items(), other.items()): if p != q: return False return True return dict.__eq__(self, other) def __ne__(self, other): return not self == other class IOBase(object): ### Flush and close ### def seek(self, pos, whence=0): raise NotImplementedError def tell(self): """Return current stream position.""" return self.seek(0, 1) def truncate(self, pos=None): raise NotImplementedError def flush(self): """Flush write buffers, if applicable. This is not implemented for read-only and non-blocking streams. """ self._checkClosed() # XXX Should this return the number of bytes written??? __closed = False def close(self): """Flush and close the IO object. This method has no effect if the file is already closed. """ if not self.__closed: self.flush() self.__closed = True def __del__(self): """Destructor. Calls close().""" # The try/except block is in case this is called at program # exit time, when it's possible that globals have already been # deleted, and then the close() call might fail. Since # there's nothing we can do about such failures and they annoy # the end users, we suppress the traceback. try: self.close() except: pass @property def closed(self): """closed: bool. True iff the file has been closed. For backwards compatibility, this is a property, not a predicate. """ return self.__closed def _checkClosed(self, msg=None): """Internal: raise an ValueError if file is closed """ if self.closed: raise ValueError("I/O operation on closed file." if msg is None else msg) ### Context manager ### def __enter__(self): """Context management protocol. Returns self.""" self._checkClosed() return self def __exit__(self, *args): """Context management protocol. Calls close()""" self.close() def readable(self): """Return whether object was opened for reading. If False, read() will raise IOError. """ return False def _checkReadable(self, msg=None): """Internal: raise an IOError if file is not readable """ if not self.readable(): raise IOError("File or stream is not readable." if msg is None else msg) ### Readline[s] ### def readline(self, limit=-1): r"""Read and return a line from the stream. If limit is specified, at most limit bytes will be read. The line terminator is always b'\n' for binary files; for text files, the newlines argument to open can be used to select the line terminator(s) recognized. """ # For backwards compatibility, a (slowish) readline(). if hasattr(self, "peek"): def nreadahead(): readahead = self.peek(1) if not readahead: return 1 readahead = "".join(readahead) n = (readahead.find("\n") + 1) or len(readahead) if limit >= 0: n = min(n, limit) return n else: def nreadahead(): return 1 if limit is None: limit = -1 elif not isinstance(limit, (int, long)): raise TypeError("limit must be an integer") res = bytearray() while limit < 0 or len(res) < limit: b = self.read(nreadahead()) if not b: break res += b if res.endswith("\n"): break return bytes(res) def __iter__(self): self._checkClosed() return self def next(self): line = self.readline() if not line: raise StopIteration return line def readlines(self, hint=None): """Return a list of lines from the stream. hint can be specified to control the number of lines read: no more lines will be read if the total size (in bytes/characters) of all lines so far exceeds hint. """ if hint is not None and not isinstance(hint, (int, long)): raise TypeError("integer or None expected") if hint is None or hint <= 0: return list(self) n = 0 lines = [] for line in self: lines.append(line) n += len(line) if n >= hint: break return lines class RawIOBase(IOBase): """Base class for raw binary I/O.""" # The read() method is implemented by calling readinto(); derived # classes that want to support read() only need to implement # readinto() as a primitive operation. In general, readinto() can be # more efficient than read(). # (It would be tempting to also provide an implementation of # readinto() in terms of read(), in case the latter is a more suitable # primitive operation, but that would lead to nasty recursion in case # a subclass doesn't implement either.) def read(self, n=-1): """Read and return up to n bytes. Returns an empty bytes object on EOF, or None if the object is set not to block and has no data to read. """ if n is None: n = -1 if n < 0: return self.readall() b = bytearray(n.__index__()) n = self.readinto(b) if n is None: return None del b[n:] return bytes(b) def readall(self): """Read until EOF, using multiple read() call.""" res = bytearray() while True: data = self.read(DEFAULT_BUFFER_SIZE) if not data: break res += data return bytes(res) def readinto(self, b): """Read up to len(b) bytes into b. Returns number of bytes read (0 for EOF), or None if the object is set not to block and has no data to read. """ raise NotImplementedError class BufferedReader(RawIOBase): def __init__(self, raw, buffer_size=DEFAULT_BUFFER_SIZE): self.raw = raw if buffer_size <= 0: raise ValueError("invalid buffer size") self.buffer_size = buffer_size self._reset_read_buf() self._read_lock = Lock() def _reset_read_buf(self): self._read_buf = "" self._read_pos = 0 def read(self, n=None): """Read n bytes. Returns exactly n bytes of data unless the underlying raw IO stream reaches EOF or if the call would block in non-blocking mode. If n is negative, read until EOF or until read() would block. """ if n is not None and n < -1: raise ValueError("invalid number of bytes to read") self._read_lock.acquire() try: return self._read_unlocked(n) finally: self._read_lock.release() def readinto(self, b): """Read up to len(b) bytes into b. Like read(), this may issue multiple reads to the underlying raw stream, unless the latter is 'interactive'. Returns the number of bytes read (0 for EOF). Raises BlockingIOError if the underlying raw stream has no data at the moment. """ # XXX This ought to work with anything that supports the buffer API data = self.read(len(b)) n = len(data) try: b[:n] = data except TypeError, err: import array if not isinstance(b, array.array): raise err b[:n] = array.array('b', data) return n def _read_unlocked(self, n=None): nodata_val = "" empty_values = ("", None) buf = self._read_buf pos = self._read_pos # Special case for when the number of bytes to read is unspecified. if n is None or n == -1: self._reset_read_buf() chunks = [buf[pos:]] # Strip the consumed bytes. current_size = 0 while True: # Read until EOF or until read() would block. chunk = self.raw.read() if chunk in empty_values: nodata_val = chunk break current_size += len(chunk) chunks.append(chunk) return "".join(chunks) or nodata_val # The number of bytes to read is specified, return at most n bytes. avail = len(buf) - pos # Length of the available buffered data. if n <= avail: # Fast path: the data to read is fully buffered. self._read_pos += n return buf[pos:pos+n] # Slow path: read from the stream until enough bytes are read, # or until an EOF occurs or until read() would block. chunks = [buf[pos:]] wanted = max(self.buffer_size, n) while avail < n: chunk = self.raw.read(wanted) if chunk in empty_values: nodata_val = chunk break avail += len(chunk) chunks.append(chunk) # n is more then avail only when an EOF occurred or when # read() would have blocked. n = min(n, avail) out = "".join(chunks) self._read_buf = out[n:] # Save the extra data in the buffer. self._read_pos = 0 return out[:n] if out else nodata_val def peek(self, n=0): """Returns buffered bytes without advancing the position. The argument indicates a desired minimal number of bytes; we do at most one raw read to satisfy it. We never return more than self.buffer_size. """ self._read_lock.acquire() try: return self._peek_unlocked(n) finally: self._read_lock.release() def _peek_unlocked(self, n=0): want = min(n, self.buffer_size) have = len(self._read_buf) - self._read_pos if have < want or have <= 0: to_read = self.buffer_size - have current = self.raw.read(to_read) if current: self._read_buf = self._read_buf[self._read_pos:] + current self._read_pos = 0 return self._read_buf[self._read_pos:] def read1(self, n): """Reads up to n bytes, with at most one read() system call.""" # Returns up to n bytes. If at least one byte is buffered, we # only return buffered bytes. Otherwise, we do one raw read. if n < 0: raise ValueError("number of bytes to read must be positive") if n == 0: return "" self._read_lock.acquire() try: self._peek_unlocked(1) return self._read_unlocked( min(n, len(self._read_buf) - self._read_pos)) finally: self._read_lock.release() def tell(self): return _BufferedIOMixin.tell(self) - len(self._read_buf) + self._read_pos def seek(self, pos, whence=0): if not (0 <= whence <= 2): raise ValueError("invalid whence value") self._read_lock.acquire() try: if whence == 1: pos -= len(self._read_buf) - self._read_pos pos = self.raw.seek(pos, whence) if pos < 0: raise IOError("tell() returned an invalid position") self._reset_read_buf() return pos finally: self._read_lock.release() def truncate(self, pos=None): # Flush the stream. We're mixing buffered I/O with lower-level I/O, # and a flush may be necessary to synch both views of the current # file state. self.flush() if pos is None: pos = self.tell() # XXX: Should seek() be used, instead of passing the position # XXX directly to truncate? return self.raw.truncate(pos) ### Flush and close ### def flush(self): if self.closed: raise ValueError("flush of closed file") self.raw.flush() def close(self): if self.raw is not None and not self.closed: self.flush() self.raw.close() def detach(self): if self.raw is None: raise ValueError("raw stream already detached") self.flush() raw = self.raw self.raw = None return raw ### Inquiries ### def seekable(self): return self.raw.seekable() def readable(self): return self.raw.readable() @property def closed(self): return self.raw.closed class IncrementalNewlineDecoder(codecs.IncrementalDecoder): r"""Codec used when reading a file in universal newlines mode. It wraps another incremental decoder, translating \r\n and \r into \n. It also records the types of newlines encountered. When used with translate=False, it ensures that the newline sequence is returned in one piece. """ def __init__(self, decoder, translate, errors='strict'): codecs.IncrementalDecoder.__init__(self, errors=errors) self.translate = translate self.decoder = decoder self.seennl = 0 self.pendingcr = False def decode(self, input, final=False): # decode input (with the eventual \r from a previous pass) if self.decoder is None: output = input else: output = self.decoder.decode(input, final=final) if self.pendingcr and (output or final): output = "\r" + output self.pendingcr = False # retain last \r even when not translating data: # then readline() is sure to get \r\n in one pass if output.endswith("\r") and not final: output = output[:-1] self.pendingcr = True # Record which newlines are read crlf = output.count('\r\n') cr = output.count('\r') - crlf lf = output.count('\n') - crlf self.seennl |= (lf and self._LF) | (cr and self._CR) \ | (crlf and self._CRLF) if self.translate: if crlf: output = output.replace("\r\n", "\n") if cr: output = output.replace("\r", "\n") return output def getstate(self): if self.decoder is None: buf = "" flag = 0 else: buf, flag = self.decoder.getstate() flag <<= 1 if self.pendingcr: flag |= 1 return buf, flag def setstate(self, state): buf, flag = state self.pendingcr = bool(flag & 1) if self.decoder is not None: self.decoder.setstate((buf, flag >> 1)) def reset(self): self.seennl = 0 self.pendingcr = False if self.decoder is not None: self.decoder.reset() _LF = 1 _CR = 2 _CRLF = 4 @property def newlines(self): return (None, "\n", "\r", ("\r", "\n"), "\r\n", ("\n", "\r\n"), ("\r", "\r\n"), ("\r", "\n", "\r\n") )[self.seennl] class TextIOWrapper(IOBase): _CHUNK_SIZE = 2048 def __init__(self, buffer, encoding=None, errors=None, newline=None, line_buffering=False): if newline is not None and not isinstance(newline, basestring): raise TypeError("illegal newline type: %r" % (type(newline),)) if newline not in (None, "", "\n", "\r", "\r\n"): raise ValueError("illegal newline value: %r" % (newline,)) if encoding is None: try: import locale except ImportError: # Importing locale may fail if Python is being built encoding = "ascii" else: encoding = locale.getpreferredencoding() if not isinstance(encoding, basestring): raise ValueError("invalid encoding: %r" % encoding) if errors is None: errors = "strict" else: if not isinstance(errors, basestring): raise ValueError("invalid errors: %r" % errors) self.buffer = buffer self._line_buffering = line_buffering self._encoding = encoding self._errors = errors self._readuniversal = not newline self._readtranslate = newline is None self._readnl = newline self._writetranslate = newline != '' self._writenl = newline or os.linesep self._encoder = None self._decoder = None self._decoded_chars = '' # buffer for text returned from decoder self._decoded_chars_used = 0 # offset into _decoded_chars for read() self._snapshot = None # info for reconstructing decoder state self._seekable = self._telling = self.buffer.seekable() @property def encoding(self): return self._encoding @property def errors(self): return self._errors @property def line_buffering(self): return self._line_buffering def seekable(self): return self._seekable def readable(self): return self.buffer.readable() def flush(self): self.buffer.flush() self._telling = self._seekable def close(self): if self.buffer is not None and not self.closed: self.flush() self.buffer.close() @property def closed(self): return self.buffer.closed if self.closed: raise ValueError("write to closed file") if not isinstance(s, unicode): raise TypeError("can't write %s to text stream" % s.__class__.__name__) length = len(s) haslf = (self._writetranslate or self._line_buffering) and "\n" in s if haslf and self._writetranslate and self._writenl != "\n": s = s.replace("\n", self._writenl) encoder = self._encoder or self._get_encoder() # XXX What if we were just reading? b = encoder.encode(s) self.buffer.write(b) if self._line_buffering and (haslf or "\r" in s): self.flush() self._snapshot = None if self._decoder: self._decoder.reset() return length def _get_encoder(self): make_encoder = codecs.getincrementalencoder(self._encoding) self._encoder = make_encoder(self._errors) return self._encoder def _get_decoder(self): make_decoder = codecs.getincrementaldecoder(self._encoding) decoder = make_decoder(self._errors) if self._readuniversal: decoder = IncrementalNewlineDecoder(decoder, self._readtranslate) self._decoder = decoder return decoder # The following three methods implement an ADT for _decoded_chars. # Text returned from the decoder is buffered here until the client # requests it by calling our read() or readline() method. def _set_decoded_chars(self, chars): """Set the _decoded_chars buffer.""" self._decoded_chars = chars self._decoded_chars_used = 0 def _get_decoded_chars(self, n=None): """Advance into the _decoded_chars buffer.""" offset = self._decoded_chars_used if n is None: chars = self._decoded_chars[offset:] else: chars = self._decoded_chars[offset:offset + n] self._decoded_chars_used += len(chars) return chars def _rewind_decoded_chars(self, n): """Rewind the _decoded_chars buffer.""" if self._decoded_chars_used < n: raise AssertionError("rewind decoded_chars out of bounds") self._decoded_chars_used -= n def _read_chunk(self): """ Read and decode the next chunk of data from the BufferedReader. """ # The return value is True unless EOF was reached. The decoded # string is placed in self._decoded_chars (replacing its previous # value). The entire input chunk is sent to the decoder, though # some of it may remain buffered in the decoder, yet to be # converted. if self._decoder is None: raise ValueError("no decoder") if self._telling: # To prepare for tell(), we need to snapshot a point in the # file where the decoder's input buffer is empty. dec_buffer, dec_flags = self._decoder.getstate() # Given this, we know there was a valid snapshot point # len(dec_buffer) bytes ago with decoder state (b'', dec_flags). # Read a chunk, decode it, and put the result in self._decoded_chars. input_chunk = self.buffer.read1(self._CHUNK_SIZE) eof = not input_chunk self._set_decoded_chars(self._decoder.decode(input_chunk, eof)) if self._telling: # At the snapshot point, len(dec_buffer) bytes before the read, # the next input to be decoded is dec_buffer + input_chunk. self._snapshot = (dec_flags, dec_buffer + input_chunk) return not eof def _pack_cookie(self, position, dec_flags=0, bytes_to_feed=0, need_eof=0, chars_to_skip=0): # The meaning of a tell() cookie is: seek to position, set the # decoder flags to dec_flags, read bytes_to_feed bytes, feed them # into the decoder with need_eof as the EOF flag, then skip # chars_to_skip characters of the decoded result. For most simple # decoders, tell() will often just give a byte offset in the file. return (position | (dec_flags<<64) | (bytes_to_feed<<128) | (chars_to_skip<<192) | bool(need_eof)<<256) def _unpack_cookie(self, bigint): rest, position = divmod(bigint, 1<<64) rest, dec_flags = divmod(rest, 1<<64) rest, bytes_to_feed = divmod(rest, 1<<64) need_eof, chars_to_skip = divmod(rest, 1<<64) return position, dec_flags, bytes_to_feed, need_eof, chars_to_skip def tell(self): if not self._seekable: raise IOError("underlying stream is not seekable") if not self._telling: raise IOError("telling position disabled by next() call") self.flush() position = self.buffer.tell() decoder = self._decoder if decoder is None or self._snapshot is None: if self._decoded_chars: # This should never happen. raise AssertionError("pending decoded text") return position # Skip backward to the snapshot point (see _read_chunk). dec_flags, next_input = self._snapshot position -= len(next_input) # How many decoded characters have been used up since the snapshot? chars_to_skip = self._decoded_chars_used if chars_to_skip == 0: # We haven't moved from the snapshot point. return self._pack_cookie(position, dec_flags) # Starting from the snapshot position, we will walk the decoder # forward until it gives us enough decoded characters. saved_state = decoder.getstate() try: # Note our initial start point. decoder.setstate(('', dec_flags)) start_pos = position start_flags, bytes_fed, chars_decoded = dec_flags, 0, 0 need_eof = 0 # Feed the decoder one byte at a time. As we go, note the # nearest "safe start point" before the current location # (a point where the decoder has nothing buffered, so seek() # can safely start from there and advance to this location). for next_byte in next_input: bytes_fed += 1 chars_decoded += len(decoder.decode(next_byte)) dec_buffer, dec_flags = decoder.getstate() if not dec_buffer and chars_decoded <= chars_to_skip: # Decoder buffer is empty, so this is a safe start point. start_pos += bytes_fed chars_to_skip -= chars_decoded start_flags, bytes_fed, chars_decoded = dec_flags, 0, 0 if chars_decoded >= chars_to_skip: break else: # We didn't get enough decoded data; signal EOF to get more. chars_decoded += len(decoder.decode('', final=True)) need_eof = 1 if chars_decoded < chars_to_skip: raise IOError("can't reconstruct logical file position") # The returned cookie corresponds to the last safe start point. return self._pack_cookie( start_pos, start_flags, bytes_fed, need_eof, chars_to_skip) finally: decoder.setstate(saved_state) def truncate(self, pos=None): self.flush() if pos is None: pos = self.tell() return self.buffer.truncate(pos) def detach(self): if self.buffer is None: raise ValueError("buffer is already detached") self.flush() buffer = self.buffer self.buffer = None return buffer def seek(self, cookie, whence=0): if self.closed: raise ValueError("tell on closed file") if not self._seekable: raise IOError("underlying stream is not seekable") if whence == 1: # seek relative to current position if cookie != 0: raise IOError("can't do nonzero cur-relative seeks") # Seeking to the current position should attempt to # sync the underlying buffer with the current position. whence = 0 cookie = self.tell() if whence == 2: # seek relative to end of file if cookie != 0: raise IOError("can't do nonzero end-relative seeks") self.flush() position = self.buffer.seek(0, 2) self._set_decoded_chars('') self._snapshot = None if self._decoder: self._decoder.reset() return position if whence != 0: raise ValueError("invalid whence (%r, should be 0, 1 or 2)" % (whence,)) if cookie < 0: raise ValueError("negative seek position %r" % (cookie,)) self.flush() # The strategy of seek() is to go back to the safe start point # and replay the effect of read(chars_to_skip) from there. start_pos, dec_flags, bytes_to_feed, need_eof, chars_to_skip = \ self._unpack_cookie(cookie) # Seek back to the safe start point. self.buffer.seek(start_pos) self._set_decoded_chars('') self._snapshot = None # Restore the decoder to its state from the safe start point. if cookie == 0 and self._decoder: self._decoder.reset() elif self._decoder or dec_flags or chars_to_skip: self._decoder = self._decoder or self._get_decoder() self._decoder.setstate(('', dec_flags)) self._snapshot = (dec_flags, '') if chars_to_skip: # Just like _read_chunk, feed the decoder and save a snapshot. input_chunk = self.buffer.read(bytes_to_feed) self._set_decoded_chars( self._decoder.decode(input_chunk, need_eof)) self._snapshot = (dec_flags, input_chunk) # Skip chars_to_skip of the decoded characters. if len(self._decoded_chars) < chars_to_skip: raise IOError("can't restore logical file position") self._decoded_chars_used = chars_to_skip # Finally, reset the encoder (merely useful for proper BOM handling) try: encoder = self._encoder or self._get_encoder() except LookupError: # Sometimes the encoder doesn't exist pass else: if cookie != 0: encoder.setstate(0) else: encoder.reset() return cookie def read(self, n=None): self._checkReadable() if n is None: n = -1 decoder = self._decoder or self._get_decoder() try: n.__index__ except AttributeError: raise TypeError("an integer is required") if n < 0: # Read everything. result = (self._get_decoded_chars() + decoder.decode(self.buffer.read(), final=True)) self._set_decoded_chars('') self._snapshot = None return result else: # Keep reading chunks until we have n characters to return. eof = False result = self._get_decoded_chars(n) while len(result) < n and not eof: eof = not self._read_chunk() result += self._get_decoded_chars(n - len(result)) return result def next(self): self._telling = False line = self.readline() if not line: self._snapshot = None self._telling = self._seekable raise StopIteration return line def readline(self, limit=None): if self.closed: raise ValueError("read from closed file") if limit is None: limit = -1 elif not isinstance(limit, (int, long)): raise TypeError("limit must be an integer") # Grab all the decoded text (we will rewind any extra bits later). line = self._get_decoded_chars() start = 0 # Make the decoder if it doesn't already exist. if not self._decoder: self._get_decoder() pos = endpos = None while True: if self._readtranslate: # Newlines are already translated, only search for \n pos = line.find('\n', start) if pos >= 0: endpos = pos + 1 break else: start = len(line) elif self._readuniversal: # Universal newline search. Find any of \r, \r\n, \n # The decoder ensures that \r\n are not split in two pieces # In C we'd look for these in parallel of course. nlpos = line.find("\n", start) crpos = line.find("\r", start) if crpos == -1: if nlpos == -1: # Nothing found start = len(line) else: # Found \n endpos = nlpos + 1 break elif nlpos == -1: # Found lone \r endpos = crpos + 1 break elif nlpos < crpos: # Found \n endpos = nlpos + 1 break elif nlpos == crpos + 1: # Found \r\n endpos = crpos + 2 break else: # Found \r endpos = crpos + 1 break else: # non-universal pos = line.find(self._readnl) if pos >= 0: endpos = pos + len(self._readnl) break if limit >= 0 and len(line) >= limit: endpos = limit # reached length limit break # No line ending seen yet - get more data' while self._read_chunk(): if self._decoded_chars: break if self._decoded_chars: line += self._get_decoded_chars() else: # end of file self._set_decoded_chars('') self._snapshot = None return line if limit >= 0 and endpos > limit: endpos = limit # don't exceed limit # Rewind _decoded_chars to just after the line ending we found. self._rewind_decoded_chars(len(line) - endpos) return line[:endpos] @property def newlines(self): return self._decoder.newlines if self._decoder else None