Issue #15068: Got rid of excessive buffering in the fileinput module.
The bufsize parameter is no longer used.
diff --git a/Lib/fileinput.py b/Lib/fileinput.py
index 021e39f..9f14762 100644
--- a/Lib/fileinput.py
+++ b/Lib/fileinput.py
@@ -64,13 +64,6 @@
disabled when standard input is read. XXX The current implementation
does not work for MS-DOS 8+3 filesystems.
-Performance: this module is unfortunately one of the slower ways of
-processing large numbers of input lines. Nevertheless, a significant
-speed-up has been obtained by using readlines(bufsize) instead of
-readline(). A new keyword argument, bufsize=N, is present on the
-input() function and the FileInput() class to override the default
-buffer size.
-
XXX Possible additions:
- optional getopt argument processing
@@ -86,6 +79,7 @@
_state = None
+# No longer used
DEFAULT_BUFSIZE = 8*1024
def input(files=None, inplace=False, backup="", bufsize=0,
@@ -207,17 +201,15 @@
self._files = files
self._inplace = inplace
self._backup = backup
- self._bufsize = bufsize or DEFAULT_BUFSIZE
self._savestdout = None
self._output = None
self._filename = None
- self._lineno = 0
+ self._startlineno = 0
self._filelineno = 0
self._file = None
+ self._readline = self._start_readline
self._isstdin = False
self._backupfilename = None
- self._buffer = []
- self._bufindex = 0
# restrict mode argument to reading modes
if mode not in ('r', 'rU', 'U', 'rb'):
raise ValueError("FileInput opening mode must be one of "
@@ -253,22 +245,18 @@
return self
def __next__(self):
- try:
- line = self._buffer[self._bufindex]
- except IndexError:
- pass
- else:
- self._bufindex += 1
- self._lineno += 1
+ line = self._readline()
+ if line:
self._filelineno += 1
return line
- line = self.readline()
- if not line:
+ if not self._file:
raise StopIteration
- return line
+ self.nextfile()
+ # Recursive call
+ return self.__next__()
def __getitem__(self, i):
- if i != self._lineno:
+ if i != self.lineno():
raise RuntimeError("accessing lines out of order")
try:
return self.__next__()
@@ -289,6 +277,7 @@
finally:
file = self._file
self._file = None
+ self._readline = self._start_readline
try:
if file and not self._isstdin:
file.close()
@@ -300,85 +289,81 @@
except OSError: pass
self._isstdin = False
- self._buffer = []
- self._bufindex = 0
def readline(self):
- try:
- line = self._buffer[self._bufindex]
- except IndexError:
- pass
- else:
- self._bufindex += 1
- self._lineno += 1
- self._filelineno += 1
- return line
- if not self._file:
- if not self._files:
- if 'b' in self._mode:
- return b''
- else:
- return ''
- self._filename = self._files[0]
- self._files = self._files[1:]
- self._filelineno = 0
- self._file = None
- self._isstdin = False
- self._backupfilename = 0
- if self._filename == '-':
- self._filename = '<stdin>'
- if 'b' in self._mode:
- self._file = getattr(sys.stdin, 'buffer', sys.stdin)
- else:
- self._file = sys.stdin
- self._isstdin = True
+ while True:
+ line = self._readline()
+ if line:
+ self._filelineno += 1
+ return line
+ if not self._file:
+ return line
+ self.nextfile()
+ # repeat with next file
+
+ def _start_readline(self):
+ if not self._files:
+ if 'b' in self._mode:
+ return b''
else:
- if self._inplace:
- self._backupfilename = (
- self._filename + (self._backup or ".bak"))
+ return ''
+ self._filename = self._files[0]
+ self._files = self._files[1:]
+ self._startlineno = self.lineno()
+ self._filelineno = 0
+ self._file = None
+ self._isstdin = False
+ self._backupfilename = 0
+ if self._filename == '-':
+ self._filename = '<stdin>'
+ if 'b' in self._mode:
+ self._file = getattr(sys.stdin, 'buffer', sys.stdin)
+ else:
+ self._file = sys.stdin
+ self._isstdin = True
+ else:
+ if self._inplace:
+ self._backupfilename = (
+ self._filename + (self._backup or ".bak"))
+ try:
+ os.unlink(self._backupfilename)
+ except OSError:
+ pass
+ # The next few lines may raise OSError
+ os.rename(self._filename, self._backupfilename)
+ self._file = open(self._backupfilename, self._mode)
+ try:
+ perm = os.fstat(self._file.fileno()).st_mode
+ except OSError:
+ self._output = open(self._filename, "w")
+ else:
+ mode = os.O_CREAT | os.O_WRONLY | os.O_TRUNC
+ if hasattr(os, 'O_BINARY'):
+ mode |= os.O_BINARY
+
+ fd = os.open(self._filename, mode, perm)
+ self._output = os.fdopen(fd, "w")
try:
- os.unlink(self._backupfilename)
+ if hasattr(os, 'chmod'):
+ os.chmod(self._filename, perm)
except OSError:
pass
- # The next few lines may raise OSError
- os.rename(self._filename, self._backupfilename)
- self._file = open(self._backupfilename, self._mode)
- try:
- perm = os.fstat(self._file.fileno()).st_mode
- except OSError:
- self._output = open(self._filename, "w")
- else:
- mode = os.O_CREAT | os.O_WRONLY | os.O_TRUNC
- if hasattr(os, 'O_BINARY'):
- mode |= os.O_BINARY
-
- fd = os.open(self._filename, mode, perm)
- self._output = os.fdopen(fd, "w")
- try:
- if hasattr(os, 'chmod'):
- os.chmod(self._filename, perm)
- except OSError:
- pass
- self._savestdout = sys.stdout
- sys.stdout = self._output
+ self._savestdout = sys.stdout
+ sys.stdout = self._output
+ else:
+ # This may raise OSError
+ if self._openhook:
+ self._file = self._openhook(self._filename, self._mode)
else:
- # This may raise OSError
- if self._openhook:
- self._file = self._openhook(self._filename, self._mode)
- else:
- self._file = open(self._filename, self._mode)
- self._buffer = self._file.readlines(self._bufsize)
- self._bufindex = 0
- if not self._buffer:
- self.nextfile()
- # Recursive call
- return self.readline()
+ self._file = open(self._filename, self._mode)
+ self._readline = self._file.readline
+ return self._readline()
def filename(self):
return self._filename
def lineno(self):
- return self._lineno
+ return self._startlineno + self._filelineno
def filelineno(self):
return self._filelineno