Lib/fileinput.py

   1 """Helper class to quickly write a loop over all standard input files.
   2
   3 Typical use is:
   4
   5     import fileinput
   6     for line in fileinput.input():
   7         process(line)
   8
   9 This iterates over the lines of all files listed in sys.argv[1:],
  10 defaulting to sys.stdin if the list is empty.  If a filename is '-' it
  11 is also replaced by sys.stdin.  To specify an alternative list of
  12 filenames, pass it as the argument to input().  A single file name is
  13 also allowed.
  14
  15 Functions filename(), lineno() return the filename and cumulative line
  16 number of the line that has just been read; filelineno() returns its
  17 line number in the current file; isfirstline() returns true iff the
  18 line just read is the first line of its file; isstdin() returns true
  19 iff the line was read from sys.stdin.  Function nextfile() closes the
  20 current file so that the next iteration will read the first line from
  21 the next file (if any); lines not read from the file will not count
  22 towards the cumulative line count; the filename is not changed until
  23 after the first line of the next file has been read.  Function close()
  24 closes the sequence.
  25
  26 Before any lines have been read, filename() returns None and both line
  27 numbers are zero; nextfile() has no effect.  After all lines have been
  28 read, filename() and the line number functions return the values
  29 pertaining to the last line read; nextfile() has no effect.
  30
  31 All files are opened in text mode.  If an I/O error occurs during
  32 opening or reading a file, the IOError exception is raised.
  33
  34 If sys.stdin is used more than once, the second and further use will
  35 return no lines, except perhaps for interactive use, or if it has been
  36 explicitly reset (e.g. using sys.stdin.seek(0)).
  37
  38 Empty files are opened and immediately closed; the only time their
  39 presence in the list of filenames is noticeable at all is when the
  40 last file opened is empty.
  41
  42 It is possible that the last line of a file doesn't end in a newline
  43 character; otherwise lines are returned including the trailing
  44 newline.
  45
  46 Class FileInput is the implementation; its methods filename(),
  47 lineno(), fileline(), isfirstline(), isstdin(), nextfile() and close()
  48 correspond to the functions in the module.  In addition it has a
  49 readline() method which returns the next input line, and a
  50 __getitem__() method which implements the sequence behavior.  The
  51 sequence must be accessed in strictly sequential order; sequence
  52 access and readline() cannot be mixed.
  53
  54 Optional in-place filtering: if the keyword argument inplace=1 is
  55 passed to input() or to the FileInput constructor, the file is moved
  56 to a backup file and standard output is directed to the input file.
  57 This makes it possible to write a filter that rewrites its input file
  58 in place.  If the keyword argument backup=".<some extension>" is also
  59 given, it specifies the extension for the backup file, and the backup
  60 file remains around; by default, the extension is ".bak" and it is
  61 deleted when the output file is closed.  In-place filtering is
  62 disabled when standard input is read.  XXX The current implementation
  63 does not work for MS-DOS 8+3 filesystems.
  64
  65 XXX Possible additions:
  66
  67 - optional getopt argument processing
  68 - specify open mode ('r' or 'rb')
  69 - specify buffer size
  70 - fileno()
  71 - isatty()
  72 - read(), read(size), even readlines()
  73
  74 """
  75
  76 import sys, os
  77
  78 _state = None
  79
  80 def input(files=(), inplace=0, backup=""):
  81     global _state
  82     if _state and _state._file:
  83         raise RuntimeError, "input() already active"
  84     _state = FileInput(files, inplace, backup)
  85     return _state
  86
  87 def close():
  88     global _state
  89     state = _state
  90     _state = None
  91     if state:
  92         state.close()
  93
  94 def nextfile():
  95     if not _state:
  96         raise RuntimeError, "no active input()"
  97     return _state.nextfile()
  98
  99 def filename():
 100     if not _state:
 101         raise RuntimeError, "no active input()"
 102     return _state.filename()
 103
 104 def lineno():
 105     if not _state:
 106         raise RuntimeError, "no active input()"
 107     return _state.lineno()
 108
 109 def filelineno():
 110     if not _state:
 111         raise RuntimeError, "no active input()"
 112     return _state.filelineno()
 113
 114 def isfirstline():
 115     if not _state:
 116         raise RuntimeError, "no active input()"
 117     return _state.isfirstline()
 118
 119 def isstdin():
 120     if not _state:
 121         raise RuntimeError, "no active input()"
 122     return _state.isstdin()
 123
 124 class FileInput:
 125
 126     def __init__(self, files=(), inplace=0, backup=""):
 127         if type(files) == type(''):
 128             files = (files,)
 129         else:
 130             files = tuple(files)
 131             if not files:
 132                 files = tuple(sys.argv[1:])
 133                 if not files:
 134                     files = ('-',)
 135         self._files = files
 136         self._inplace = inplace
 137         self._backup = backup
 138         self._savestdout = None
 139         self._output = None
 140         self._filename = None
 141         self._lineno = 0
 142         self._filelineno = 0
 143         self._file = None
 144         self._isstdin = 0
 145         self._backupfilename = None
 146
 147     def __del__(self):
 148         self.close()
 149
 150     def close(self):
 151         self.nextfile()
 152         self._files = ()
 153
 154     def __getitem__(self, i):
 155         if i != self._lineno:
 156             raise RuntimeError, "accessing lines out of order"
 157         line = self.readline()
 158         if not line:
 159             raise IndexError, "end of input reached"
 160         return line
 161
 162     def nextfile(self):
 163         savestdout = self._savestdout
 164         self._savestdout = 0
 165         if savestdout:
 166             sys.stdout = savestdout
 167
 168         output = self._output
 169         self._output = 0
 170         if output:
 171             output.close()
 172
 173         file = self._file
 174         self._file = 0
 175         if file and not self._isstdin:
 176             file.close()
 177
 178         backupfilename = self._backupfilename
 179         self._backupfilename = 0
 180         if backupfilename and not self._backup:
 181             try: os.unlink(backupfilename)
 182             except: pass
 183
 184         self._isstdin = 0
 185
 186     def readline(self):
 187         if not self._file:
 188             if not self._files:
 189                 return ""
 190             self._filename = self._files[0]
 191             self._files = self._files[1:]
 192             self._filelineno = 0
 193             self._file = None
 194             self._isstdin = 0
 195             self._backupfilename = 0
 196             if self._filename == '-':
 197                 self._filename = '<stdin>'
 198                 self._file = sys.stdin
 199                 self._isstdin = 1
 200             else:
 201                 if self._inplace:
 202                     self._backupfilename = (
 203                         self._filename + (self._backup or ".bak"))
 204                     try: os.unlink(self._backupfilename)
 205                     except os.error: pass
 206                     # The next three lines may raise IOError
 207                     os.rename(self._filename, self._backupfilename)
 208                     self._file = open(self._backupfilename, "r")
 209                     self._output = open(self._filename, "w")
 210                     self._savestdout = sys.stdout
 211                     sys.stdout = self._output
 212                 else:
 213                     # This may raise IOError
 214                     self._file = open(self._filename, "r")
 215         line = self._file.readline()
 216         if line:
 217             self._lineno = self._lineno + 1
 218             self._filelineno = self._filelineno + 1
 219             return line
 220         self.nextfile()
 221         # Recursive call
 222         return self.readline()
 223
 224     def filename(self):
 225         return self._filename
 226
 227     def lineno(self):
 228         return self._lineno
 229
 230     def filelineno(self):
 231         return self._filelineno
 232
 233     def isfirstline(self):
 234         return self._filelineno == 1
 235
 236     def isstdin(self):
 237         return self._isstdin
 238
 239 def _test():
 240     import getopt
 241     inplace = 0
 242     backup = 0
 243     opts, args = getopt.getopt(sys.argv[1:], "ib:")
 244     for o, a in opts:
 245         if o == '-i': inplace = 1
 246         if o == '-b': backup = a
 247     for line in input(args, inplace=inplace, backup=backup):
 248         if line[-1:] == '\n': line = line[:-1]
 249         if line[-1:] == '\r': line = line[:-1]
 250         print "%d: %s[%d]%s %s" % (lineno(), filename(), filelineno(),
 251                                    isfirstline() and "*" or "", line)
 252     print "%d: %s[%d]" % (lineno(), filename(), filelineno())
 253
 254 if __name__ == '__main__':
 255     _test()