coding_placer.py

#!/usr/local/bin/python
# -*- coding:latin-1 -*-

# Inserts the above encoding lines in every file it finds within the current directory
# and its subdirectories. Tries to find similarities, in case the lines are already
# there, but spelled incorrectly or not as per PEP...something.

import time
import os
import subprocess
from difflib import SequenceMatcher

correct_lines   = [ "#!/usr/local/bin/python",  "# -*- coding:latin-1 -*-" ]

# similarity thresholds to test the existing first two lines. If values are below 60%, then
# the assumption is the existing lines are not encoding lines, and new ones get prepended.
replace_thres   = 75    # similarity >= 75%, replace line.
warning_thres   = 70    # >= 70 < 75, replace but warn user of low similarity.
noreplace_thres = 60    # >= 60 < 70, querry the user.

def cls():                      subprocess.call( 'cls', shell=True, cwd='./' )
def similar( a, b ):            return SequenceMatcher( None, a, b ).ratio()
def concat( a, b ): return ''.join( [ a, b ] )

def check_for_replace( a, b ):
    percent = similar( a, b )*100
    if   percent == 100:             return ( -1, -1, percent )     # ignore, it's fine
    elif percent >= replace_thres:   return (  1,  0, percent )     # replace, no warning
    elif percent >= warning_thres:   return (  1,  1, percent )     # replace, w/ warning
    elif percent >  noreplace_thres: return (  1,  0, percent )  \
                                        if ask_user( a, b ) else ( 0, 0, percent )
    else:                           return (  0,  0, percent )

def ask_user( a, b ):
    inpt = ''
    while inpt.lower() not in ( 'y', 'n' ):
        inpt = raw_input( 'Found %s. Replace the string? (y/n) ' % ( b ) )
    return inpt in 'y'

def prepend( fpath ):
    with open( fpath, 'r' ) as f:
        writen_something = False
        lines = list( f.read().split('\n') )
        tmp_lines_copy = lines
        for i, c in enumerate( correct_lines ):
            replace, warn, percent = check_for_replace( c, tmp_lines_copy[i] )
            if replace == -1:   # ignore, already correct in the file
                continue
            elif replace == 1:  # replace
                if warn == 1:
                    print "- WARNING: string '%s' only %.1f similar..." % ( lines[i], percent ),
                lines[i] = c
                writen_something = True
            else:               # prepend
                lines.insert( i, c )
                writen_something = True
        if writen_something:
            # add a blank line after the second one, to keep things neet
            if len( lines[2] ):
                lines[1] = concat( lines[1], '\n' )
            # a little quick hack to not add an extra \n at the end
            if not len( lines[-1] ):
                del lines[-1]
            with open( fpath, 'w' ) as fout:
                for l in lines:
                    fout.write( concat( l, '\n' ) )
            return True
        return False

def main():
    start = time.time()
    num_files = sum( [ len( files ) for r, d, files in os.walk(".\\") ] )
    cls()
    print "\n\t%d .py files found.\n\n" % ( num_files )
    for directory, d, files in os.walk('.\\'):
        # os.walk gives me paths w/ backslashes, so I use them for consistency
        for f in files:
            if f.endswith('.py'):
                if not directory.endswith('\\'):
                    fpath = concat( concat( directory, '\\' ), f )
                else:
                    fpath = concat( directory, f )

                print " %s" % ( fpath ),
                if prepend( fpath ):
                    print "Done."
                else:
                    print "is fine."
    end = time.time()
    print "\n\t %d files in %.3f seconds." % ( num_files, end-start )


if __name__ == "__main__":
    main()