#!/usr/bin/python2

#
# Copyright 2004 Michael Kefeder (h.t.d@gmx.de)
# 
# This program renames files
# It converts filenames from ANY supported encoding to UTF-8, this enables
#    a pretty good co-existence of netatalk 2.0 and samba 3.0
#    If character decoding fails then the name is quoted in URL-style
#    (special-chars become %xx codes)
# It converts CAP-sequences from ANY supported encoding to UTF-8
# It removes characters illegal for windows: */\?<>:|",
#    whitespace from beginning and end of filenames,
#    dots from beginning and end of filenames
#    unless disabled by cmd-line parameter
# Filenames that would be truncated to a zero length string will be renamed
#    to a quote from the Movie "Finding Nemo"
#    e.g. ' ...' -=> 'Shark bait, oo ha ha!'
#
# TODO:
# * check if filename-length exceeds 32 characters (mac os9, afp2.x limit)
#   --max-namelength <LENGTH> (default 32)
#   --check-namelength "only_duplicates|always"
#   if filename too long: <LENGTH> - 1 and iterate from 0-9 to find a new name
#      no name found: <LENGTH> - 2 and iterate from 0-99 to find a new name
#      no name found: this is insane, stop processing!
# * documentation (manpage?)

#
# FUTURE:
# * cmd-line switch "--newfs-encoding=<ENCODING>" to allow
#   output in other encodings than utf-8
# * cmd-line switch to list available encodings 
#   (not implemented in python, point user at python2 documentation: 
#   http://docs.python.org/lib/node127.html and simply check if the
#   passed argument is an existing codec before running.)
# * cmd-line switch to set what to do if de-/en-coding of characters fails (abort, ignore,
#   replace,...)
# * cmd-line switches for VETO_FILES and VETO_DIRS
# * cmd-line switch for invalid characters

import os
import sys
import string, urllib
import re
import getopt

# presets
#
INVALID_CHARS = unicode ('*/\?<>:|"')
VETO_DIRS = (".AppleDB", ".AppleDouble", ".AppleDesktop")
VETO_FILES = ("Icon\r", ".htaccess", ":2eDS_Store", ".DS_Store")

def usage():
  print "USAGE: %s [-r|--recursive] <PATH>" % (sys.argv[0])
  print "\t -v|--verbose displays 'real-filename' ('CAPdecoded') -=> 'new-filename'"
  print "\t --notest ACTUALLY moves the files!"
  print "\t --no-capdecode disable decoding of CAP encoded sequences"
  print "\t --no-winvalid disable creation of windows-compatible names"
  print "\t --no-db-update disable updating of CNID-database"
  print "\t --cap-encoding=<encoding> Character Encoding of CAP-sequences\n\t   in filenames"
  print "\t   default: MAC-ROMAN"
  print "\t --fs-encoding=<encoding> Character Encoding of filenames\n\t   as they are in filesystem"
  print "\t   default: UTF-8"
  print "\t --cnid-db-path=<DB_PATH> Path to .AppleDB directory"
  print "\t   default: <PATH>/.AppleDB"
  print "\t -h|--help for help (this message)"
  print "NOTE: all output is UTF-8 encoded!"
  print "for a list of supported encodings see:"
  print "http://docs.python.org/lib/node127.html"
  sys.exit(1)

#if len (sys.argv) < 2:
#  usage()

try:
    opts, args = getopt.getopt(sys.argv[1:], "rvh", 
	["recursive", "verbose", "help", "notest", "no-capdecode", "no-winvalid", "no-db-update", "cap-encoding=", "fs-encoding=", "cnid-db-path="])
except getopt.GetoptError:
    usage()

pRecursive = 0
pVerbose = 0
pNoTest = 0
pNoCAP = 0
pNoWIN = 0
pNoDB = 0
pCAPenc = 'mac-roman'
pFSenc = 'utf-8'
pDB_PATH = ''
for o, a in opts:
  if o in ("-r", "--recursive"):
    pRecursive = 1
  if o in ("-v", "--verbose"):
    pVerbose = 1
  if o == "--notest":
    pNoTest = 1
  if o == "--no-capdecode":
    pNoCAP = 1
  if o == "--no-winvalid":
    pNoWIN = 1
  if o == "--no-db-update":
    pNoDB = 1
  if o == "--cap-encoding":
    pCAPenc = a
  if o == "--fs-encoding":
    pFSenc = a
  if o == "--cnid-db-path":
    pDB_PATH = a
  if o in ("-h", "--help"):
    usage()

if len(args) == 0:
  usage()

# python 2.2 workarounds
try:
  x = UnicodeDecodeError
except NameError:
  UnicodeDecodeError = UnicodeError
else:
  del x
  
start_path = args[0]

if pNoDB == 0:
	from cnid2db import cnid_db
	cnid2 = cnid_db()
	if pDB_PATH is '':
		cnid2.open (start_path)
	else:
		cnid2.open (pDB_PATH)

if pNoTest == 1:
	recursive = ''
	if pRecursive == 1:
		recursive = ' RECURSIVELY'
	sys.stderr.write ("This will rename all files in '%s'%s. Are you sure? (Y/n): " % (start_path, recursive))
	sure = sys.stdin.readline()
	if string.strip(sure) != 'Y':
		sys.exit(1)
	sys.stderr.write ("Are you abso-fuckin'-lutely sure? (Y/n): ")
	sure = sys.stdin.readline()
	if string.strip(sure) != 'Y':
		sys.exit(1)
else:
	sys.stderr.write ("To enable actual renaming use the --notest cmd-line option, see --help for more info.\n")

def isValidUTF8 (s):
	return re.compile ("^(([\0-\x7F])|([\xC2-\xDF][\x80-\xBF])|((([\xE0][\xA0-\xBF])|([\xE1-\xEC\xEE-\xEF][\x80-\xBF])|([\xED][\x80-\x9F]))[\x80-\xBF])|((([\xF0][\x90-\xBF])|([\xF1-\xF3][\x80-\xBF])|([\xF4][\x80-\x8F]))[\x80-\xBF][\x80-\xBF]))*$").match(s)

def stripChar (c, s):
	while len(s) > 0 and s[0] == c:
		s = s[1:]
	while len(s) > 0 and s[-1] == c:
		s = s[:-1]
	return s

def invalidencoding2unicode (filename, fs_encoding):
	retval = unicode ('')
	for character in filename:
		try:
			retval += unicode (character, fs_encoding)
		except UnicodeDecodeError:
			retval += urllib.quote (character)
	return retval

def filename2unicode (filename, fs_encoding):
	"""
	converts filename from current encoding to unicode if the filename is not valid UTF-8
	this is useful to not convert existing utf-8 encoded names again, and again, and...
	if decoding fails the bad characters are quoted URL conform (%xx encoded)
	"""
	if not isValidUTF8 (filename):
		try:
			return unicode (filename, fs_encoding)
		except UnicodeDecodeError:
			return invalidencoding2unicode(filename, fs_encoding)#unicode (urllib.quote (filename))
	else:
		return unicode (filename, 'utf-8')

def cap2unicode (cap_encoded, cap_encoding = "mac-roman"):
	"""
	decodes ':xx' encoded characters from cap_encoding to unicode
	"""
	retval = unicode ('')
	c = 0
	p = string.find (cap_encoded, ":", 0)
	if p is not -1:
		while c < len(cap_encoded):
			if cap_encoded[c] is unicode (':'):
				hexcode = cap_encoded[c+1:c+3]
				if hexcode[0] in string.hexdigits and hexcode[1] in string.hexdigits:
					value = int(hexcode, 16)
					try:
						retval += unicode ("%c" % value, cap_encoding)
					except UnicodeDecodeError:
						retval += urllib.quote ("%c" % value)
					c += 2
				else:
					retval += ':'
			else:
				retval += cap_encoded[c]
			c += 1
	else:
		retval = cap_encoded
	return retval # unicode (retval, cap_encoding)

def createValidFilename (brokenFilename, fixedFilename = ""):
	"""
	replaces invalid characters for windows with '_'
	strips whitespace from beg./end of names
	strips dots from beg./end of names
	runs at least twice per name to check if output of 'createValidFilename' is really a valid name
	it recursively continues until filename has a length of 0 or stays the same after validation.
	example filename why recursion is necessary: '...  file.extension  ...'
	    when the dots are stripped invalid spaces are left
	"""
	brokenFilename = string.strip (brokenFilename) # remove whitespace from start and end of filename
	retval = unicode("")
	for character in brokenFilename[:]:
		if character in INVALID_CHARS:
			retval += unicode ("_")
		else:
			retval += character
	retval = stripChar (unicode('.'), retval)
	if len(retval) == 0:
		return unicode("Shark bait, oo ha ha!")
	if retval == fixedFilename:
		return retval
	else:
		return createValidFilename (retval, retval) # run recursively because invalid filenames like ' ... ... ... .... ..' result in yet again invalid names, and again, and again....

def createUniqueFilename (newfilename, newfiles, files):
	# rename only if newfilename does NOT already exist
	if newfilename in newfiles or newfilename.encode ('utf-8') in files:
		sys.stdout.write (" WARNING: '%s' exists!" % (newfilename.encode ('utf-8')))
		i = 0
		maxlen = 30 # NOTE: maximum length for atalk v2 is 31 chars, for a new name it needs to append new chars & stay within limit
		tmpfilename = newfilename[:]
		while newfilename in newfiles or newfilename.encode ('utf-8') in files:
			newfilename = "%s%d" % (tmpfilename[:maxlen], i)
			i += 1
			if i == 10:
				maxlen -= 1
			if i > 99: # protection against crazy circumstances
				sys.stderr.write ("OMG, do something about that!")
				sys.exit(1)
	return newfilename

# statistics
ntotalf = 0
ntotald = 0
nrenamedf = 0
nrenamedd = 0
maxdepth = 0

def scanDirectory (path, recursive = 0, depth = 0):
	global ntotalf, ntotald, nrenamedf, nrenamedd, maxdepth
	if maxdepth < depth:
		maxdepth = depth
	os.chdir(path)
	print "%2dCWD: '%s'" % (depth, os.getcwd())
	path = '.'
	files = os.listdir (path)
	files.sort()
	newfiles = [] # list of filenames which have been processed. Is needed to check if a file exists before moving
	oldfiles = [] # list of filenames already processed
	while len (files) > 0:
		# emulate .pop() for lists, inefficient! needed to check for filenames which have been truncated at the beginning
		curfilename = files[0]
		stat_data = os.stat (curfilename)
		cur_inode = stat_data.st_ino
		cur_device = stat_data.st_dev
		files = files [1:]
		if os.path.isfile (curfilename):
			ntotalf += 1
			if curfilename in VETO_FILES:
				continue
			newfilename = filename2unicode(curfilename, pFSenc)
			sys.stdout.write ("%2dFILE: '%s'" % (depth, newfilename.encode('utf-8')))
			if pNoCAP == 0:
				newfilename = cap2unicode (newfilename, pCAPenc)
				deCAPfilename = newfilename[:]
			if pNoWIN == 0:
				newfilename = createValidFilename (newfilename)
			newfilename = createUniqueFilename (newfilename, newfiles, files)
			newfiles.append (newfilename)
			newfilename = newfilename.encode ('utf-8')
			extras = ""
			if curfilename != newfilename:
				nrenamedf += 1
				if pNoTest == 1:
					os.rename(curfilename, newfilename)
					if os.path.isdir ("./.AppleDouble") and os.path.isfile (os.path.join ("./.AppleDouble", curfilename)):
						os.rename(os.path.join ("./.AppleDouble", curfilename), os.path.join ("./.AppleDouble", newfilename))
						extras += "R"
					if pNoDB == 0:
						cnid2.renameFromDEVINO (cur_device, cur_inode, newfilename)
						extras += "D"
				else:
					if pNoDB == 0:
						if cnid2.DEVINOexists (cur_device, cur_inode) is not 0:
							sys.stdout.write (" WARNING dev:%d ino:%d not in DB!" % (cur_device, cur_inode))
				if pVerbose == 0:
					sys.stdout.write (" -=> '%s' %s\n" % (newfilename, extras))
				else:
					sys.stdout.write (" ('%s') -=> '%s' %s\n" % (deCAPfilename.encode ('utf-8'), newfilename, extras))
			else:
				sys.stdout.write("\n") # newline to separate unprocessed filenames
			sys.stdout.flush()
		elif os.path.isdir (curfilename):
			ntotald += 1
			if curfilename in VETO_DIRS:
				continue
			newfilename = filename2unicode(curfilename, pFSenc)
			sys.stdout.write ("%2dDIRECTORY: '%s'" % (depth, newfilename.encode('utf-8')))
			if pNoCAP == 0:
				newfilename = cap2unicode (newfilename, pCAPenc)
				deCAPfilename = newfilename[:]
			if pNoWIN == 0:
				newfilename = createValidFilename (newfilename)
			newfilename = createUniqueFilename (newfilename, newfiles, files)
			newfiles.append (newfilename)
			newfilename = newfilename.encode ('utf-8')
			extras = ""
			if curfilename != newfilename:
				nrenamedd += 1
				if pNoTest == 1:
					os.rename(curfilename, newfilename)
					if os.path.isdir ("./.AppleDouble") and os.path.isfile (os.path.join ("./.AppleDouble", curfilename)):
						os.rename(os.path.join ("./.AppleDouble", curfilename), os.path.join ("./.AppleDouble", newfilename))
						extras += "R"
					if pNoDB == 0:
						cnid2.renameFromDEVINO (cur_device, cur_inode, newfilename)
						extras += "D"
				else:
					if pNoDB == 0:
						if cnid2.DEVINOexists (cur_device, cur_inode) is not 0:
							sys.stdout.write (" WARNING dev:%d ino:%d not in DB!" % (cur_device, cur_inode))
				if pVerbose == 0:
					sys.stdout.write (" -=> '%s' %s\n" % (newfilename, extras))
				else:
					sys.stdout.write (" ('%s') -=> '%s' %s\n" % (deCAPfilename.encode ('utf-8'), newfilename, extras))
			else:
				sys.stdout.write("\n") # newline to separate unprocessed filenames
			sys.stdout.flush()
			if recursive == 1:
				if pNoTest == 0:
					newfilename = curfilename
				scanDirectory(newfilename, 1, depth + 1)
	os.chdir('..')

scanDirectory (start_path, pRecursive)
sys.stderr.write ("Depth of deepest directory tree:  %d\n" % maxdepth)
sys.stderr.write ("Total # of Files processed:       %d\n" % ntotalf)
sys.stderr.write ("Total # of Directories processed: %d\n" % ntotald)
sys.stderr.write ("Total # of Files renamed:         %d\n" % nrenamedf)
sys.stderr.write ("Total # of Directories renamed:   %d\n" % nrenamedd)
# close Database
if pNoDB == 0:
	cnid2.close()