| Index: trunk/extensions/ConfirmEdit/captcha.py |
| — | — | @@ -26,17 +26,24 @@ |
| 27 | 27 | # 2008-01-06: Add regex check to skip words containing other than a-z |
| 28 | 28 | |
| 29 | 29 | import random |
| 30 | | -import Image |
| 31 | | -import ImageFont |
| 32 | | -import ImageDraw |
| 33 | | -import ImageEnhance |
| 34 | | -import ImageOps |
| 35 | | -import math, string, md5 |
| 36 | | -import getopt |
| | 30 | +import math |
| | 31 | +import hashlib |
| | 32 | +from optparse import OptionParser |
| 37 | 33 | import os |
| 38 | 34 | import sys |
| 39 | 35 | import re |
| 40 | 36 | |
| | 37 | +try: |
| | 38 | + import Image |
| | 39 | + import ImageFont |
| | 40 | + import ImageDraw |
| | 41 | + import ImageEnhance |
| | 42 | + import ImageOps |
| | 43 | +except: |
| | 44 | + sys.exit("This script requires the Python Imaging Library - http://www.pythonware.com/products/pil/") |
| | 45 | + |
| | 46 | +nonalpha = re.compile('[^a-z]') # regex to test for suitability of words |
| | 47 | + |
| 41 | 48 | # Does X-axis wobbly copy, sandwiched between two rotates |
| 42 | 49 | def wobbly_copy(src, wob, col, scale, ang): |
| 43 | 50 | x, y = src.size |
| — | — | @@ -48,7 +55,6 @@ |
| 49 | 56 | # Do a cheap bounding-box op here to try to limit work below |
| 50 | 57 | bbx = rot.getbbox() |
| 51 | 58 | if bbx == None: |
| 52 | | - print "whoops" |
| 53 | 59 | return src |
| 54 | 60 | else: |
| 55 | 61 | l, t, r, b= bbx |
| — | — | @@ -102,13 +108,13 @@ |
| 103 | 109 | # save the image, in format determined from filename |
| 104 | 110 | im.save(file_name) |
| 105 | 111 | |
| 106 | | -def gen_subdir(basedir, hash, levels): |
| | 112 | +def gen_subdir(basedir, md5hash, levels): |
| 107 | 113 | """Generate a subdirectory path out of the first _levels_ |
| 108 | 114 | characters of _hash_, and ensure the directories exist |
| 109 | 115 | under _basedir_.""" |
| 110 | 116 | subdir = None |
| 111 | 117 | for i in range(0, levels): |
| 112 | | - char = hash[i] |
| | 118 | + char = md5hash[i] |
| 113 | 119 | if subdir: |
| 114 | 120 | subdir = os.path.join(subdir, char) |
| 115 | 121 | else: |
| — | — | @@ -124,9 +130,9 @@ |
| 125 | 131 | word = word1+word2 |
| 126 | 132 | if verbose: |
| 127 | 133 | print "word is %s" % word |
| 128 | | - r = re.compile('[^a-z]'); |
| 129 | | - if r.search(word): |
| 130 | | - print "skipping word pair '%s' because it contains non-alphabetic characters" % word |
| | 134 | + if nonalpha.search(word): |
| | 135 | + if verbose: |
| | 136 | + print "skipping word pair '%s' because it contains non-alphabetic characters" % word |
| 131 | 137 | return None |
| 132 | 138 | |
| 133 | 139 | for naughty in blacklist: |
| — | — | @@ -137,13 +143,14 @@ |
| 138 | 144 | return word |
| 139 | 145 | |
| 140 | 146 | def pick_word(words, blacklist, verbose): |
| 141 | | - while True: |
| | 147 | + for x in range(1000): # If we can't find a valid combination in 1000 tries, just give up |
| 142 | 148 | word = try_pick_word(words, blacklist, verbose) |
| 143 | 149 | if word: |
| 144 | 150 | return word |
| | 151 | + sys.exit("Unable to find valid word combinations") |
| 145 | 152 | |
| 146 | 153 | def read_wordlist(filename): |
| 147 | | - return [string.lower(x.strip()) for x in open(wordlist).readlines()] |
| | 154 | + return [x.strip().lower() for x in open(wordlist).readlines()] |
| 148 | 155 | |
| 149 | 156 | if __name__ == '__main__': |
| 150 | 157 | """This grabs random words from the dictionary 'words' (one |
| — | — | @@ -153,47 +160,51 @@ |
| 154 | 161 | To check a reply, hash it in the same way with the same salt and |
| 155 | 162 | secret key, then compare with the hash value given. |
| 156 | 163 | """ |
| 157 | | - font = "VeraBd.ttf" |
| 158 | | - wordlist = "awordlist.txt" |
| 159 | | - blacklistfile = None |
| 160 | | - key = "CHANGE_THIS_SECRET!" |
| 161 | | - output = "." |
| 162 | | - count = 20 |
| 163 | | - fill = 0 |
| 164 | | - dirs = 0 |
| 165 | | - verbose = False |
| | 164 | + parser = OptionParser() |
| | 165 | + parser.add_option("--wordlist", help="A list of words (required)", metavar="WORDS.txt") |
| | 166 | + parser.add_option("--key", help="The passphrase set as $wgCaptchaSecret (required)", metavar="KEY") |
| | 167 | + parser.add_option("--output", help="The directory to put the images in - $wgCaptchaDirectory (required)", metavar="DIR") |
| | 168 | + parser.add_option("--font", help="The font to use (required)", metavar="FONT.ttf") |
| | 169 | + parser.add_option("--font-size", help="The font size (default 40)", metavar="N", type='int', default=40) |
| | 170 | + parser.add_option("--count", help="The maximum number of images to make (default 20)", metavar="N", type='int', default=20) |
| | 171 | + parser.add_option("--blacklist", help="A blacklist of words that should not be used", metavar="FILE") |
| | 172 | + parser.add_option("--fill", help="Fill the output directory to contain N files, overrides count, cannot be used with --dirs", metavar="N", type='int') |
| | 173 | + parser.add_option("--dirs", help="Put the images into subdirectories N levels deep - $wgCaptchaDirectoryLevels", metavar="N", type='int') |
| | 174 | + parser.add_option("--verbose", "-v", help="Show debugging information", action='store_true') |
| 166 | 175 | |
| 167 | | - opts, args = getopt.getopt(sys.argv[1:], "", ["font=", "wordlist=", "blacklist=", "key=", "output=", "count=", "fill=", "dirs=", "verbose"]) |
| 168 | | - for o, a in opts: |
| 169 | | - if o == "--font": |
| 170 | | - font = a |
| 171 | | - if o == "--wordlist": |
| 172 | | - wordlist = a |
| 173 | | - if o == "--blacklist": |
| 174 | | - blacklistfile = a |
| 175 | | - if o == "--key": |
| 176 | | - key = a |
| 177 | | - if o == "--output": |
| 178 | | - output = a |
| 179 | | - if o == "--count": |
| 180 | | - count = int(a) |
| 181 | | - if o == "--fill": |
| 182 | | - fill = int(a) |
| 183 | | - if o == "--dirs": |
| 184 | | - dirs = int(a) |
| 185 | | - if o == "--verbose": |
| 186 | | - verbose = True |
| | 176 | + opts, args = parser.parse_args() |
| | 177 | + |
| | 178 | + if opts.wordlist: |
| | 179 | + wordlist = opts.wordlist |
| | 180 | + else: |
| | 181 | + sys.exit("Need to specify a wordlist") |
| | 182 | + if opts.key: |
| | 183 | + key = opts.key |
| | 184 | + else: |
| | 185 | + sys.exit("Need to specify a key") |
| | 186 | + if opts.output: |
| | 187 | + output = opts.output |
| | 188 | + else: |
| | 189 | + sys.exit("Need to specify an output directory") |
| | 190 | + if opts.font and os.path.exists(opts.font): |
| | 191 | + font = opts.font |
| | 192 | + else: |
| | 193 | + sys.exit("Need to specify the location of a font") |
| 187 | 194 | |
| | 195 | + blacklistfile = opts.blacklist |
| | 196 | + count = opts.count |
| | 197 | + fill = opts.fill |
| | 198 | + dirs = opts.dirs |
| | 199 | + verbose = opts.verbose |
| | 200 | + fontsize = opts.font_size |
| | 201 | + |
| 188 | 202 | if fill: |
| 189 | | - # Option processing order is not guaranteed, so count the output |
| 190 | | - # files after... |
| 191 | 203 | count = max(0, fill - len(os.listdir(output))) |
| 192 | 204 | |
| 193 | 205 | words = read_wordlist(wordlist) |
| 194 | 206 | words = [x for x in words |
| 195 | | - if len(x) <= 5 and len(x) >= 4 and x[0] != "f" |
| 196 | | - and x[0] != x[1] and x[-1] != x[-2] |
| 197 | | - and (not "'" in x)] |
| | 207 | + if len(x) in (4,5) and x[0] != "f" |
| | 208 | + and x[0] != x[1] and x[-1] != x[-2]] |
| 198 | 209 | |
| 199 | 210 | if blacklistfile: |
| 200 | 211 | blacklist = read_wordlist(blacklistfile) |
| — | — | @@ -204,11 +215,12 @@ |
| 205 | 216 | word = pick_word(words, blacklist, verbose) |
| 206 | 217 | salt = "%08x" % random.randrange(2**32) |
| 207 | 218 | # 64 bits of hash is plenty for this purpose |
| 208 | | - hash = md5.new(key+salt+word+key+salt).hexdigest()[:16] |
| 209 | | - filename = "image_%s_%s.png" % (salt, hash) |
| | 219 | + md5hash = hashlib.md5(key+salt+word+key+salt).hexdigest()[:16] |
| | 220 | + filename = "image_%s_%s.png" % (salt, md5hash) |
| 210 | 221 | if dirs: |
| 211 | | - subdir = gen_subdir(output, hash, dirs) |
| | 222 | + subdir = gen_subdir(output, md5hash, dirs) |
| 212 | 223 | filename = os.path.join(subdir, filename) |
| 213 | 224 | if verbose: |
| 214 | 225 | print filename |
| 215 | | - gen_captcha(word, font, 40, os.path.join(output, filename)) |
| | 226 | + gen_captcha(word, font, fontsize, os.path.join(output, filename)) |
| | 227 | + |