summaryrefslogtreecommitdiffstats
path: root/module/plugins/captcha/captcha.py
diff options
context:
space:
mode:
Diffstat (limited to 'module/plugins/captcha/captcha.py')
-rw-r--r--module/plugins/captcha/captcha.py56
1 files changed, 35 insertions, 21 deletions
diff --git a/module/plugins/captcha/captcha.py b/module/plugins/captcha/captcha.py
index cc07f50cf..b67ce9b9e 100644
--- a/module/plugins/captcha/captcha.py
+++ b/module/plugins/captcha/captcha.py
@@ -2,11 +2,11 @@
from __future__ import with_statement
-import GifImagePlugin
-import Image
-import JpegImagePlugin
-import PngImagePlugin
-import TiffImagePlugin
+try:
+ from PIL import Image, GifImagePlugin, JpegImagePlugin, PngImagePlugin, TiffImagePlugin
+except ImportError:
+ import Image, GifImagePlugin, JpegImagePlugin, PngImagePlugin, TiffImagePlugin
+
import logging
import os
import subprocess
@@ -16,30 +16,34 @@ from os.path import abspath, join
class OCR(object):
- __name__ = "OCR"
- __type__ = "ocr"
+ __name__ = "OCR"
+ __type__ = "ocr"
__version__ = "0.1"
__description__ = """OCR base plugin"""
- __author_name__ = "pyLoad Team"
- __author_mail__ = "admin@pyload.org"
+ __license__ = "GPLv3"
+ __authors__ = [("pyLoad Team", "admin@pyload.org")]
def __init__(self):
self.logger = logging.getLogger("log")
+
def load_image(self, image):
self.image = Image.open(image)
self.pixels = self.image.load()
self.result_captcha = ''
+
def unload(self):
"""delete all tmp images"""
pass
+
def threshold(self, value):
self.image = self.image.point(lambda a: a * value + 10)
+
def run(self, command):
"""Run a command"""
@@ -50,29 +54,32 @@ class OCR(object):
popen.stderr.close()
self.logger.debug("Tesseract ReturnCode %s Output: %s" % (popen.returncode, output))
+
def run_tesser(self, subset=False, digits=True, lowercase=True, uppercase=True):
- #self.logger.debug("create tmp tif")
+ #tmpTif = tempfile.NamedTemporaryFile(suffix=".tif")
+ try:
+ tmpTif = open(join("tmp", "tmpTif_%s.tif" % self.__name__), "wb")
+ tmpTif.close()
+
+ #tmpTxt = tempfile.NamedTemporaryFile(suffix=".txt")
+ tmpTxt = open(join("tmp", "tmpTxt_%s.txt" % self.__name__), "wb")
+ tmpTxt.close()
- #tmp = tempfile.NamedTemporaryFile(suffix=".tif")
- tmp = open(join("tmp", "tmpTif_%s.tif" % self.__name__), "wb")
- tmp.close()
- #self.logger.debug("create tmp txt")
- #tmpTxt = tempfile.NamedTemporaryFile(suffix=".txt")
- tmpTxt = open(join("tmp", "tmpTxt_%s.txt" % self.__name__), "wb")
- tmpTxt.close()
+ except IOError, e:
+ self.logError(e)
+ return
self.logger.debug("save tiff")
- self.image.save(tmp.name, 'TIFF')
+ self.image.save(tmpTif.name, 'TIFF')
if os.name == "nt":
tessparams = [join(pypath, "tesseract", "tesseract.exe")]
else:
tessparams = ["tesseract"]
- tessparams.extend( [abspath(tmp.name), abspath(tmpTxt.name).replace(".txt", "")] )
+ tessparams.extend( [abspath(tmpTif.name), abspath(tmpTxt.name).replace(".txt", "")] )
if subset and (digits or lowercase or uppercase):
- #self.logger.debug("create temp subset config")
#tmpSub = tempfile.NamedTemporaryFile(suffix=".subset")
tmpSub = open(join("tmp", "tmpSub_%s.subset" % self.__name__), "wb")
tmpSub.write("tessedit_char_whitelist ")
@@ -99,22 +106,25 @@ class OCR(object):
self.logger.debug(self.result_captcha)
try:
- os.remove(tmp.name)
+ os.remove(tmpTif.name)
os.remove(tmpTxt.name)
if subset and (digits or lowercase or uppercase):
os.remove(tmpSub.name)
except:
pass
+
def get_captcha(self, name):
raise NotImplementedError
+
def to_greyscale(self):
if self.image.mode != 'L':
self.image = self.image.convert('L')
self.pixels = self.image.load()
+
def eval_black_white(self, limit):
self.pixels = self.image.load()
w, h = self.image.size
@@ -125,6 +135,7 @@ class OCR(object):
else:
self.pixels[x, y] = 0
+
def clean(self, allowed):
pixels = self.pixels
@@ -170,6 +181,7 @@ class OCR(object):
self.pixels = pixels
+
def derotate_by_average(self):
"""rotate by checking each angle and guess most suitable"""
@@ -244,6 +256,7 @@ class OCR(object):
self.pixels = pixels
+
def split_captcha_letters(self):
captcha = self.image
started = False
@@ -283,6 +296,7 @@ class OCR(object):
return letters
+
def correct(self, values, var=None):
if var:
result = var