def run():
# Argument parser
parser = argparse.ArgumentParser(description='Plainsight: a textual steganography tool to make your data look like prose.' + \
'\nHomepage and examples: http://github.com/rw/plainsight' + \
'\nAuthor: Robert Winslow (@robert_winslow)',
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog=textwrap.dedent('''\
Example invocations:
Encipher:
echo "the meeting place is Union Square at 5pm" | plainsight --mode=encipher -f a-book.txt > ciphertext
Decipher:
cat ciphertext | plainsight --mode=decipher -f a-book.txt
'''))
parser.add_argument(
'-m',
'--mode',
metavar='MODE',
type=str,
nargs=1,
required=True,
choices=['encipher', 'decipher'],
help='Operating mode. Valid options are: encipher, decipher.')
parser.add_argument('-c', '--context', metavar='N', type=int, nargs='?', default=3, choices=xrange(11),
help='The number of tokens to use as context. ' + \
'Choose a number from 0 through 10. Default is 3. ' + \
'The time to build a model increases [superlinearly] with this, but so does similarity to the source texts.')
parser.add_argument('-f', '--model-file', metavar='FILE', type=argparse.FileType('rb'), nargs='+',
required=True,
help='One or more files to use when creating the lingustic model. ' + \
'The data from the file will be used verbatim. ' + \
'Example input: a downloaded a Project Guttenberg book.')
parser.add_argument('-u', '--model-url', metavar='URL', type=str, nargs='*',
help='(Not implemented.) One or more URLs to use when creating the lingustic model. ' + \
'The data at the URL will be stripped of markup. ' + \
'Example input: http://www.gutenberg.org/ebooks/1661.txt.utf8 (Sherlock Holmes).')
# Parse the arguments
args = parser.parse_args()
MODEL_FILES = args.model_file
MODE = args.mode[0]
CONTEXT = max(0, args.context - 1)
INPUT, OUTPUT, LOGGING = sys.stdin, sys.stdout, sys.stderr
# generate model from input text
model = Model(CONTEXT)
# load files for language model:
if MODEL_FILES is not None:
sys.stderr.write('Adding models:\n')
for f in MODEL_FILES:
start = time()
model_text = data.take_char_input(f)
model.add_text(model_text, CONTEXT)
sys.stderr.write('Model: %s added in %.02fs (context == %d)\n' \
% (f.name, time() - start, CONTEXT))
sys.stderr.write('input is "%s", ' % INPUT.name)
sys.stderr.write('output is "%s"\n' % OUTPUT.name)
sys.stderr.write('\n')
# encipher or decipher the input
if MODE == 'encipher':
cleartext = data.take_binary_input(INPUT)
ciphertext = cipher(model, CONTEXT, cleartext, 'encipher')
sys.stderr.write('\n')
sys.stdout.write(ciphertext)
elif MODE == 'decipher':
ciphertext = data.to_words(data.take_char_input(INPUT))
cleartext = cipher(model, CONTEXT, ciphertext, 'decipher')
sys.stderr.write('\n')
sys.stdout.write(cleartext.tobytes())
def add_text(self, text, context):
phrases = data.to_phrases(context, data.to_words(text))
for p in phrases:
self.add_phrase(p)
self.update_root_count()
def add_text(self, text, context):
phrases = data.to_phrases(context, data.to_words(text))
for p in phrases:
self.add_phrase(p)
self.update_root_count()
def run():
# Argument parser
parser = argparse.ArgumentParser(
description="Plainsight: a textual steganography tool to make your data look like prose."
+ "\nHomepage and examples: http://github.com/rw/plainsight"
+ "\nAuthor: Robert Winslow (@robert_winslow)",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog=textwrap.dedent(
"""\
Example invocations:
Encipher:
echo "the meeting place is Union Square at 5pm" | plainsight --mode=encipher -f a-book.txt > ciphertext
Decipher:
cat ciphertext | plainsight --mode=decipher -f a-book.txt
"""
),
)
parser.add_argument(
"-m",
"--mode",
metavar="MODE",
type=str,
nargs=1,
required=True,
choices=["encipher", "decipher"],
help="Operating mode. Valid options are: encipher, decipher.",
)
parser.add_argument(
"-c",
"--context",
metavar="N",
type=int,
nargs="?",
default=3,
choices=xrange(11),
help="The number of tokens to use as context. "
+ "Choose a number from 0 through 10. Default is 3. "
+ "The time to build a model increases [superlinearly] with this, but so does similarity to the source texts.",
)
parser.add_argument(
"-f",
"--model-file",
metavar="FILE",
type=argparse.FileType("rb"),
nargs="+",
required=True,
help="One or more files to use when creating the lingustic model. "
+ "The data from the file will be used verbatim. "
+ "Example input: a downloaded a Project Guttenberg book.",
)
parser.add_argument(
"-u",
"--model-url",
metavar="URL",
type=str,
nargs="*",
help="(Not implemented.) One or more URLs to use when creating the lingustic model. "
+ "The data at the URL will be stripped of markup. "
+ "Example input: http://www.gutenberg.org/ebooks/1661.txt.utf8 (Sherlock Holmes).",
)
# Parse the arguments
args = parser.parse_args()
MODEL_FILES = args.model_file
MODE = args.mode[0]
CONTEXT = max(0, args.context - 1)
INPUT, OUTPUT, LOGGING = sys.stdin, sys.stdout, sys.stderr
# generate model from input text
model = Model(CONTEXT)
# load files for language model:
if MODEL_FILES is not None:
sys.stderr.write("Adding models:\n")
for f in MODEL_FILES:
start = time()
model_text = data.take_char_input(f)
model.add_text(model_text, CONTEXT)
sys.stderr.write("Model: %s added in %.02fs (context == %d)\n" % (f.name, time() - start, CONTEXT))
sys.stderr.write('input is "%s", ' % INPUT.name)
sys.stderr.write('output is "%s"\n' % OUTPUT.name)
sys.stderr.write("\n")
# encipher or decipher the input
if MODE == "encipher":
cleartext = data.take_binary_input(INPUT)
ciphertext = cipher(model, CONTEXT, cleartext, "encipher")
sys.stderr.write("\n")
sys.stdout.write(ciphertext)
elif MODE == "decipher":
ciphertext = data.to_words(data.take_char_input(INPUT))
cleartext = cipher(model, CONTEXT, ciphertext, "decipher")
sys.stderr.write("\n")
sys.stdout.write(cleartext.tobytes())