From e3d190b10c4b9016d74598e459293e6e82851b2e Mon Sep 17 00:00:00 2001
From: heffnercj <heffnercj@gmail.com>
Date: Sun, 17 Nov 2013 11:07:54 -0500
Subject: [PATCH] Basic signature scans now work in python3; other features TBD
---
src/binwalk/__init__.py | 5 ++++-
src/binwalk/common.py | 2 ++
src/binwalk/compat.py | 4 ++--
src/binwalk/parser.py | 8 ++++----
4 files changed, 12 insertions(+), 7 deletions(-)
diff --git a/src/binwalk/__init__.py b/src/binwalk/__init__.py
index 6bb34dc..6451666 100644
--- a/src/binwalk/__init__.py
+++ b/src/binwalk/__init__.py
@@ -539,8 +539,11 @@ class Binwalk(object):
results = []
results_offset = -1
+ # In python3 we need a bytes object to pass to magic.buffer
+ candidate_data = str2bytes(data[i+candidate:i+candidate+fd.MAX_TRAILING_SIZE])
+
# Pass the data to libmagic, and split out multiple results into a list
- for magic_result in self.parser.split(self.magic.buffer(data[i+candidate:i+candidate+fd.MAX_TRAILING_SIZE])):
+ for magic_result in self.parser.split(self.magic.buffer(candidate_data)):
i_set_results_offset = False
diff --git a/src/binwalk/common.py b/src/binwalk/common.py
index d192abf..4fa9a4a 100644
--- a/src/binwalk/common.py
+++ b/src/binwalk/common.py
@@ -158,6 +158,8 @@ class BlockFile(io.FileIO):
data = self.read(self.READ_BLOCK_SIZE + self.MAX_TRAILING_SIZE)
if data and data is not None:
+ data = bytes2str(data)
+
# Get the actual length of the read in data
dlen = len(data)
seek_offset = dlen - self.READ_BLOCK_SIZE
diff --git a/src/binwalk/compat.py b/src/binwalk/compat.py
index e81e322..8212651 100644
--- a/src/binwalk/compat.py
+++ b/src/binwalk/compat.py
@@ -33,7 +33,7 @@ def str2bytes(string):
For cross compatibility between Python 2 and Python 3 strings.
'''
if isinstance(string, type('')) and sys.version_info.major > 2:
- return bytes(string, 'ascii')
+ return bytes(string, 'latin1')
else:
return string
@@ -42,7 +42,7 @@ def bytes2str(bs):
For cross compatibility between Python 2 and Python 3 strings.
'''
if isinstance(bs, type(b'')) and sys.version_info.major > 2:
- return bs.decode('ascii')
+ return bs.decode('latin1')
else:
return bs
diff --git a/src/binwalk/parser.py b/src/binwalk/parser.py
index 1a4468a..9e323c1 100644
--- a/src/binwalk/parser.py
+++ b/src/binwalk/parser.py
@@ -141,6 +141,7 @@ class MagicParser:
try:
for line in io.FileIO(file_name).readlines():
+ line = bytes2str(line)
line_count += 1
# Check if this is the first line of a signature entry
@@ -164,7 +165,7 @@ class MagicParser:
# Keep writing lines of the signature to the temporary magic file until
# we detect a signature that should not be included.
if include:
- self.fd.write(line)
+ self.fd.write(str2bytes(line))
self.build_signature_set()
except Exception as e:
@@ -191,14 +192,14 @@ class MagicParser:
# Quick and dirty pre-filter. We are only concerned with the first line of a
# signature, which will always start with a number. Make sure the first byte of
# the line is a number; if not, don't process.
- if bytes2str(line[:1]) < '0' or bytes2str(line[:1]) > '9':
+ if line[:1] < '0' or line[:1] > '9':
return None
try:
# Split the line into white-space separated parts.
# For this to work properly, replace escaped spaces ('\ ') with '\x20'.
# This means the same thing, but doesn't confuse split().
- line_parts = bytes2str(line).replace('\\ ', '\\x20').split()
+ line_parts = line.replace('\\ ', '\\x20').split()
entry['offset'] = line_parts[0]
entry['type'] = line_parts[1]
# The condition line may contain escaped sequences, so be sure to decode it properly.
@@ -281,7 +282,6 @@ class MagicParser:
Returns an ordered list of offsets inside of data at which candidate offsets were found.
'''
candidate_offsets = []
- data = bytes2str(data)
for regex in self.signature_set:
candidate_offsets += [match.start() for match in regex.finditer(data) if match.start() < end]
--
libgit2 0.26.0