python3 fixes

Author: angea

scripts/jp2.py

@@ -33,7 +33,7 @@ def isValid(d):
 l2 = len(d2)
 
 suffix = b"".join([
-  struct.pack(">I", 0x100 + 8),b "free",
+  struct.pack(">I", 0x100 + 8),b"free",
   b"\0" * (0x100 - 8),
   struct.pack(">I", 8 + l1), b"free",
   d1,

scripts/jpg.py

@@ -28,7 +28,7 @@ def comments(s, delta=0):
   block2 = f.read()
 
 # skip the signature, split by scans (usually the biggest segments)
-c1 = d1[2:].split("\xff\xda")
+c1 = d1[2:].split(b"\xff\xda")
 
 if max(len(i) for i in c1) >= 65536 - 8:
   print("ERROR: The first image file has a segment that is too big!" +
@@ -53,7 +53,7 @@ def comments(s, delta=0):
 \\==============/
 vvvvvvvvvvvvvvvv""".splitlines())
 
-suffix = "".join([
+suffix = b"".join([
   # fake comment to jump over the first image chunk (typically small)
   b"\xff\xfe",
     struct.pack(">H", 0x100 + len(c1[0]) - 2 + 4),

scripts/pdf.py

@@ -18,36 +18,36 @@ def EnclosedString(d, starts, ends):
   return d[off:d.find(ends, off)]
 
 def getCount(d):
-  s = EnclosedString(d, "/Count ", "/")
+  s = EnclosedString(d, b"/Count ", b"/")
   count = int(s)
   return count
 
 def procreate(l): # :p
-  return " 0 R ".join(l) + " 0 R"
+  return b" 0 R ".join(l) + b" 0 R"
 
 def adjustPDF(contents):
   """dumb [start]xref fix: fixes old-school xref with no holes, with hardcoded \\n"""
-  startXREF = contents.find("\nxref\n0 ") + 1
-  endXREF = contents.find(" \n\n", startXREF) + 1
+  startXREF = contents.find(b"\nxref\n0 ") + 1
+  endXREF = contents.find(b" \n\n", startXREF) + 1
   origXref = contents[startXREF:endXREF]
-  objCount = int(origXref.splitlines()[1].split(" ")[1])
+  objCount = int(origXref.splitlines()[1].split(b" ")[1])
   print("object count: %i" % objCount)
 
   xrefLines = [
-    "xref",
-    "0 %i" % objCount,
+    b"xref",
+    b"0 %i" % objCount,
     # mutool declare its first xref like this
-    "0000000000 00001 f "
+    b"0000000000 00001 f "
     ]
 
   i = 1
   while i < objCount:
     # doesn't support comments at the end of object declarations
-    off = contents.find("\n%i 0 obj\n" % i) + 1
-    xrefLines.append("%010i 00000 n " % (off))
+    off = contents.find(b"\n%i 0 obj\n" % i) + 1
+    xrefLines.append(b"%010i 00000 n " % (off))
     i += 1
 
-  xref = "\n".join(xrefLines)
+  xref = b"\n".join(xrefLines)
 
   # XREF length should be unchanged
   try:
@@ -58,8 +58,8 @@ def adjustPDF(contents):
 
   contents = contents[:startXREF] + xref + contents[endXREF:]
 
-  startStartXref = contents.find("\nstartxref\n", endXREF) + len("\nstartxref\n")
-  endStartXref = contents.find("\n%%%%EOF", startStartXref)
+  startStartXref = contents.find(b"\nstartxref\n", endXREF) + len(b"\nstartxref\n")
+  endStartXref = contents.find(b"\n%%%%EOF", startStartXref)
   contents = contents[:startStartXref] + "%i" % startXREF + contents[endStartXref:]
 
   return contents
@@ -88,12 +88,12 @@ def adjustPDF(contents):
 COUNT2 = getCount(d2)
 
 
-kids = EnclosedString(dm, "/Kids[", "]")
+kids = EnclosedString(dm, b"/Kids[", b"]")
 
 # we skip the first dummy, and the last " 0 R" string
-pages = kids[:-4].split(" 0 R ")[1:]
+pages = kids[:-4].split(b" 0 R ")[1:]
 
-template = """%%PDF-1.4
+template = b"""%%PDF-1.4
 
 1 0 obj
 <<
@@ -132,7 +132,7 @@ def adjustPDF(contents):
 contents = template % locals()
 
 # adjust parents for the first set of pages
-contents += dm[dm.find("5 0 obj"):].replace("/Parent 2 0 R", "/Parent 3 0 R", COUNT1)
+contents += dm[dm.find(b"5 0 obj"):].replace(b"/Parent 2 0 R", b"/Parent 3 0 R", COUNT1)
 
 # not necessary (will be fixed by mutool anyway) but avoids warnings
 contents = adjustPDF(contents)
@@ -148,8 +148,8 @@ def adjustPDF(contents):
 
 # some mutool versions do different stuff :(
 cleaned = cleaned.replace(
-  " 65536 f \n0000000016 00000 n \n",
-  " 65536 f \n0000000018 00000 n \n",
+  b" 65536 f \n0000000016 00000 n \n",
+  b" 65536 f \n0000000018 00000 n \n",
   1)
 
 with open("pdf1.bin", "rb") as f:
@@ -158,8 +158,8 @@ def adjustPDF(contents):
 with open("pdf2.bin", "rb") as f:
   prefix2 = f.read()
 
-file1 = prefix1 + "\n" + cleaned[192:]
-file2 = prefix2 + "\n" + cleaned[192:]
+file1 = prefix1 + b"\n" + cleaned[192:]
+file2 = prefix2 + b"\n" + cleaned[192:]
 
 with open("collision1.pdf", "wb") as f:
   f.write(file1)

scripts/pdfpe.py

@@ -18,11 +18,11 @@ def EnclosedString(d, starts, ends):
   return d[off:d.find(ends, off)]
 
 def getCount(d):
-  s = EnclosedString(d, "/Count ", "/")
+  s = EnclosedString(d, b"/Count ", b"/")
   count = int(s)
   return count
 
-template = """%%PDF-1.3
+template = b"""%%PDF-1.3
 %%\xC2\xB5\xC2\xB6
 
 1 0 obj
@@ -135,7 +135,7 @@ def relocateSections(d, SecTblOff, SecCount, delta):
 
 dm = dm[dm.find(b"5 0 obj"):]
 dm = dm.replace(b"/Parent 2 0 R", b"/Parent 4 0 R")
-dm = dm.replace("/Root 1 0 R", b"/Root 3 0 R")
+dm = dm.replace(b"/Root 1 0 R", b"/Root 3 0 R")
 
 pe = relocateSections(pe, SecTblOff, NumSec, ALIGN - SectsStart)
 Sections = pe[SectsStart - SECTIONEXTRA:]

scripts/pe.py

@@ -14,7 +14,7 @@
 import sys, struct, hashlib, os
 
 def getPEhdr(d):
-  PEoffset = d.find("PE\0\0")
+  PEoffset = d.find(b"PE\0\0")
   peHDR = d[PEoffset:]
 
   Machine = struct.unpack("H", peHDR[4:4+2])[0]
@@ -78,8 +78,8 @@ def relocateSections(d, SecTblOff, SecCount, delta):
 
 
 # best PE validation ever :p
-assert d1.startswith("MZ")
-assert d2.startswith("MZ")
+assert d1.startswith(b"MZ")
+assert d2.startswith(b"MZ")
 
 PEoff1, HdrLen1, NumSec1, SecTblOff1, SectsStart1 = getPEhdr(d1)
 PEoff2, HdrLen2, NumSec2, SecTblOff2, SectsStart2 = getPEhdr(d2)
@@ -110,9 +110,9 @@ def relocateSections(d, SecTblOff, SecCount, delta):
 
 suffix = b"".join([
   d1[PEoff1:PEoff1+HdrLen1],
-    (HEADER2OFF - (HEADER1OFF + HdrLen1)) * "\0",
+    (HEADER2OFF - (HEADER1OFF + HdrLen1)) * b"\0",
   d2[PEoff2:PEoff2+HdrLen2],
-    (SECTIONSOFF - (HEADER2OFF + HdrLen2)) * "\0",
+    (SECTIONSOFF - (HEADER2OFF + HdrLen2)) * b"\0",
   Sections1,
   Sections2
   ])

scripts/pileup.py

@@ -20,7 +20,7 @@ def dprint(s):
 
 def setDWORD(d, offset, s):
   assert len(s) == 4
-  return "".join([
+  return b"".join([
     d[:offset],
     s,
     d[offset+4:]
@@ -33,10 +33,10 @@ def relocateMP4(d, delta):
   # finds and relocates all Sample Tables Chunk Offset tables
   # TODO: support 64 bits `co64` tables ()
   offset = 0
-  tablecount = d.count("stco")
+  tablecount = d.count(b"stco")
   dprint("stco found: %i" % tablecount)
   for i in range(tablecount):
-    offset = d.find("stco", offset)
+    offset = d.find(b"stco", offset)
     dprint("current offset: %0X" % offset)
 
     length   = struct.unpack(">I", d[offset-4:offset])[0]
@@ -76,11 +76,11 @@ def EnclosedString(d, starts, ends):
   return d[off:d.find(ends, off)]
 
 def getCount(d):
-  s = EnclosedString(d, "/Count ", "/")
+  s = EnclosedString(d, b"/Count ", b"/")
   count = int(s)
   return count
 
-template = """%%PDF-1.3
+template = b"""%%PDF-1.3
 %%\xC2\xB5\xC2\xB6
 
 1 0 obj
@@ -111,7 +111,7 @@ def getCount(d):
 
 
 def getPEhdr(d):
-  PEoffset = d.find("PE\0\0")
+  PEoffset = d.find(b"PE\0\0")
   peHDR = d[PEoffset:]
 
   Machine = struct.unpack("H", peHDR[4:4+2])[0]
@@ -199,13 +199,13 @@ def relocateSections(d, SecTblOff, SecCount, delta):
 
 buffer = b"".join([
   pe[PEoff:PEoff+HdrLen],
-    (ALIGN - HdrLen - PEOFFSET - SECTIONEXTRA) * "\0",
+    (ALIGN - HdrLen - PEOFFSET - SECTIONEXTRA) * b"\0",
   Sections,
   ])
 
 lenPE = len(buffer) # for PNG file
 
-buffer += "CRC3" + png
+buffer += b"CRC3" + png
 lenPEPNG = len(buffer) # for MP4 file
 
 buffer += relocateMP4(mp4, lenPEPNG + 0x560) # hardcoded offset of the PE in our computed prefix
@@ -245,7 +245,7 @@ def relocateSections(d, SecTblOff, SecCount, delta):
 # MP4 free atom ###############################################################
 
 cleaned = setDWORD(cleaned, 0x540,struct.pack(">L", lenPEPNG + 0x20))
-cleaned = setDWORD(cleaned, 0x544, "free")
+cleaned = setDWORD(cleaned, 0x544, b"free")
 
 
 # PNG fixes 

scripts/pngStd.py

@@ -43,7 +43,7 @@ def get_data(args):
     # 00-20 - original common header
     d1[:0x21],
     # 21-46 - padding chunk
-    b"\0\0\0\x1a", b"aNGE", b":MD5 ISREALLY DEAD NOW!!1!", "ROFL",
+    b"\0\0\0\x1a", b"aNGE", b":MD5 ISREALLY DEAD NOW!!1!", b"ROFL",
 
     # 47-C7 - collision chunk