Unify interface. Use python logger. Optimize.

tools/dumphunks.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python2.7 -B
 
 import sys
-from objtools.hunk import HunkParser
+from objtools import hunk
 
 
 if __name__ == '__main__':
@@ -9,6 +9,6 @@ if __name__ == '__main__':
     print 'Parsing "%s".' % path
     print ''
 
-    for hunk in HunkParser(path):
-      hunk.dump()
+    for h in hunk.ReadFile(path):
+      h.dump()
       print ''

tools/objtools/aout.py

@@ -1,4 +1,5 @@
 from collections import namedtuple, Sequence
+import logging
 import os
 import struct
 from util import hexdump
@@ -12,6 +13,9 @@ class Header(namedtuple('Header', ('mid', 'magic', 'text', 'data', 'bss',
 
   @classmethod
   def decode(cls, data):
+    if len(data) != 32:
+      raise ValueError('Not a valid a.out header!')
+
     mid, magic, text, data, bss, syms, entry, trsize, drsize = \
         struct.unpack('>HHIIIIIII', data)
 
@@ -169,8 +173,8 @@ class Aout(object):
     self._path = path
 
     with open(path) as data:
-      print 'Reading {0} of size {1} bytes.'.format(
-        repr(path), os.stat(path)[6])
+      logging.debug('Reading %r of size %d bytes.',
+                    path, os.stat(path).st_size)
 
       self._header = Header.decode(data.read(32))
 
@@ -183,7 +187,7 @@ class Aout(object):
       strings = data.read()
 
       if str_size != len(strings):
-        print 'Warning: wrong size of string table!'
+        logging.warn('Wrong size of string table!')
 
       self._strings = StringTable.decode(strings)
 
@@ -226,3 +230,9 @@ class Aout(object):
       for reloc in self._text_relocs:
         print ' ', reloc.as_string(self._strings)
       print ''
+
+
+def ReadFile(path):
+  aout = Aout()
+  aout.read(path)
+  return aout

tools/objtools/ar.py

@@ -1,3 +1,4 @@
+import logging
 import os
 import struct
 
@@ -12,7 +13,7 @@ class ArEntry(namedtuple('ArEntry',
         struct.unpack('16s12s6s6s8s10s2s', ar.read(60))
 
     length = int(length.strip())
-    modtime = int(modtime.strip())
+    modtime = int(modtime.strip() or '0')
     owner = int(owner.strip() or '0')
     group = int(group.strip() or '0')
     mode = mode.strip() or '100644'
@@ -38,11 +39,18 @@ def ReadFile(path):
 
   with open(path) as ar:
     if ar.read(8) != '!<arch>\n':
-      raise RuntimeError('%s is not an ar archive', path)
+      raise ValueError('%s is not an ar archive' % path)
 
     ar_size = os.stat(path).st_size
 
+    logging.debug('Reading ar archive %r of size %d bytes.', path, ar_size)
+
     while ar.tell() < ar_size:
-      entries.append(ArEntry.decode(ar))
+      # Some archives have version information attached at the end of file,
+      # that confuses ArEntry parser, so just skip it.
+      try:
+        entries.append(ArEntry.decode(ar))
+      except struct.error:
+        break
 
   return entries

tools/objtools/hunk.py

@@ -1,3 +1,4 @@
+import logging
 import os
 import struct
 import textwrap
@@ -35,6 +36,8 @@ HunkMap = {
   'HUNK_ABSRELOC16': 1022
 }
 
+HunkMapRev = dict((v, k) for k, v in HunkMap.items())
+
 HunkExtMap = {
   'EXT_SYMB': 0,          # symbol table
   'EXT_DEF': 1,           # relocatable definition
@@ -53,6 +56,8 @@ HunkExtMap = {
   'EXT_ABSREF8': 139      # 8 bit absolute reference to symbol
 }
 
+HunkExtMapRev = dict((v, k) for k, v in HunkExtMap.items())
+
 # Any hunks that have the HUNKB_ADVISORY bit set will be ignored if they
 # aren't understood.  When ignored, they're treated like HUNK_DEBUG hunks.
 # NOTE: this handling of HUNKB_ADVISORY started as of V39 dos.library!  If
@@ -80,11 +85,10 @@ class Hunk(object):
   def getType(number):
     number &= 0x1fffffff
 
-    for name, value in HunkMap.items():
-      if value == number:
-        return name
-
-    raise ValueError('Unknown Hunk: %d' % number)
+    try:
+      return HunkMapRev[number]
+    except KeyError:
+      raise ValueError('Unknown Hunk: %d' % number)
 
   @staticmethod
   def getFlags(number):
@@ -324,11 +328,10 @@ class HunkExt(Hunk):
 
   @staticmethod
   def getType(number):
-    for name, value in HunkExtMap.items():
-      if value == number:
-        return name
-
-    raise ValueError('Unknown HunkExt: %d' % number)
+    try:
+      return HunkExtMapRev[number]
+    except KeyError:
+      raise ValueError('Unknown HunkExt: %d' % number)
 
   @classmethod
   def parse(cls, hf):
@@ -594,7 +597,7 @@ HunkClassMap = {
 }
 
 
-def HunkParser(path):
+def ReadFile(path):
   with HunkFile(path) as hf:
     hunks = []
     units = 0
@@ -621,7 +624,7 @@ def HunkParser(path):
       try:
         hunks.append(hunk.parse(hf))
       except ValueError:
-        print 'Parse error at position 0x%x.' % hf.tell()
+        logging.error('Parse error at position 0x%x.', hf.tell())
         util.hexdump(hf.read())
 
     return hunks