This commit was generated by cvs2svn to compensate for changes in r30,

which included commits to RCS files with non-trunk default branches.
(This used to be commit 3a69cffb062d4f1238b8cae10481c1f2ea4d3d8b)

4 files changed, 411 insertions, 0 deletions

diff --git a/source4/python/examples/tdbpack/.cvsignore b/source4/python/examples/tdbpack/.cvsignore
new file mode 100644
index 0000000000..52e4e61140
--- /dev/null
+++ b/source4/python/examples/tdbpack/.cvsignore
@@ -0,0 +1,2 @@
+*.pyc
+*.pyo
diff --git a/source4/python/examples/tdbpack/oldtdbutil.py b/source4/python/examples/tdbpack/oldtdbutil.py
new file mode 100644
index 0000000000..ac435b8bac
--- /dev/null
+++ b/source4/python/examples/tdbpack/oldtdbutil.py
@@ -0,0 +1,144 @@
+#!/usr/bin/python
+#############################################################
+# tdbutil
+# 
+# Purpose:
+#   Contains functions that are used to pack and unpack data
+# from Samba's tdb databases.  Samba sometimes represents complex
+# data structures as a single value in a database.  These functions
+# allow other python scripts to package data types into a single python
+# string and unpackage them.
+#
+#
+# XXXXX: This code is no longer used; it's just here for testing
+# compatibility with the new (much faster) C implementation.
+#
+############################################################## 
+import string
+
+def pack(format,list):
+   retstring = ''
+   listind = 0
+   
+   # Cycle through format entries
+   for type in format:
+      # Null Terminated String
+      if (type == 'f' or type == 'P'):
+         retstring = retstring + list[listind] + "\000"
+      # 4 Byte Number
+      if (type == 'd'):
+         retstring = retstring + PackNum(list[listind],4)
+      # 2 Byte Number
+      if (type == 'w'):
+         retstring = retstring + PackNum(list[listind],2)
+      # Pointer Value
+      if (type == 'p'):
+         if (list[listind]):
+            retstring = retstring + PackNum(1,4)
+         else:
+            retstring = retstring + PackNum(0,4)
+      # Buffer and Length
+      if (type == 'B'):
+         # length
+         length = list[listind]
+         retstring = retstring + PackNum(length,4)
+         length = int(length)
+         listind = listind + 1
+         # buffer
+         retstring = retstring + list[listind][:length]
+         
+      listind = listind + 1
+      
+   return retstring
+
+def unpack(format,buffer):
+   retlist = []
+   bufind = 0
+   
+   lasttype = ""
+   for type in format:
+      # Pointer Value
+      if (type == 'p'):
+         newvalue = UnpackNum(buffer[bufind:bufind+4])
+         bufind = bufind + 4
+         if (newvalue):
+            newvalue = 1L
+         else:
+            newvalue = 0L
+         retlist.append(newvalue)
+      # Previous character till end of data
+      elif (type == '$'):
+         if (lasttype == 'f'):
+            while (bufind < len(buffer)):
+               newstring = ''
+               while (buffer[bufind] != '\000'):
+                  newstring = newstring + buffer[bufind]
+                  bufind = bufind + 1
+               bufind = bufind + 1
+               retlist.append(newstring)
+      # Null Terminated String
+      elif (type == 'f' or type == 'P'):
+         newstring = ''
+         while (buffer[bufind] != '\000'):
+            newstring = newstring + buffer[bufind]
+            bufind = bufind + 1
+         bufind = bufind + 1
+         retlist.append(newstring)
+      # 4 Byte Number
+      elif (type == 'd'):
+         newvalue = UnpackNum(buffer[bufind:bufind+4])
+         bufind = bufind + 4
+         retlist.append(newvalue)
+      # 2 Byte Number
+      elif (type == 'w'):
+         newvalue = UnpackNum(buffer[bufind:bufind+2])
+         bufind = bufind + 2
+         retlist.append(newvalue)
+      # Length and Buffer
+      elif (type == 'B'):
+         # Length
+         length = UnpackNum(buffer[bufind:bufind+4])
+         bufind = bufind + 4 
+         retlist.append(length)
+         length = int(length)
+         # Buffer
+         retlist.append(buffer[bufind:bufind+length])
+         bufind = bufind + length
+         
+      lasttype = type
+
+   return ((retlist,buffer[bufind:]))
+
+def PackNum(myint,size):
+    retstring = ''
+    size = size * 2
+    hint = hex(myint)[2:]
+
+    # Check for long notation
+    if (hint[-1:] == 'L'):
+       hint = hint[:-1]
+    
+    addon = size - len(hint)
+    for i in range(0,addon):
+       hint = '0' + hint
+    
+    while (size > 0):
+       val = string.atoi(hint[size-2:size],16)
+       retstring = retstring + chr(val)
+       size = size - 2
+    
+    return retstring
+   
+def UnpackNum(buffer):
+   size = len(buffer)
+   mystring = ''
+
+   for i in range(size-1,-1,-1):
+      val = hex(ord(buffer[i]))[2:]
+      if (len(val) == 1):
+         val = '0' + val
+      mystring = mystring + val
+   if (len(mystring) > 4):
+      return string.atol(mystring,16)
+   else:
+      return string.atoi(mystring,16)
diff --git a/source4/python/examples/tdbpack/tdbtimetrial.py b/source4/python/examples/tdbpack/tdbtimetrial.py
new file mode 100755
index 0000000000..be6404899d
--- /dev/null
+++ b/source4/python/examples/tdbpack/tdbtimetrial.py
@@ -0,0 +1,12 @@
+#! /usr/bin/python2.2
+
+def run_trial():
+    # import tdbutil
+    from samba.tdbpack import pack
+    
+    for i in xrange(500000):
+        pack("ddffd", (10, 2, "mbp", "martin", 0))
+        #s = "\n\0\0\0" + "\x02\0\0\0" + "mbp\0" + "martin\0" + "\0\0\0\0"
+
+if __name__ == '__main__':
+    run_trial()
diff --git a/source4/python/examples/tdbpack/test_tdbpack.py b/source4/python/examples/tdbpack/test_tdbpack.py
new file mode 100755
index 0000000000..837600f789
--- /dev/null
+++ b/source4/python/examples/tdbpack/test_tdbpack.py
@@ -0,0 +1,253 @@
+#! /usr/bin/env python2.2
+
+__doc__ = """test case for samba.tdbpack functions
+
+tdbpack provides a means of pickling values into binary formats
+compatible with that used by the samba tdbpack()/tdbunpack()
+functions.
+
+Numbers are always stored in little-endian format; strings are stored
+in either DOS or Unix codepage as appropriate.
+
+The format for any particular element is encoded as a short ASCII
+string, with one character per field."""
+
+# Copyright (C) 2002 Hewlett-Packard.
+
+__author__ = 'Martin Pool <mbp@sourcefrog.net>'
+
+import unittest
+import oldtdbutil
+import samba.tdbpack
+
+both_unpackers = (samba.tdbpack.unpack, oldtdbutil.unpack)
+both_packers = (samba.tdbpack.pack, oldtdbutil.pack)
+
+
+
+# #             ('B', [10, 'hello'], '\x0a\0\0\0hello'),
+#              ('BB', [11, 'hello\0world', 3, 'now'],
+#               '\x0b\0\0\0hello\0world\x03\0\0\0now'),
+#              ('pd', [1, 10], '\x01\0\0\0\x0a\0\0\0'),
+#              ('BBB', [5, 'hello', 0, '', 5, 'world'],
+#               '\x05\0\0\0hello\0\0\0\0\x05\0\0\0world'),
+
+             # strings are sequences in Python, there's no getting away
+             # from it
+#             ('ffff', 'evil', 'e\0v\0i\0l\0'),
+#              ('BBBB', 'evil',                   
+#               '\x01\0\0\0e'
+#               '\x01\0\0\0v'
+#               '\x01\0\0\0i'
+#               '\x01\0\0\0l'),
+
+#              ('', [], ''),
+
+#              # exercise some long strings
+#              ('PP', ['hello' * 255, 'world' * 255],
+#               'hello' * 255 + '\0' + 'world' * 255 + '\0'),
+#              ('PP', ['hello' * 40000, 'world' * 50000],
+#               'hello' * 40000 + '\0' + 'world' * 50000 + '\0'),
+#              ('B', [(5*51), 'hello' * 51], '\xff\0\0\0' + 'hello' * 51),
+#              ('BB', [(5 * 40000), 'hello' * 40000,
+#                      (5 * 50000), 'world' * 50000],
+#               '\x40\x0d\x03\0' + 'hello' * 40000 + '\x90\xd0\x03\x00' + 'world' * 50000),
+
+    
+class PackTests(unittest.TestCase):
+    symm_cases = [
+             ('w', [42], '\x2a\0'),
+             ('www', [42, 2, 69], '\x2a\0\x02\0\x45\0'),
+             ('wd', [42, 256], '\x2a\0\0\x01\0\0'),
+             ('w', [0], '\0\0'),
+             ('w', [255], '\xff\0'),
+             ('w', [256], '\0\x01'),
+             ('w', [0xdead], '\xad\xde'),
+             ('w', [0xffff], '\xff\xff'),
+             ('p', [0], '\0\0\0\0'),
+             ('p', [1], '\x01\0\0\0'),
+             ('d', [0x01020304], '\x04\x03\x02\x01'),
+             ('d', [0x7fffffff], '\xff\xff\xff\x7f'),
+             ('d', [0x80000000L], '\x00\x00\x00\x80'),
+             ('d', [0x80000069L], '\x69\x00\x00\x80'),
+             ('d', [0xffffffffL], '\xff\xff\xff\xff'),
+             ('d', [0xffffff00L], '\x00\xff\xff\xff'),
+             ('ddd', [1, 10, 50], '\x01\0\0\0\x0a\0\0\0\x32\0\0\0'),
+             ('ff', ['hello', 'world'], 'hello\0world\0'),
+             ('fP', ['hello', 'world'], 'hello\0world\0'),
+             ('PP', ['hello', 'world'], 'hello\0world\0'),
+             ('B', [0, ''], '\0\0\0\0'),
+# old implementation is wierd when string is not the right length             
+#             ('B', [2, 'hello'], '\x0a\0\0\0hello'),
+             ('B', [5, 'hello'], '\x05\0\0\0hello'),
+             ]
+
+    def test_symmetric(self):
+        """Cookbook of symmetric pack/unpack tests
+        """
+        for packer in [samba.tdbpack.pack]: # both_packers:
+            for unpacker in both_unpackers:
+                for format, values, expected in self.symm_cases:
+                    out_packed = packer(format, values)
+                    self.assertEquals(out_packed, expected)
+                    out, rest = unpacker(format, expected)
+                    self.assertEquals(rest, '')
+                    self.assertEquals(list(values), list(out))
+
+    def test_large(self):
+        """Test large pack/unpack strings"""
+        large_cases = [('w' * 1000, xrange(1000)), ]
+        for packer in both_packers:
+            for unpacker in both_unpackers:
+                for format, values in large_cases:
+                    packed = packer(format, values)
+                    out, rest = unpacker(format, packed)
+                    self.assertEquals(rest, '')
+                    self.assertEquals(list(values), list(out))
+
+                    
+    def test_pack(self):
+        """Cookbook of expected pack values
+
+        These can't be used for the symmetric test because the unpacked value is
+        not "canonical".
+        """
+        cases = [('w', (42,), '\x2a\0'),
+                 ]
+
+        for packer in both_packers:
+            for format, values, expected in cases:
+                self.assertEquals(packer(format, values), expected)
+
+    def test_unpack_extra(self):
+        # Test leftover data
+        for unpacker in both_unpackers:
+            for format, values, packed in self.symm_cases:
+                out, rest = unpacker(format, packed + 'hello sailor!')
+                self.assertEquals(rest, 'hello sailor!')
+                self.assertEquals(list(values), list(out))
+
+
+    def test_pack_extra(self):
+        """Leftover values when packing"""
+        cases = [
+            ('d', [10, 20], [10]),
+            ('d', [10, 'hello'], [10]),
+            ('ff', ['hello', 'world', 'sailor'], ['hello', 'world']),
+            ]
+        for unpacker in both_unpackers:
+            for packer in both_packers:
+                for format, values, chopped in cases:
+                    bin = packer(format, values)
+                    out, rest = unpacker(format, bin)
+                    self.assertEquals(list(out), list(chopped))
+                    self.assertEquals(rest, '')
+
+
+    def test_unpack(self):
+        """Cookbook of tricky unpack tests"""
+        cases = [
+                 # Apparently I couldn't think of any tests that weren't
+                 # symmetric :-/
+                 ]
+        for unpacker in both_unpackers:
+            for format, values, expected in cases:
+                out, rest = unpacker(format, expected)
+                self.assertEquals(rest, '')
+                self.assertEquals(list(values), list(out))
+
+
+    def test_pack_failures(self):
+        """Expected errors for incorrect packing"""
+        cases = [('w', []),
+#                 ('w', ()),
+#                 ('w', {}),
+                 ('ww', [2]),
+                 ('w', 2),
+#                  ('w', None),
+                 ('wwwwwwwwwwww', []),
+#                 ('w', [0x60A15EC5L]),
+#                 ('w', [None]),
+                 ('d', []),
+                 ('p', []),
+                 ('f', [2]),
+                 ('P', [None]),
+                 ('P', ()),
+                 ('f', [hex]),
+                 ('fw', ['hello']),
+#                  ('f', [u'hello']),
+                 ('B', [2]),
+                 (None, [2, 3, 4]),
+                 (ord('f'), [20]),
+                 # old code doesn't distinguish string from seq-of-char
+#                 (['w', 'w'], [2, 2]),
+                 # old code just ignores invalid characters
+#                 ('Q', [2]),
+#                 ('fQ', ['2', 3]),
+#                 ('fQ', ['2']),
+                 (2, [2]),
+                 # old code doesn't typecheck format
+#                 ({}, {})
+                 ]
+        for packer in both_packers:
+            for format, values in cases:
+                try:
+                    packer(format, values)
+                except StandardError:
+                    pass
+                else:
+                    raise AssertionError("didn't get exception: format %s, values %s, packer %s"
+                                         % (`format`, `values`, `packer`))
+
+
+    def test_unpack_failures(self):
+        """Expected errors for incorrect unpacking"""
+        cases = [
+# This ought to be illegal, but the old code doesn't prohibit it
+#                ('$', '', ValueError),
+#                ('Q', '', ValueError),
+#                ('Q$', '', ValueError),
+                 ('f', '', IndexError),
+                 ('d', '', IndexError),
+# This is an illegal packing, but the old code doesn't trap                 
+#                 ('d', '2', IndexError),
+#                 ('d', '22', IndexError),
+#                 ('d', '222', IndexError),
+#                 ('p', '\x01\0', IndexError),
+#                ('w', '2', IndexError),
+#                ('B', '\xff\0\0\0hello', IndexError),
+#                  ('B', '\xff\0', IndexError),
+                 ('w', '', IndexError),
+                 ('f', 'hello', IndexError),
+                 ('f', '', IndexError),
+#                ('B', '\x01\0\0\0', IndexError),
+#                 ('B', '\x05\0\0\0hell', IndexError),
+                 ('B', '\xff\xff\xff\xff', ValueError),
+#                 ('B', 'foobar', IndexError),
+#                 ('BB', '\x01\0\0\0a\x01', IndexError),
+                 ]
+
+        for unpacker in both_unpackers:
+            for format, values, throwable_class in cases:
+                try:
+                    unpacker(format, values)
+                except StandardError:
+                    pass
+                else:
+                    raise AssertionError("didn't get exception: format %s, values %s, unpacker %s"
+                                         % (`format`, `values`, `unpacker`))
+
+    def test_unpack_repeated(self):
+        cases = [(('df$',
+                  '\x00\x00\x00\x00HP C LaserJet 4500-PS\x00Windows 4.0\x00\\print$\\WIN40\\0\\PSCRIPT.DRV\x00\\print$\\WIN40\\0\\PSCRIPT.DRV\x00\\print$\\WIN40\\0\\PSCRIPT.DRV\x00\\print$\\WIN40\\0\\PSCRIPT.HLP\x00\x00RAW\x00\\print$\\WIN40\\0\\readme.wri\x00\\print$\\WIN40\\0\\pscript.drv\x00\\print$\\WIN40\\0\\pscript.hlp\x00'),
+                 ([0L, 'HP C LaserJet 4500-PS', 'Windows 4.0', '\\print$\\WIN40\\0\\PSCRIPT.DRV', '\\print$\\WIN40\\0\\PSCRIPT.DRV', '\\print$\\WIN40\\0\\PSCRIPT.DRV', '\\print$\\WIN40\\0\\PSCRIPT.HLP', '', 'RAW', '\\print$\\WIN40\\0\\readme.wri', '\\print$\\WIN40\\0\\pscript.drv', '\\print$\\WIN40\\0\\pscript.hlp'], ''))]
+        for unpacker in both_unpackers:
+            for input, expected in cases:
+                result = apply(unpacker, input)
+                if result != expected:
+                    raise AssertionError("%s:\n     input: %s\n    output: %s\n  expected: %s" % (`unpacker`, `input`, `result`, `expected`))
+        
+
+if __name__ == '__main__':
+    unittest.main()
+