diff --git a/tools/reduce.py b/tools/reduce.py
new file mode 100644
index 000000000..604304607
--- /dev/null
+++ b/tools/reduce.py
@@ -0,0 +1,233 @@
+
+import subprocess
+import os
+import sys
+
+CMD = None
+EXPECTED = None
+SEGFAULT = False
+FILE = None
+BACKUPFILE = None
+for arg in sys.argv[1:]:
+  if arg.startswith('--cmd='):
+    CMD = arg[arg.find('=')+1:]
+  elif arg.startswith('--expected='):
+    EXPECTED = arg[arg.find('=')+1:]
+  elif arg.startswith('--file='):
+    FILE = arg[arg.find('=')+1:]
+    BACKUPFILE = FILE + '.bak'
+  elif arg == '--segfault':
+    SEGFAULT = True
+
+if CMD is None:
+  print('Abort: No --cmd')
+  sys.exit(1)
+
+if SEGFAULT == False and EXPECTED is None:
+  print('Abort: No --expected')
+  sys.exit(1)
+
+if FILE is None:
+  print('Abort: No --file')
+  sys.exit(1)
+
+print('CMD='+CMD)
+if SEGFAULT:
+  print('EXPECTED=SEGFAULT')
+else:
+  print('EXPECTED='+EXPECTED)
+print('FILE='+FILE)
+
+def runtool():
+  p = subprocess.Popen(CMD.split(), stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+  comm = p.communicate()
+  if SEGFAULT:
+    if p.returncode != 0:
+      return True
+  elif p.returncode == 0:
+    out = comm[0] + '\n' + comm[1]
+    if (out.find('error:') < 0) and (out.find(EXPECTED) > 0):
+      return True
+  return False
+
+def writefile(filename, filedata):
+  f = open(filename, 'wt')
+  for line in filedata:
+    f.write(line)
+  f.close()
+
+def replaceandrun(what, filedata, i, line):
+  print(what + ' ' + str(i+1) + '/' + str(len(filedata)) + '..')
+  bak = filedata[i]
+  filedata[i] = line
+  writefile(FILE, filedata)
+  if runtool() == True:
+    print('pass')
+    writefile(BACKUPFILE, filedata)
+    return True
+  print('fail')
+  filedata[i] = bak
+  return False
+
+def replaceandrun2(what, filedata, i, line1, line2):
+  print(what + ' ' + str(i+1) + '/' + str(len(filedata)) + '..')
+  bak1 = filedata[i]
+  bak2 = filedata[i+1]
+  filedata[i] = line1
+  filedata[i+1] = line2
+  writefile(FILE, filedata)
+  if runtool() == True:
+    print('pass')
+    writefile(BACKUPFILE, filedata)
+  else:
+    print('fail')
+    filedata[i] = bak1
+    filedata[i+1] = bak2
+
+def removecomments(filedata):
+  for i in range(len(filedata)):
+    line = filedata[i]
+    if line.find('//') >= 0:
+      replaceandrun('remove comment', filedata, i, line[:line.find('//')].rstrip())
+
+def checkpar(line):
+  par = 0
+  for c in line:
+    if c=='(' or c=='[':
+      par = par + 1
+    elif c==')' or c==']':
+      par = par - 1
+      if par<0:
+        return False
+  return par == 0
+
+def combinelines(filedata):
+  if len(filedata) < 3:
+    return
+  for i in range(len(filedata)-1):
+    fd1 = filedata[i].rstrip()
+    if fd1.endswith(','):
+      fd2 = filedata[i+1].lstrip()
+      replaceandrun2('combine lines', filedata, i, fd1+fd2, '')
+
+def removeline(filedata):
+  stmt = True
+  for i in range(len(filedata)):
+    line = filedata[i]
+    strippedline = line.strip()
+
+    if len(strippedline) == 0:
+      continue
+
+    if stmt and strippedline[-1]==';' and checkpar(line) and line.find('{')<0 and line.find('}')<0:
+      replaceandrun('remove line', filedata, i, '')
+
+    elif stmt and strippedline.find('{') > 0 and strippedline.find('}') == len(strippedline) - 1:
+      replaceandrun('remove line', filedata, i, '')
+
+    if ';{}'.find(strippedline[-1]) >= 0:
+      stmt = True
+    else:
+      stmt = False
+
+
+def removeincludes(filedata):
+  for i in range(len(filedata)):
+    if filedata[i].startswith('#include '):
+      replaceandrun('remove #include', filedata, i, '')
+
+def removeemptyblocks(filedata):
+  if len(filedata) < 3:
+    return
+
+  i = 0
+  while i < len(filedata):
+    if filedata[i].strip() == '':
+      filedata.pop(i)
+    else:
+      i = i + 1
+
+  for i in range(len(filedata)-1):
+    fd1 = filedata[i].rstrip()
+    fd2 = filedata[i+1].strip()
+    if checkpar(fd1) and fd1.endswith('{') and fd2 == '}':
+      replaceandrun2('remove block', filedata, i, '', '')
+
+def removenamespaces(filedata):
+  if len(filedata) < 3:
+    return filedata
+
+  stmt = True
+
+  for i in range(len(filedata)):
+    line = filedata[i]
+    strippedline = line.strip()
+
+    if len(strippedline) == 0:
+      continue
+
+    stmt1 = stmt
+    if ';{}'.find(strippedline[-1]) >= 0:
+      stmt = True
+    else:
+      stmt = False
+
+    if stmt1 == False:
+      continue
+
+    if strippedline.find('}') < 0 and strippedline.find('{') == len(strippedline)-1:
+      i2 = i + 1
+      level = 1
+      while i2 < len(filedata) and level > 0:
+        #print(str(i2)+':'+str(level)+':'+filedata[i2])
+        for c in filedata[i2]:
+          if c == '}':
+            level = level - 1
+            if level <= 0:
+              break
+          elif c == '{':
+            level = level + 1
+        i2 = i2 + 1
+      if level == 0 and (filedata[i2-1].strip().endswith('}') or filedata[i2-1].strip().endswith('};')):
+        #print(str(i)+';'+str(i2))
+        filedata2 = list(filedata)
+        ii = i
+        while ii < i2:
+          filedata2[ii] = ''
+          ii = ii + 1
+        if replaceandrun('remove codeblock', filedata2, i, ''):
+          filedata = filedata2
+  return filedata
+
+# reduce..
+print('Make sure error can be reproduced...')
+if runtool() == False:
+     print("Cannot reproduce")
+     sys.exit(1)
+
+f = open(FILE, 'rt')
+filedata = f.readlines()
+f.close()
+
+writefile(BACKUPFILE, filedata)
+
+print('remove comments...')
+removecomments(filedata)
+print('combine lines..')
+combinelines(filedata)
+print('remove namespaces...')
+filedata = removenamespaces(filedata)
+print('remove includes...')
+removeincludes(filedata)
+print('remove line...')
+removeline(filedata)
+print('remove includes...')
+removeincludes(filedata)
+print('remove empty blocks...')
+removeemptyblocks(filedata)
+print('remove namespaces...')
+filedata = removenamespaces(filedata)
+
+
+
+writefile(FILE, filedata)