Our Enterprise Engine has been specifically designed to facilitate the development of enterprise solutions such as cloud or in-house services. Our engine offers the same SDK as our Cerbero Suite Advanced and has already analyzed billions of files.
Security
Our engine has been designed taking into account any type of security issue when analyzing malicious files: buffer overflows, integer overflows, infinite loops, infinite recursion, decompression bombs, denial-of-service etc. It has successfully and safely analyzed billions of files throughout the years.
Cross-Platform
Like Cerbero Suite our Enterprise Engine is cross-platform. Currently we offer our Enterprise Engine solution for both Windows (x86, x64) and Linux (x64). It is also compatible with older version of Windows (e.g. XP) and Linux (e.g. Ubuntu 12).
Documentation
We provide in-depth documentation for our SDK. The documentation not only covers the API, but details key concepts and includes many code examples.
Embedding
Our Enterprise Engine is deployed as an embeddable module: a Dynamic-Link Library on Windows and a Shared Library on Linux. It’s possible to embed the engine from both C/C++ and Python 3.
Embedding the engine from Python is extremely simple:
from ProEngine import * # initialize the engine proEngineInit() # from now on the SDK can be accessed from Pro.PDF import * # ... # finalize the engine before exiting proEngineFinal()
Embedding the engine from C/C++ is also very simple: it only requires including the ProEngine header and specifying the location of the engine on disk:
#define PRO_ENGINE_INIT
#include "ProEngine.h"
int main()
{
/* initialize the engine */
if (!proEngineInit("/path/to/the/engine", ProEngine_InitPython))
return -1;
/* from now on the SDK can be accessed */
/* finalize the engine before exiting */
proEngineFinal();
return 0;
}
Simplicity
Our SDK is not only intuitive, but also flexible. For example, extracting JavaScript code from a PDF document can be achieved by creating a hook extension. The following code snippet is all that is necessary to extract JavaScript code from any PDF document, even if encrypted or contained in an archive.
from Pro.Core import *
def printJSEntry(sp, xml, tnode):
# data node
dnode = xml.findChild(tnode, "d")
if not dnode:
return
# we let Cerbero extract the JavaScript for us
params = NTStringVariantHash()
params.insert("op", "js")
idnode = xml.findChild(dnode, "id")
if idnode:
params.insert("id", int(xml.value(idnode), 16))
ridnode = xml.findChild(dnode, "rid")
if idnode:
params.insert("rid", int(xml.value(ridnode), 16))
js = sp.customOperation(params)
# print out the JavaScript
print("JS CODE")
print("-------")
print(js)
def pdfExtractJS(sp, ud):
xml = sp.getReportXML()
# object node
onode = xml.findChild(None, "o")
if onode:
# scan node
snode = xml.findChild(onode, "s")
if snode:
# enumerate scan entries
tchild = xml.firstChild(snode)
while tchild:
if xml.name(tchild) == "t":
# type attribute
tattr = xml.findAttribute(tchild, "t")
# check if it's a JavaScript entry
if tattr and int(xml.value(tattr)) == CT_JavaScript:
printJSEntry(sp, xml, tchild)
tchild = xml.nextSibling(tchild)
By using hooks and other extensions our scanning engine does most of the heavy lifting!
Alternatively, our SDK can be used to perform custom parsing operations, without relying on the scanning engine. The following code snippet parses a PDF document and for every PDF object it prints out the ID, its dictionary and the decoded stream content. The code even contains logic to detect unreferenced objects, which can be found in corrupted or malicious PDF documents, and decrypts encrypted PDF documents.
# iterate through all the objects in a PDF
from ProEngine import *
def parsePDF(fname):
# open the file
c = createContainerFromFile(fname)
if c.isNull():
print("error: couldn't open file")
return
# load the file as PDF
pdf = PDFObject()
if not pdf.Load(c):
print("error: invalid file format")
return
# parse all referenced objects
objtable = pdf.BuildObjectTable()
# detect unreferenced objects
# (corrupted or malicious PDFs may contain them)
pdf.DetectObjects(objtable)
# store the object table internally
pdf.SetObjectTable(objtable)
# process PDF encryption
if not pdf.ProcessEncryption():
print("warning: couldn't decrypt file")
# [optional] sort objects by ID
oids = []
it = objtable.iterator()
while it.hasNext():
oid, _ = it.next()
oids.append(oid)
oids.sort()
# iterate through the objects
for oid in oids:
# print out the object id
print("\nOBJECT ID:", oid >> 32, "\n")
# parse the object
ret, dictn, content, info = pdf.ParseObject(objtable, oid)
if not ret:
print("warning: couldn't parse object %d" % (oid,))
continue
# print out the object dictionary
it = dictn.iterator()
while it.hasNext():
k, v = it.next()
print(" ", k, "-", v)
# print out the decoded object stream
content = pdf.DecodeObjectStream(content, dictn, oid)
if not content:
continue
out = NTTextBuffer()
out.printHex(content)
print("\n", out.buffer)
if __name__ == "__main__":
import sys
proEngineInit()
from Pro.Core import *
from Pro.PDF import *
parsePDF(sys.argv[1])
proEngineFinal()
This is the output of a single PDF object:
OBJECT ID: 50
/Length - 371
/Filter - /FlateDecode
0 1 2 3 4 5 6 7 8 9 A B C D E F Ascii
0000 2F 43 49 44 49 6E 69 74 2F 50 72 6F 63 53 65 74 /CIDInit/ProcSet
0010 20 66 69 6E 64 72 65 73 6F 75 72 63 65 20 62 65 findresource be
0020 67 69 6E 0A 31 32 20 64 69 63 74 20 62 65 67 69 gin.12 dict begi
0030 6E 0A 62 65 67 69 6E 63 6D 61 70 0A 2F 43 49 44 n.begincmap./CID
0040 53 79 73 74 65 6D 49 6E 66 6F 3C 3C 0A 2F 52 65 SystemInfo<<./Re
0050 67 69 73 74 72 79 20 28 41 64 6F 62 65 29 0A 2F gistry (Adobe)./
0060 4F 72 64 65 72 69 6E 67 20 28 55 43 53 29 0A 2F Ordering (UCS)./
0070 53 75 70 70 6C 65 6D 65 6E 74 20 30 0A 3E 3E 20 Supplement 0.>>
...
The following code snippet scans all the files contained in a specified directory and its sub-directories:
import os
from ProEngine import *
def scanFiles(scandir):
rootdir = os.path.realpath(scandir)
for dirname, _, filelist in os.walk(rootdir):
print("Scanning directory: %s" % dirname)
for fname in filelist:
fullpath = os.path.join(dirname, fname)
print('\tScanning %s... ' % fname, end ="")
r = proEngineScanFile(fullpath, return_report=True)
print("OK" if r else "FAIL")
if __name__ == "__main__":
proEngineInit()
scanFiles("/path/to/scan")
proEngineFinal()
From C/C++ it’s possible to execute Python scripts, evaluate Python code, retrieve Python variables and scan files. In the following code snippet a global variable is set from Python and then retrieved from C.
#define PRO_ENGINE_INIT
#include "ProEngine.h"
int main()
{
/* initialize the engine */
if (!proEngineInit("/path/to/the/engine", ProEngine_InitPython))
return -1;
/* set a variable in Python and retrieve it from C */
proEngineEvalPythonCode("some_value ='kawaii'");
char *r = proEngineGetPythonGlobalVariable("some_value");
if (r != NULL && strcmp(r, "kawaii") == 0)
puts("SUCCESS");
else
puts("FAIL");
proEngineFreeMemory(r);
/* finalize the engine before exiting */
proEngineFinal();
return 0;
}
Our SDK makes it trivial to pass variables between Python and C/C++.
Speed
Although our SDK is in Python, our engine is written in C++ and is multi-threaded. This design decision guarantees maximum speed, while also giving our customers the capability to write cross-platform code, which is also compatible across different versions of the engine.
Stability
Since our SDK is in Python, our customers don’t need to worry about rebuilding their project when the engine is updated. Not only that, but we take great care not to introduce breaking changes to our SDK: we don’t want our customers to worry about their code not working anymore when updating the engine.
Range
Our SDK is vast. It features support for dozens of file formats, disassembly, decompiling, emulation, file scanning, signature matching, file carving, decompression, decryption and much more.
If you’re in doubt whether our engine fits your use-case, you can contact us for a free consultation!
Quality
We make sure our engine keeps up with the latest threats and challenges presented by file formats which are difficult to analyze. We offer state-of-the-art support for various file types such as Adobe PDF and Microsoft Office.
Flexible Licensing
Our Enterprise Engine is licensed on a per-customer basis. The licensing depends upon the scope of the project. If you are interested in a quotation, please get in touch with us.
Suite Discounts
Purchasing a license of our Enterprise Engine comes with discounted lab licenses for Cerbero Suite!
By using Cerbero Suite, engineers can interactively debug parsing issues, analyze edge cases, use our Python editor and create graphical applications that work together with the engine.
Priority Support
Our Enterprise Engine customers have access to our priority support. We understand the criticality of enterprise applications and hence provide the fastest support for our engine.
