libleeloo is a C++ library (with Python bindings) that help manages intervals of integers with fast and memory-wise randomisation of their elements. It can be used directly in any C++ program or through the Python bindings.
One common usage of this library is to aggregate and randomize IPv4 ranges, which can be found in large-scale IP scanners. See the Examples section for more information.
You need at least GCC 4.7.3 (4.7.2 crashes) or Clang 3.4, a Linux-based system, and these libraries:
- Boost Random >= 1.49
- Intel Threading Building Blocks (https://www.threadingbuildingblocks.org/) >= 4.0
- Boost Serialization >= 1.49 for random state persistance (thus not necessary)
These tools are also needed:
- cmake
Support for other OS-es and compilers may be provied in the future.
Under Debian-based systems, this can be installed thanks to:
# aptitude install cmake libtbb-dev libboost-random-dev libboost-serialization-dev
Under Fedora, this can be done thanks to:
# yum install cmake tbb-devel boost-devel
This project uses CMake. To compile it, follow these steps:
$ cd /path/to/src
$ mkdir build && cd build
$ cmake -DLEELOO_BOOST_SERIALIZE=1 -DCMAKE_BUILD_TYPE=Release ..
$ make
Use -DLEELOO_BOOST_SERIALIZE=1 if you want to support random state persistance through the list_intervals_random* classes (needed for nodescan).
The supported build types are :
- debug: compiles with no optimisations and debug symbols (-g)
- relwithdebinfo : compiles with full optimisations and debug symbols (-O3 -march=native -g)
- release : compiles with full optimisations and no debug symbol (-O3 -march=native)
This will compile the library. Then, as root, you can install it :
# make install
In order to install the python bindings, you need these dependencies:
- Python development environment
- Boost Python >= 1.49 for Python bindings (thus not necessary)
Under Debian-based system, this can be installed like this:
# aptitude install libboost-python-dev python-dev
Under Fedora, you can do:
# yum install boost-python-devel python-devel
Then, do the following :
$ cd /path/to/src
$ cd bindings/python
$ mkdir build && cd build
$ cmake -DCMAKE_BUILD_TYPE=Release -DPYTHON_VERSION=2.7 ..
$ make
Then, as root :
# make install
The PYTHON_VERSION variable tells CMake for which Python version the bindings must be compiled. Fedora systems need to manually specify the boost python library name with the -DPYTHON_BOOST=python argument to the cmake command line.
After building the library, just run as root:
# make install
This will install the library to /usr/local, and the Python module to the current Python installation.
Debian packages can be downloaded on Quarkslab's public debian repository. Packages for the stable (wheezy) and unstable (sid) versions are available.
Instructions are available here to add this repository to your system.
Then, you simply have to do:
# apt-get update
# apt-get install libleeloo
Development packages are also available:
# apt-get install libleeloo-dev
Python 2.7 and 3 bindings are also available as debian packages:
# apt-get install python-pyleeloo python3-pyleeloo
The "leeloo-aggregate" and "leeloo-ip-ranges-rand" tools are available in a separate "leeloo-tools" package:
# apt-get install leeloo-tools
Some examples can be found in the tools/ and tests/ directory.
One tool is provided for randomizing IP ranges :
$ ./leeloo-ip-ranges-rand
Usage: ./leeloo-ip-ranges-rand range1 [-][range2] [-][range3] ...
This tool will aggregate all the given ranges and will write a random order of the corresponding IP.
where range can be described as:
CIDR notation: 192.168.1.0/24
Ranges: 10.4-5.8.9-250
Single IP: 192.168.1.10
A '-' symbol before any range will remove it from the final set.
You can also read a blog post about some math used in this library in Quarkslab's blog here : http://blog.quarkslab.com/unique-random-number-set-computation.html
Any remarks, patches, gift-cards and other related stuff can can go to:
See AUTHORS file.
See LICENSE file.