Download and Installation¶
- Install Python 2.7.X or 3.4+.
- Download and install Scapy.
- Follow the platform-specific instructions (depedencies).
- (Optional): Install additional software for special features.
- Run Scapy with root privileges.
Each of these steps can be done in a different way depending on your platform and on the version of Scapy you want to use.
At the moment, there are two different versions of Scapy:
- Scapy v2.x. The current up-to-date version. It consists of several files packaged in the standard distutils way. Scapy v2 <= 2.3.3 needs Python 2.5, Scapy v2 > 2.3.3 needs Python 2.7 or 3.4+.
- Scapy v1.x (deprecated). It does not support Python 3. It consists of only one file and works on Python 2.4, so it might be easier to install. Moreover, your OS may already have specially prepared packages or ports for it. The last version is v1.2.2.
In Scapy v2 use
from scapy.all import * instead of
from scapy import *.
Installing Scapy v2.x¶
The following steps describe how to install (or update) Scapy itself. Dependent on your platform, some additional libraries might have to be installed to make it actually work. So please also have a look at the platform specific chapters on how to install those requirements.
The following steps apply to Unix-like operating systems (Linux, BSD, Mac OS X). For Windows, see the special chapter below.
Make sure you have Python installed before you go on.
To get the latest versions, with bugfixes and new features, but maybe not as stable, see the development version.
$ pip install scapy
$ cd /tmp $ wget --trust-server-names https://github.com/secdev/scapy/archive/master.zip # or wget -O master.zip https://github.com/secdev/scapy/archive/master.zip $ unzip master.zip $ cd master $ sudo python setup.py install
Current development version¶
If you always want the latest version with all new features and bugfixes, use Scapy’s Git repository:
Install the Git version control system. For example, on Debian/Ubuntu use:
$ sudo apt-get install git
or on OpenBSD:
$ doas pkg_add git
Check out a clone of Scapy’s repository:
$ git clone https://github.com/secdev/scapy
Install Scapy in the standard distutils way:
$ cd scapy $ sudo python setup.py install
Then you can always update to the latest version:
$ git pull $ sudo python setup.py install
You can run scapy without installing it using the
run_scapy (unix) or
run_scapy.bat (Windows) script or running it directly from the executable zip file (see the previous section).
Installing Scapy v1.2 (Deprecated)¶
As Scapy v1 consists only of one single Python file, installation is easy: Just download the last version and run it with your Python interpreter:
$ wget https://raw.githubusercontent.com/secdev/scapy/v18.104.22.168/scapy.py $ sudo python scapy.py
Optional software for special features¶
For some special features, you have to install more software.
Most of those software are installable via
Here are the topics involved and some examples that you can use to try if your installation was successful.
plot()needs Matplotlib. It is installable via
pip install matplotlib
>>> p=sniff(count=50) >>> p.plot(lambda x:len(x))
Note: PyX requires version 0.12.1 on Python 2.7. This means that on Python 2.7, it needs to be installed via
pip install pyx==0.12.1. Otherwise
pip install pyx
>>> p=IP()/ICMP() >>> p.pdfdump("test.pdf")
>>> p=readpcap("myfile.pcap") >>> p.conversations(type="jpg", target="> test.jpg")
>>> a,u=traceroute(["www.python.org", "google.com","slashdot.org"]) >>> a.trace3D()
Cryptography is installable via
pip install cryptography
>>> enc=rdpcap("weplab-64bit-AA-managed.pcap") >>> enc.show() >>> enc >>> conf.wepkey="AA\x00\x00\x00" >>> dec=Dot11PacketList(enc).toEthernet() >>> dec.show() >>> dec
PKI operations and TLS decryption. cryptography is also needed.
>>> load_module("nmap") >>> nmap_fp("192.168.0.1") Begin emission: Finished to send 8 packets. Received 19 packets, got 4 answers, remaining 4 packets (0.88749999999999996, ['Draytek Vigor 2000 ISDN router'])
Scapy can run natively on Linux, without libdnet and libpcap.
Just use the standard packages:
$ sudo apt-get install tcpdump graphviz imagemagick python-gnuplot python-cryptography python-pyx
Scapy optionally uses python-cryptography v1.7 or later. It has not been packaged for
apt in less recent OS versions (e.g. Debian Jessie). If you need the cryptography-related methods, you may install the library with:
# pip install cryptography
Here’s how to install Scapy on Fedora 9:
# yum install git python-devel # cd /tmp # git clone https://github.com/secdev/scapy # cd scapy # python setup.py install
Some optional packages:
# yum install graphviz python-cryptography sox PyX gnuplot numpy # cd /tmp # wget http://heanet.dl.sourceforge.net/sourceforge/gnuplot-py/gnuplot-py-1.8.tar.gz # tar xvfz gnuplot-py-1.8.tar.gz # cd gnuplot-py-1.8 # python setup.py install
Mac OS X¶
On Mac OS X, Scapy does not work natively. You need to install Python bindings to use libdnet and libpcap. You can choose to install using either Homebrew or MacPorts. They both work fine, yet Homebrew is used to run unit tests with Travis CI.
Install using Homebrew¶
$ brew update
Install Python bindings:
$ brew install --with-python libdnet $ brew install https://raw.githubusercontent.com/secdev/scapy/master/.travis/pylibpcap.rb $ sudo brew install --with-python libdnet $ sudo brew install https://raw.githubusercontent.com/secdev/scapy/master/.travis/pylibpcap.rb
Install using MacPorts¶
$ sudo port -d selfupdate
Install Python bindings:
$ sudo port install py-libdnet py-pylibpcap
Here’s how to install Scapy on OpenBSD 5.9+
$ doas pkg_add py-libpcap py-libdnet git $ cd /tmp $ git clone http://github.com/secdev/scapy $ cd scapy $ doas python2.7 setup.py install
Optional packages (OpenBSD only)¶
# pkg_add py-cryptography
gnuplot and its Python binding:
# pkg_add gnuplot py-gnuplot
Graphviz (large download, will install several GNOME libraries)
# pkg_add graphviz
ImageMagick (takes long to compile)
# cd /tmp # ftp ftp://ftp.openbsd.org/pub/OpenBSD/4.3/ports.tar.gz # cd /usr # tar xvfz /tmp/ports.tar.gz # cd /usr/ports/graphics/ImageMagick/ # make install
PyX (very large download, will install texlive etc.)
# pkg_add py-pyx
# wget http://git.netfilter.org/ebtables/plain/ethertypes -O /etc/ethertypes
python-bz2 (for UTscapy)
# pkg_add python-bz2
Scapy is primarily being developed for Unix-like systems and works best on those platforms. But the latest version of Scapy supports Windows out-of-the-box. So you can use nearly all of Scapy’s features on your Windows machine as well.
If you update from Scapy-win v22.214.171.124 to Scapy v2 remember to use
from scapy.all import * instead of
from scapy import *.
You need the following software packages in order to install Scapy on Windows:
- Python: Python 2.7.X or 3.4+. After installation, add the Python installation directory and its Scripts subdirectory to your PATH. Depending on your Python version, the defaults would be
- Npcap: the latest version. Default values are recommanded. Scapy will also work with Winpcap.
- Scapy: latest development version from the Git repository. Unzip the archive, open a command prompt in that directory and run “python setup.py install”.
Just download the files and run the setup program. Choosing the default installation options should be safe.
For your convenience direct links are given to the version that is supported (Python 2.7 and 3.4+). If these links do not work or if you are using a different Python version (which will surely not work), just visit the homepage of the respective package and look for a Windows binary. As a last resort, search the web for the filename.
After all packages are installed, open a command prompt (cmd.exe) and run Scapy by typing
scapy. If you have set the PATH correctly, this will find a little batch file in your
C:\Python27\Scripts directory and instruct the Python interpreter to load Scapy.
If really nothing seems to work, consider skipping the Windows version and using Scapy from a Linux Live CD – either in a virtual machine on your Windows host or by booting from CDROM: An older version of Scapy is already included in grml and BackTrack for example. While using the Live CD you can easily upgrade to the latest Scapy version by using the above installation methods.
- You may not be able to capture WLAN traffic on Windows. Reasons are explained on the Wireshark wiki and in the WinPcap FAQ. Try switching off promiscuous mode with
- Packets sometimes cannot be sent to localhost (or local IP addresses on your own host).
As Winpcap is becoming old, it’s recommended to use Npcap instead. Npcap is part of the Nmap project.
- If you get the message ‘Winpcap is installed over Npcap.’ it means that you have installed both winpcap and npcap versions, which isn’t recommended.
- You may uninstall winpcap from your Program Files, then you will need to remove:
- And if you are on an x64 machine:
To use npcap instead. Those files are not removed by the Winpcap un-installer.
2. If you get the message ‘The installed Windump version does not work with Npcap’ it surely means that you have installed an old version of Windump. Download the correct one on https://github.com/hsluoyz/WinDump/releases
In some cases, it could also mean that you had installed Npcap and Winpcap, and that Windump is using Winpcap. Fully delete Winpcap using the above method to solve the problem.
Build the documentation offline¶
The instructions to build the HTML version are:
(activate a virtualenv) pip install sphinx cd doc/scapy make html
You can now open the resulting HTML file
_build/html/index.html in your favorite web browser.
To use the ReadTheDocs’ template, you will have to install the corresponding theme with:
pip install sphinx_rtd_theme
pyreverse you can build a UML representation of the Scapy source code’s object hierarchy. Here is an
example of how to build the inheritance graph for the Fields objects :
(activate a virtualenv) pip install pylint cd scapy/ pyreverse -o png -p fields scapy/fields.py
This will generate a
classes_fields.png picture containing the inheritance hierarchy. Note that you can provide as many
modules or packages as you want, but the result will quickly get unreadable.
To see the dependencies between the DHCP layer and the ansmachine module, you can run:
pyreverse -o png -p dhcp_ans scapy/ansmachine.py scapy/layers/dhcp.py scapy/packet.py
In this case, Pyreverse will also generate a
packages_dhcp_ans.png showing the link between the different python modules provided.