Obtaining COD

COD is an open-access database, and you can freely obtain all data contained in it. You can get the database using one of the following methods outlined below:


Using Subversion

COD, PCOD and TCOD are available as Subversion repositories, open for anonymous download – you are welcome to update your copies on a regular basis (say, each midnight). This is probably the best and the most complete method to get the databases – after the initial checkout (which, admittedly, can take long), you will have a possibility to update your working copies of the COD, PCOD and TCOD databases very quickly, since only new additions and corrections will be transferred on each subsequent update. In addition, Subversion will give you access to all change logs of the (P,T)COD database, and the possibility to retrieve any previous revision of any file.

Access to the repositories are available at:




Under Unix or Linux, one can use the following command to retrieve the initial copy (it will create a directory cod/ for you):

svn co svn://www.crystallography.net/cod

and the following commands to update it:

cd cod
svn update

Please note that the checkout of the whole COD will take somewhat more space on your hard disk than plain rsync-downloaded data, since Subversion keeps original archived copies of the files, in case you modify them and want to compare with or revert to the original without accessing the server (the full checkout of the COD repository as of 2011-02-08 will take 11G, and just the cif/ subdirectory will occupy 5.2G).

Due to Subversion's sensitivity to locales that can lead to introduction of non-ASCII symbols in CIF files therefore corrupting them, it is recommended to use English locale for all checkouts/updates:

LC_TIME=en_US.UTF-8 svn co svn://www.crystallography.net/cod

Using 'rsync'

We have also implemented the rsync access to our CIF files. This method will also enable you to update your COD copy efficiently, only fetching the changes, but it will only give you access to the latest revision, and will not give tools to view logs. The rsync address of the repository is:


You can download the COD data with the following Linux command:

mkdir -p cif; rsync -av --delete rsync://www.crystallography.net/cif/ cif/
mkdir -p hkl; rsync -av --delete rsync://www.crystallography.net/hkl/ hkl/

Using plain http

If you cannot or do not want to use the above mentioned methods, you can download the whole COD dataset in a single archive. Please note however that these archives are inevitably incomplete – we only archive them once in a while, probably once a three or four months, and COD sometimes gets updated daily. Also, each new download will fetch all files, old and new, consuming unnecessarily yours and ours bandwidth. Use these files only if you absolutely cannot retrieve COD and PCOD using Subversion or rsync:




When you have downloaded the archive and unpacked it once, you can later update it using rsync (but not Subversion).

The current version and the older snapshots are available at


Finally, using the Web browser or 'wget', the whole CIF tree can be fetched at:


Please do not overload our servers and make a dozen of second pauses before fetching subsequent files. In this way you can get an up-to-date version of the COD contents, but efficient updates are not possible with wget. But of course you can use rsync to update you copy :)


The most up-to-date version of PCOD can be found on http://cod.ensicaen.fr/pcod/, but we also maintain a mirror.

As mentioned before, the PCOD and COD are available as Subversion repositories, open for anonymous download – you are welcome to update your copies on a regular basis (say, each midnight):



The whole CIF tree is available at


There is currently no pre-packed version of PCOD (.zip or .tgz files), but if you would need one, please let me know, I will make one and publish it on the Web.