Changing the way the world does text.
Our newest product is ICgrep, our parallel bitstream implementation of the widely used regular expression processing utility called grep, which is built-in to the Linux and Mac-OS operating systems, and a commonly downloaded utility for the Windows platform (which itself uses the built-in FINDSTR utility). Regular expressions are widely used to identify patterns in data files and data streams, with applications security, search engines, biomedical research, in database systems, and in a wide range of big data applications
Our ICgrep technology was presented at the 2014 Parallel Architectures and
Compilation Technology (PACT) conference and can be found in the paper on Bitwise Data Parallelism in Regular Expression Matching. More recent developments were presented in Performance Matters - A New Algorithmic Approach for Fast Unicode Regular Expression Processing at Internationalization and Unicode Conference 39.
ICgrep is available as open source software through SFU's Centre for Open Software Technology and Applications Research (COSTAR). Browse the ICgrep directory by visiting http://parabix.costar.sfu.ca/browser/tags/icgrep1.0. Executables can also be downloaded from our downloads page, along with the document icGREP_Demonstration.pdf which shows the use and performance of ICgrep.The technical webpage for ICgrep is available at icgrep.net.
The distribution below includes a prebuilt LLVM install, for those who wish to just recompile ICgrep. To try out the Windows implementation, please contact firstname.lastname@example.org. The Linux implementation can be checked-out and downloaded by using Apache™ Subversion® with the following command: