Your developers probably already use Sourcegraph to search open-source code. Now you can run Sourcegraph securely on your network for your company's code, so your developers can be more productive at work.
The core of Sourcegraph’s code search is a highly optimized repository text search algorithm tuned for high OS file system cache hit rates and effective use of memory. Each Sourcegraph search node caches a subset of your repositories and performs searches using this algorithm, and a supervisor node routes queries. The overall system is horizontally and almost-linearly scalable. In a recent benchmark, doubling the number of nodes from 4 to 8 resulted in the same query completing in 60% of the time.
Sourcegraph code search connects to your team's code host, so all of your repositories are searchable and up to date.
Optimized on-the-fly searching means you never see stale results. You can search for code that was pushed just seconds before.
Use complex regular expressions, sub-word matches, include and exclude paths, language filters, and more.
Search over any historical commit, not just the tip of a single branch. No pre-indexing required.
The web app is easy to use for everyone and highly efficient for power users. Find what you're looking for with the query interface (or just type and go), share result URLs, and navigate around code.
Sourcegraph code search integrates with your team's existing tools: code hosts, editors, and web browsers. All this means you can perform code searches right when you need.
The API exposes what you need to build more advanced tooling, using complex queries, multi-revision searches, authorship data, revision history, and more.
Sourcegraph code search is securely deployed on your company's servers. We'll help you set it up if you prefer.
Install Sourcegraph code search in 10-15 minutes from a Docker image or prebuilt VM. It doesn't need data persistence/storage, so it's easy to manage.