AI-powered code search for the terminal and beyond.
Project description
Codebased
Codebased is the most powerful code search tool that runs on your computer.
Here's why it's great:
- Search, simplified: Combines semantic and full-text search to find what you're looking for, not just what you typed.
- Search code, not just text: Searches for complete code structures, not just lines of text, in 11 languages.
- Ignores uninteresting files: Respects your
.gitignore
file(s) and ignores hidden directories. - Instant editing: Selecting a search result opens your favorite editor at the correct file and line number.
- Fast: Indexes your code in seconds, searches in milliseconds, and updates in real-time as you edit your code.
- Open-source, runs locally: No code is stored on remote servers.
Getting Started
The fastest way to install Codebased is with pipx:
pipx install codebased
Verify the installation by running:
codebased --version
If this fails, please double-check your pipx installation.
Next, run the following command in a Git repository to start searching:
codebased search
The first time you run Codebased, it will create a configuration file at ~/.codebased/config.toml
.
It will prompt you for an OpenAI key, you can access a testing key on the Discord.
Once this is finished, codebased
will create an index of your codebase, stored in .codebased
at the root of your
repository.
This takes seconds for most projects, but can take a few minutes for large projects.
Most of the time is spent creating embeddings using the OpenAI API.
Once the index is created, a terminal UI will open with a search bar. At this point, you can start typing a search query and results will appear as you type.
- You can use the arrow keys and the mouse to navigate the results.
- A preview of the selected result is displayed.
- Pressing enter on the highlighted result opens the file in your editor at the correct line number.
- Pressing escape returns to the search bar.
- As you edit your code, the index will be updated in real-time, and future searches will reflect your changes.
Codebased will run stat
on all non-ignored files in your repository, which can take a few seconds, but after that
will listen for filesystem events, so it's recommended to use the TUI.
Development
If you'd like to contribute, bug fixes are welcome, as well as anything in the list of issues.
Especially welcome is support for your favorite language, as long as:
- There's a tree-sitter grammar for it.
- There are Python bindings for it maintained by the excellent amaanq.
Also, if there's anything ripgrep does that Codebased doesn't, feel free to file an issue / PR.
Clone the repository:
git clone https://github.com/codebased-sh/codebased.git
Install the project's dependencies (requires poetry, using a virtual environment is recommended):
poetry install
Run the tests (some tests require an OPENAI_API_KEY
environment variable, usage is de minimis):
poetry run pytest
Appendix
Languages
- C
- C#
- C++
- Go
- Java
- JavaScript
- PHP
- Python
- Ruby
- Rust
- TypeScript
- HTML / CSS
- SQL
- Shell
- Swift
- Lua
- Kotlin
- Dart
- R
- Assembly language
- OCaml
- Zig
- Haskell
- Elixir
- Erlang
- TOML?
- YAML?
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file codebased-0.6.2.tar.gz
.
File metadata
- Download URL: codebased-0.6.2.tar.gz
- Upload date:
- Size: 31.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.1 CPython/3.9.20
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | cf8368885d8173b5c091f97c499d63a1a2fcafe8990247c79b35e622abb11655 |
|
MD5 | 7ce346e35f21ffca40325d5f740cce95 |
|
BLAKE2b-256 | 9ab066f4f7e943ddf1790d1102098a8f63f9ff7003879fb69695982e9df23b70 |
File details
Details for the file codebased-0.6.2-py3-none-any.whl
.
File metadata
- Download URL: codebased-0.6.2-py3-none-any.whl
- Upload date:
- Size: 36.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.1 CPython/3.9.20
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 740a878a28f10647f2dd0db38860eeb4c42f50f184e1e273be3e82b526ef9a07 |
|
MD5 | fc5abb8d75f3ef105cb06c7652458cfd |
|
BLAKE2b-256 | d310f548cde04a7ebc686f2e77815bd13575063c28330e5b432b0b41c99a6c32 |