Integrate GitLab Data with LlamaIndex

The GitLab Readers package uses python-gitlab to pull content from GitLab projects, via two separate readers: GitLabRepositoryReader (reads through files in a repo by branch or commit) and GitLabIssuesReader (loads issues, either by project ID or by group ID). An access token or OAuth token is required for private groups and projects.

Use it when you need GitLab repository files or issues loaded into LlamaIndex documents -- for example a specific file at a given branch, or all open issues across a project or group. It requires a GitLab client (gitlab.Gitlab(...)) and, for private groups or projects, an access or OAuth token, so it is not usable against private content without one.

Install with:

pip install llama-index-readers-gitlab

For repository files:

import gitlab
from llama_index.readers.gitlab import GitLabRepositoryReader

gitlab_client = gitlab.Gitlab("https://gitlab.com")

project_repo_reader = GitLabRepositoryReader(
    gitlab_client=gitlab_client,
    project_id=project_id,
    verbose=True,
)

docs = project_repo_reader.load_data(file_path="README.rst", ref="develop")

For issues, by project or by group:

import gitlab
from llama_index.readers.gitlab import GitLabIssuesReader

gitlab_client = gitlab.Gitlab("https://gitlab.com")

### load issues by project id
project_issues_reader = GitLabIssuesReader(
    gitlab_client=gitlab_client,
    project_id=project_id,
    verbose=True,
)

docs = project_issues_reader.load_data(
    state=GitLabIssuesReader.IssueState.OPEN
)

### load issues by group id
group_issues_reader = GitLabIssuesReader(
    gitlab_client=gitlab_client,
    group_id=group_id,
    verbose=True,
)

docs = group_issues_reader.load_data(state=GitLabIssuesReader.IssueState.OPEN)

Developers building LlamaIndex pipelines that need GitLab repository content or issue tracking data, at either the project or group level.