Read a single file's contents or metadata from a GitLab repository and list a repository's file tree via the REST API, including pagination and recursive listing.

Documented steps

1. URL-encode the target file's path as one segment (slashes become %2F), e.g. app/models/key.rb -> app%2Fmodels%2Fkey%2Erb.
2. GET https://gitlab.example.com/api/v4/projects/:id/repository/files/:file_path?ref=main to retrieve file_name, size, base64-encoded content, blob_id, commit_id, last_commit_id, and execute_filemode; use ref=HEAD to target the default branch without knowing its name.
3. Use GET .../repository/files/:file_path/raw?ref=main to fetch raw file bytes directly (no base64 decoding needed); accessible without authentication for public repos.
4. Issue a HEAD request to .../repository/files/:file_path?ref=main to get only metadata via X-Gitlab-* response headers (X-Gitlab-Blob-Id, X-Gitlab-Commit-Id, X-Gitlab-File-Path, X-Gitlab-Size, X-Gitlab-Last-Commit-Id, X-Gitlab-Encoding, X-Gitlab-Execute-Filemode) without downloading the body.
5. GET https://gitlab.example.com/api/v4/projects/:id/repository/tree?path=subdir&ref=main&recursive=true to list directory entries (id, name, type: blob|tree, path, mode); omit recursive for a flat single-level listing.
6. For consecutive pages of a large tree, prefer keyset pagination: pass pagination=keyset and per_page, then use page_token from the previous response instead of offset-based page= paging.
7. Docs: https://docs.gitlab.com/api/repository_files/, https://docs.gitlab.com/api/repositories/