sqlexplore 0.4.0

SQL exploration over data files with your CLI

sqlexplore

SQL exploration over data files with your CLI

Currently in development (Feb 2026).

Install

Requires Python 3.13+.

pip install sqlexplore

Optional (uv tool install):

uv tool install sqlexplore

Usage

Local file:

sqlexplore ./data/example.parquet

Remote file URL:

sqlexplore https://example.com/data_file.csv

Show version:

sqlexplore --version

Remote URL behavior:

• Supports http:// and https:// URLs ending in .csv, .tsv, .txt, .parquet, or .pq.
• Downloads to <app-user-dir>/downloads/<filename> by default.
• Use --download-dir /your/path to override the download location.
• If local download target already exists, prints warning and exits (no overwrite by default). Use --overwrite to replace it.
• Logs download details before normal app flow: remote/local path, progress, elapsed time, and file size.
• In TUI mode, the Activity pane shows app version on load, then startup download log lines (if any).
• Activity pane logs executed SQL for startup query runs, manual SQL, and helper/slash-command generated SQL.

Result formatting:

• JSON syntax highlighting is auto-applied for VARCHAR/text columns when sampled values look like JSON objects/arrays.
• Detection samples only a few visible rows (not full columns) to keep rendering fast.
• Highlighting is disabled when query result row count is over 100,000.
• .txt sources load as one line per row with extra metrics columns: line_number, line_length, line_hash, word_count, mean_word_length, median_word_length, max_word_length, min_word_length.

Autocomplete behavior:

• Completions auto-open from engine context for SQL clauses and helper command arguments (for example SELECT and /top ).
• Up/Down navigates the completion menu whenever it is visible.
• Tab accepts the selected completion when the menu is visible; otherwise it inserts editor indentation.
• Esc closes the completion menu.
• Ctrl+Space opens completion mode explicitly.

Known limitations

• Python 3.12 and below are not supported.
• Remote downloads only support .csv, .tsv, .txt, .parquet, .pq.

Links

• https://github.com/dylanhogg/sqlexplore
• https://pypi.org/project/sqlexplore/