photon-import

name	photon-import
description	Clone or update the vmware/photon repository and import commit history for all branches (3.0, 4.0, 5.0, 6.0, common, master) into a local SQLite database (photon_commits.db). Use this skill when the blogger droid needs fresh commit data, when the database is missing or stale, or when the user requests a data refresh.

Photon OS Commit Importer

Purpose

Populate or update the photon_commits.db SQLite database with commit history from https://github.com/vmware/photon. This database is the data source for the photon-summarize skill.

Prerequisites

• git available on PATH
• Python 3.9+ with sqlite3 (stdlib) and tqdm (auto-installed)
• Network access to github.com (for initial clone / fetch)

Instructions

1. Check current database status

python3 "$FACTORY_PROJECT_DIR/.factory/skills/photon-import/importer.py" \
  --db-path "$FACTORY_PROJECT_DIR/photon_commits.db" \
  --check

Review the JSON output. If exists is false or any branch count is 0, proceed to step 2.

2. Run the import

python3 "$FACTORY_PROJECT_DIR/.factory/skills/photon-import/importer.py" \
  --db-path "$FACTORY_PROJECT_DIR/photon_commits.db" \
  --repo-dir "$FACTORY_PROJECT_DIR/photon" \
  --branches 3.0 4.0 5.0 6.0 common master \
  --since-date "2021-01-01"

The script prints progress to stderr and outputs a JSON summary to stdout with the number of new commits per branch.

3. Verify

Confirm the JSON output shows total_new >= 0 and no errors. If total_new is 0 on a fresh run, investigate network or git issues.

Arguments Reference

Flag	Default	Description
--db-path	photon_commits.db	Path to the SQLite database
--repo-dir	photon	Path to the local git clone
--branches	all 6 branches	Space-separated branch list
--since-date	none (all history)	Only import commits after this date
--check	off	Report DB status without importing

Database Schema

The importer creates the full database schema (shared with summarizer.py via db_schema.py), including both the commits and summaries tables. This means either script can run first on a fresh database.

Output

JSON object to stdout. The --check mode also reports summaries status:

{
  "branches": {
    "3.0": {"commits": 1500, "latest": "2025-01-15T..."},
    "5.0": {"commits": 800, "latest": "2025-01-20T..."}
  },
  "summaries": {
    "3.0": {"count": 12},
    "5.0": {"count": 8}
  },
  "total_new": 52
}