Skip to content

CHANGELOG.md

Latest commit

 

History

History
207 lines (125 loc) · 8.63 KB

CHANGELOG.md

File metadata and controls

207 lines (125 loc) · 8.63 KB

Changelog

All notable changes to this project will be documented in this file.

The format follows Keep a Changelog and this project adheres to Semantic Versioning.

[0.9.1] - 2026-03-10

Fixed

  • restore (PostgreSQL plain, MySQL): SQL files produced by pg_dump -C, pgAdmin, or mysqldump --databases embed \connect, CREATE DATABASE, and USE directives that redirected the client to the source database, ignoring --db. The plain SQL content is now filtered to strip these directives before being piped to the database client, so the restore always targets the database specified via --db.

[0.9.0] - 2026-03-04

Added

  • MongoDB engine support (--engine mongodb). Requires mongosh, mongodump, and mongorestore binaries.
  • Backup format: archive (mongodump --archive --gzip) producing .mongodump files. Plain format is not supported by MongoDB and will exit with a clear error.
  • hdx inspect <file>.mongodump: inspect MongoDB archive dumps offline via mongorestore --dryRun.
  • MongoDB added to hdx cloud upload and hdx restore s3:// workflows (engine-agnostic path, no changes required).
  • mongodb added to npm keywords.

[0.8.2] - 2026-03-04

Added

  • hdx cloud list [path] now defaults to directory mode: shows only immediate children (files and subdirectories) at the given level, similar to ls. Pass --recursive to list all objects regardless of depth.
  • hdx cloud upload <file> [--prefix PREFIX] command: upload an existing local file to the configured S3 bucket.
  • hdx backup --no-keep flag: delete the local backup file after a successful upload (requires --upload).
  • hdx cloud delete now verifies the key exists before attempting deletion. Exits with an error if the key is not found or is a directory prefix.

Fixed

  • hdx cloud list <path> now accepts path as a positional argument. Previously, passing a path without --prefix was silently ignored.
  • hdx cloud download now saves files to ~/.herdux/backups/ by default, creating the directory if needed. Previously, files were saved to the current working directory.

[0.8.1] - 2026-03-03

Fixed

  • hdx cloud list no longer crashes with a stack overflow when the bucket contains a very large number of objects. Display is now capped at 200 items with a truncation notice.

[0.8.0] - 2026-03-03

Added

  • hdx cloud config command: configure cloud storage settings (bucket, region, access-key, secret-key, endpoint). Credentials stored in ~/.herdux/config.json; env vars AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY take priority.
  • hdx cloud list [--prefix PREFIX] command: list backup files in the configured S3 bucket. Displays key, size, and last modified date in a table.
  • hdx cloud download <key> [-o DIR] command: download a backup file from the configured S3 bucket to a local directory.
  • hdx cloud delete <key> [-y] command: delete a backup file from the configured S3 bucket. Requires confirmation unless --yes is passed.
  • hdx backup --upload [prefix] flag: after a successful backup, upload the file to the configured S3 bucket. Accepts an optional key prefix (e.g. backups/).
  • hdx restore s3://bucket/key --db mydb: restore directly from an S3 URL. Downloads to a temp file, restores, then cleans up automatically.
  • Support for any S3-compatible provider (Cloudflare R2, MinIO, DigitalOcean Spaces) via hdx cloud config endpoint URL.

[0.7.0] - 2026-03-02

Added

  • hdx docker list command: lists running postgres, mysql, and mariadb containers. Displays name, engine type, mapped host port, and status in a table. Accepts --all to include stopped containers.
  • hdx docker start <name> command: starts a stopped database container.
  • hdx docker stop <name> command: stops a running container. Accepts --remove to also remove it after stopping.

[0.6.0] - 2026-03-02

Added

  • hdx inspect <file> command: inspects the contents of a backup file without connecting to a database. Supports .dump (PostgreSQL custom, via pg_restore --list), .sql (any engine, extracts CREATE statements), and .db / .sqlite (SQLite, via sqlite3 .schema). Completely offline.

Fixed

  • MySQL engine now rejects --format custom with a clear error message. Previously, passing --format custom for a MySQL backup was silently ignored and produced a plain SQL file regardless. MySQL only supports plain SQL format via mysqldump.

[0.5.0] - 2026-03-02

Added

  • SQLite engine support (--engine sqlite). File-based, no server required. Databases stored at ~/.herdux/sqlite/ by default; override with --host <dir>.
  • test:e2e:sqlite npm script and e2e-sqlite CI job. No Docker required; SQLite is installed via apt-get in CI.
  • SQLite E2E workflow test covering the full lifecycle: doctor, version, config, create, list, backup (custom and plain), drop, restore, cleanup.
  • sqlite added to npm keywords.

Fixed

  • Connection resolver now skips port-based auto-discovery for portless engines. Previously, running any command with --engine sqlite without a port would incorrectly trigger server scanning and exit with "No running SQLite servers found".
  • sqlite added to the valid engines list in src/index.ts. Previously, passing --engine sqlite was rejected with "Unknown engine".

[0.4.3] - 2026-03-01

Added

  • Help examples via .addHelpText for all 9 commands, visible with hdx <command> --help.
  • Input validation for database names: rejects names containing spaces and special shell characters.
  • Improved command descriptions for clarity across all commands.

[0.4.2] - 2026-02-28

Added

  • CLI integration test harness (CliRunner) with isolated temp directories. Tests run the compiled CLI as a subprocess.
  • AI agent workflow guides in .agents/workflows/: pre-commit.md, testing.md, new-engine.md, new-command.md, refactoring.md.
  • Coverage thresholds enforced in CI: 95% statements, functions, lines, and branches for src/commands/.

Changed

  • All unit tests parametrized with describe.each(engines) to run against both PostgreSQL and MySQL simultaneously.
  • Extracted promptServerSelection and resolveEngineType from resolve-connection.ts into private helpers, reducing function size and improving testability.

[0.4.1] - 2026-02-28

Fixed

  • --engine flag now correctly filters the server profile selection prompt. Previously, passing --engine mysql would still show PostgreSQL profiles in the interactive list.

[0.4.0] - 2026-02-27

Added

  • MySQL engine support (--engine mysql). Requires mysql and mysqldump binaries.
  • Multi-engine architecture: all commands are now engine-agnostic behind the IDatabaseEngine interface.
  • --engine global flag accepted by all commands.
  • Engine type saved per server profile and in global config (herdux config set engine mysql).
  • GitHub Actions CI, Dependabot config, and Husky pre-commit hooks.

[0.3.1] - 2026-02-26

Fixed

  • PostgreSQL restore now ignores ownership and ACL assignments (--no-owner, --no-acl), preventing errors when restoring production dumps to a local environment where production roles do not exist.
  • Non-fatal pg_restore warnings (e.g. missing roles) are now reported informatively instead of causing the command to fail.

[0.3.0] - 2026-02-26

Changed

  • Full architecture refactor: decoupled all CLI commands from PostgreSQL internals and introduced the abstract IDatabaseEngine contract (src/core/interfaces/).
  • Engine implementations moved to src/infra/engines/, commands moved to src/commands/.
  • Added comprehensive E2E test suite using Docker Compose.
  • Added comprehensive unit test suite with Jest and ts-jest.

[0.2.2] - 2026-02-25

Changed

  • Minor documentation and badge updates.

[0.2.0] - 2026-02-25

Added

  • hdx short alias for the herdux binary.
  • CLI version now reads dynamically from package.json at runtime.

[0.1.1] - 2026-02-24

Changed

  • Renamed application to herdux (from previous internal name).
  • Published as herdux-cli on npm.

[0.1.0] - 2026-02-23

Added

  • Initial release. PostgreSQL-only CLI.
  • Commands: version, list, create, drop, backup, restore, clean, doctor, config.
  • Named server profiles with herdux config add.
  • Global defaults with herdux config set.
  • Auto-discovery of running PostgreSQL servers on common ports.
  • Backup in custom (pg_dump -Fc) and plain (SQL) formats.
  • Restore with automatic format detection and database auto-creation.
  • Interactive multi-select bulk cleanup (herdux clean) with optional safety backup.