v0.10.0

v0.9.1...v0.10.0

Overview

Breaking changes

• The debug console command read has been renamed to eval

• YouTube text search is now supported
  • Configure how many results to choose from are shown with the max_search_results config key
• Config key media_filter.allowed_urls is now media_filter.allowed_queries
• Tracks can now be looped via the -loop command
• You can now have the bot prompt for confirmation when using -remove by setting the
  confirm_on_remove config key to true

Added

• Added bot command -loop (method cogs.voice.VoiceCog.toggle_loop()) (#34)
• Added debug bot command -buttons
• Added config key confirm_on_remove (bool)
  • If true, using -remove will prompt the user to confirm their choice
• Added config key max_search_results (#39)
• Added title named group to markdown header regexes in const
  • const.MD_HEADER_REGEX
  • const.MD_H1_REGEX
  • const.MD_H2_REGEX
  • const.MD_H3_REGEX
• Added tool module tools.testdataytdl
• Added enum member const.EmojiStr.LOOP (#34)
• Added enum member const.EmojiStr.LOOP_ONE (#34)
• Added attribute cogs.voice.MediaItem.uploader
• Added attribute cogs.voice.VoiceCog.loop (#34)
• Added class cogs.util.DynamicButtonsView
• Added constant const.EMOJI_DIGITS
• Added constant const.HTTP_REGEX
• Added constant const.YTDL_SEARCH_PREFIX_REGEX (#39)
• Added classmethod const.EmojiStr.from_int()
• Added parameter confirm_callback to method cogs.voice.MediaItem.from_url()
  • Takes a function which receives the tuple of constructed MediaItem objects and must return
    True for the method to return those objects, returning an empty tuple if False
• Added method cogs.voice.VoiceCog.loop_state_embed() (#34)
• Added method cogs.voice.VoiceCog._prompt_media_item_choice() (#39)
• Added function util.group_by()
• Added function util.nop()
  • "No-op"; takes any arguments and keyword arguments and does nothing
• Added function util.nop_ret()
  • Returns a function with takes any arguments and keyword arguments and returns a specified value
  • util.nop_true and util.nop_false also added for convenience which return True and False
    respectively
• Added mock function tests.conftest.mock_ytdl()
• Added proper handling for discord.ext.commands.errors.BadLiteralArgument in
  bot.on_command_error()

Changed

• Renamed config key media_filter.allowed_queries
• Renamed console command debug read to debug eval
• Renamed attribute config.MediaFilterConfig.allowed_urls to .allowed_queries
• Renamed method console.LydianConsole.debug_read() to .debug_eval()
• Renamed method config.Config.filter_media_url() to .filter_query_url()
• Renamed function util.assure() to asserts()
• util.asserts() now raises the builtin AssertionError instead of the now-removed custom
  exception errors.AssuranceError
• Method cogs.util.DropdownView.__init__() parameter options now accepts Iterable instead of
  just list
• util.Cache objects can now be effectively disabled with the enabled attribute, which prevents
  setting keys and retrieving keys from its data dictionary
• Multiple methods in MediaItem now use .get(...) or ... instead of setting a default for
  .get() in yt-dlp extracted info dictionary usage
• Method cogs.voice.VoiceCog.play() now takes a variable argument query_parts instead of url
  (#39)
  • This allows queries with spaces to be used as one whole string
• Attribute cogs.voice.VoiceCog._play_calls now requires a third item for its tuples: either
  'url' or 'search' (#39)
• Method cogs.voice.VoiceCog._try_to_queue() now takes an optional choose parameter (#39)
  • Prompts the user to choose one track to queue out of the extracted items
• Various messages that mention a URL in queueing now mention "query" since searching is supported
  (#39)

Removed

• Removed class tests.ReadOnlyDict
• Removed exception errors.AssuranceError
• Removed exception tests.ReadOnlyModifiedError
• Removed function tests._raise_for_readonly()

Fixed

• Fixed markdown header regexes matching more header symbols than intended, they now correctly match
  only their respective header level
• Fixed check_for_stable_only config key not getting checked when checking updates at bot startup
  (#42)

v0.9.1

v0.8.0...v0.9.1

This release is intended to replace the now-yanked v0.9.0, which due to improper merging did not contain any of the intended changes.

Overview

• Vote-skipping is now supported, optionally requiring a minimum number of users (a percentage of
  the channel or an exact count) to use -skip before the current track is skipped. (#16)
  • To use it, add the [vote_skipping] table to your lydian-config.toml like below and adjust
    the values as desired.

[vote_skipping]
enabled = true
threshold_type = "percentage"
percentage = 50
literal = 3

---

Added

• Added debug console command debug store
• Added debug bot command -dropdown (method cogs.debug.DebugCog.dropdown())
• Added module help
• Added constant const.DOCSTRING_PARAM_REGEX
• Added constants const.GH_CHANGELOG_WEB and const.GH_CHANGELOG_RAW
• Added constants to const:
  • MD_HEADER_REGEX
  • MD_H1_REGEX
  • MD_H2_REGEX
  • MD_H3_REGEX
• Added enum members const.EmojiStr.BACK and const.EmojiStr.GEAR
• Added classes cogs.util.ArrowButtonsView and cogs.util.DropdownView
• Added class cogs.voice.VoteSkip
• Added attribute cogs.voice.MediaItem.user_id to replace .user
• Added method console.LydianConsole._debug_evaluate_in_context()
• Added functions to cogs.util:
  • cog_emoji()
  • command_signature()
  • paginated_message()
• Added functions to util:
  • cog_commands()
  • getclass()
  • get_text_sections()
  • mention()
• Added pytest fixtures mock_bot and mock_discord_user

Changed

• util.dirsize() now only counts file sizes and ignores directories
• cogs.voice.MediaItem objects now store who queued the item as their int user ID rather than
  the actual discord.Member object, which allows them to be deep-copied
• Moved update.GH_API_ROOT, update.GH_REPO_API_ROOT, and update.USER_AGENT to const
  • Value of USER_AGENT updated to use the package's name instead of lydian-update-checker
• Swapped arguments of util.first_where() to make the predicate first, matching built-ins like
  map and filter

Removed

• Removed attribute cogs.voice.MediaItem.user, (replaced by .user_id)

Fixed

• Fixed cogs.voice.MediaItem objects always refreshing when advancing queue if their
  thumbnail_url attribute is truthy (incorrectly wrote
  (not item.duration) or (item.thumbnail_url) instead of
  (not item.duration) or (not item.thumbnail_url))
• Fixed ZoneInfo not working correctly on systems without an IANA database by including tzdata
  as a dependency

v0.8.0

v0.7.0...v0.8.0

Overview

Breaking changes:

• Some TOML configuration keys are now properly nullable, using the 'n/a' value to indicate
  such.
  • Config keys max_duration, media_dir_warn_threshold, inactivity_timeout, and
    lonely_timeout now use 'n/a' instead of -1 or 0 to indicate they should be disabled

---

• Lydian can now optionally check for new releases at startup, toggled with the check_for_updates
  configuration key (or LYDIAN_CHECK_UPDATES environment variable).
  • You can check for updates manually by running lydian-update.
  • By default, pre-releases are skipped during this check. Set the check_for_stable_only config
    key to false to include pre-releases.
• The media queue now has a shuffle mode, use -shuffle <on|off> to enable or disable it.
• Lydian now includes exception arguments in "unexpected error" messages.
  • Full tracebacks are still only available in Lydian's logs.
• A temporary (deletes after 10 seconds) message is now sent when skipping a track

---

Added

• Added config key bot_console (boolean)
  • env: LYDIAN_BOT_CONSOLE
• Added config key check_for_updates (boolean)
  • env: LYDIAN_CHECK_UPDATES
• Added config key check_for_stable_only (boolean)
  • env: LYDIAN_CHECK_STABLE_ONLY
• Added console commands version and updates
• Added module update
• Added debug bot commands -argstr and -argint
• Added bot commands:
  • -issues
  • -repo
  • -shuffle
• Added attribute cogs.voice.VoiceCog.shuffle
• Added attributes converter_env and converter_toml to config.ConfigFieldMeta
• Added command methods cogs.debug.DebugCog.argstr() and cogs.debug.DebugCog.argint()
• Added command methods cogs.general.GeneralCog.repo() and cogs.general.GeneralCog.issues()
• Added command method cogs.voice.VoiceCog.toggle_shuffle()
• Added method console.LydianConsole.version()
• Added classmethod tools.todos.Todo.parse_todos()
  • Parses Todo objects directly from a string without needing file paths
• Added method tools.todos.Todo.with_file()
• Added function util.exc_str()

Changed

• Package version is now set dynamically using __version__ from __init__.py
• Bot command -hello now accepts aliases
• The discord.py exceptions BadArgument and MissingRequiredArgument are now properly handled
• Logging methods in cogs.voice.YTDLLogHandler now log with depth=2 for more accuracy
• Renamed in const:
  • COLOR_INFO → EMBED_COLOR_INFO
  • COLOR_OK → EMBED_COLOR_OK
  • COLOR_WARN → EMBED_COLOR_WARN
  • COLOR_ERROR → EMBED_COLOR_ERROR
• A status message is now sent when retrieving info for a track before downloading begins (#30)
• Tool todos now accepts multiple directories to search
• In tools.todos:
  • Todo.content is now an attribute and not a method
  • Todo.header is now a cached property and no longer accepted as a constructor argument
  • Todo.file is now an optional attribute
  • find_todos() now takes a variable positional argument paths instead of the single
    positional argument source_dir, still str | Path
• Moved function config.env_to_bool() to util.FromStr.to_bool()
• Renamed classmethod util.FromStr.filesize() to util.FromStr.to_filesize()
• Unexpected errors caught in bot.on_command_error now have their exception args sent along with
  the bot's message instead of only saying to check the logs

Fixed

• Fixed util.tabulate() raising an error for empty data (#32)
• Fixed todos dev tool skipping # TODO lines with no author (#33)

v0.7.0

v0.6.1...v0.7.0

Overview

Breaking changes:

• TOML configuration keys now use underscores (_) instead of hyphens (-)
• Config key log_level under logging is now level

---

• [feature] A duration limit for media items can now be set using the max-duration config key, which
  specifies the limit in seconds using a positive integer (setting to 0 will disable the limit)
  • max-duration-allow-unknown specifies whether tracks with an unknown duration can be played
    or not when a duration limit is in place, defaulting to false
    • If no limit is set, this key is unused and tracks of unknown duration are always allowed
    • To save time on queueing, Lydian does not fetch a track's duration until it needs to, meaning
      items queued from a playlist URL won't have their duration checked against the limit until
      they are next in line
• [feature] lydian can now be run with the -V or --version option to display the installed Lydian
  version and exit without starting the bot
• [feature] The "now playing" view now shows the approximate current timestamp of the playing media
  (#13)
  • Slight network delays will likely cause this number to be slightly inaccurate, but it should
    remain close enough for reference
• [fix] Fixed filesize strings in the config with single digits raising an error due to a typo in the
  regex used to match them

---

Added

• Added the -V/--version option to the lydian package command
  • When used, Lydian's version is printed out and the script immediately exits
• Added config key max-duration (integer)
• Added config key max-duration-allow-unknown (boolean)
• Added class config.Config.ConfigField
• Added class config.Config.ConfigFieldMeta
• Added method config.Config.set()
• Added property config.Config.fields
• Added function util.compose()
• Added support for validator functions in config.Config fields, which can be given with the
  'validators' key of a given field's metadata argument as a single function or iterable of
  functions

Changed

• Having debug mode enabled no longer forces the log level to "DEBUG" level
• cogs.voice.MediaItem.duration_str now returns ?:?? for an unknown duration instead of None
• cogs.voice.MediaItem.embed() now accepts an optional timestamp keyword argument to show the
  current time alongside its duration
• TOML configuration keys are now in snake_case instead of kebab-case; though the latter is more
  common in TOML, it's ultimately unnecessary to keep converting keys between the two cases where
  needed
• const.LogLevel is now an IntEnum subclass instead of StrEnum
• config.Config updating and loading from TOML is now handled in the class itself in a simpler
  fashion than using util.DataclassUpdateMixin (#26)

Removed

• Removed mixin class util.DataclassUpdateMixin
  • Only used for updating the configuration object, which now handles updating in its own class

Fixed

• Fixed single-digit filesize numbers in TOML config not matching pattern
• Queue no longer stops when encountering an error, will now skip problematic items until one can be
  played or the queue is empty (#25)

v0.6.1

v0.6.0...v0.6.1

Fixed

• Fixed Lydian completely failing on Windows due to an issue with a path highlighter (#28)

v0.6.0

This is the first public pre-release version for Lydian.

Lydian is currently in a beta state, before its v1.0.0 release it may be unstable or subject to a number of bugs. Pre-releases should work reasonably well, but it should likely be kept to smaller servers for the time being. Below is a basic user-facing overview of Lydian's current commands and features; more detail can be found in Lydian's full changelog, and a longer overview of the project and how to use the bot is at the README.

General features

• Can keep a queue up to a user-defined limit (or no limit) of tracks to play
• Supports any source URL that yt-dlp supports: https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md
• Can queue full playlists/albums from URLs (this can be limited or disabled)
• Filters can be set for URLs or yt-dlp extractor names to control what is allowed to play: https://github.com/svioletg/lydian-discord-bot#url-and-extractor-filtering
• Has role-based permissions to control who can use which commands

Bot commands

"Author" or "command author" refers to the user issuing the command.
Arguments surrounded by angle brackets (<>) are required, while arguments in square brackets ([]) are optional. Brackets should not be included in the actual argument values.

-join

Joins the author's voice channel if they are connected to one, leaving the voice channel it was previously in if applicable. Typically unnecessary to use, as -play will make the bot do this automatically if not connected already.

-leave

Leaves the voice channel the bot is connected to, if any. The user issuing this command must be in the same voice channel as the bot.

-play

If not given a URL, tries to resume a paused or stopped player. Otherwise, the URL is added to the bot's queue, playing immediately if the queue was previously empty.

-pause

Pauses the player. Can be resumed with -play.

-stop

Stops the player. This behaves like most media players do, meaning the currently playing track is saved and will be played from the beginning of -play is used again after stopping.

-skip

Skips the currently item (whether playing, paused, or stopped), starting the next item in the queue if one exists.

-nowplaying

Shows what's currently playing.

-queue [page]

Shows the queue. If the queue is longer than 20 items, it will be separated into pages, where a specific page can be viewed with -queue 2, -queue 3, etc.

-remove <index>

Removes an item at the index position of the queue.

-move <from> <to>

Moves an item from position from of the queue to position to.

-clear

Removes every item from the queue, leaving it empty.

-help

Shows a list of every command available and a brief description of it. This is very bare at the moment, and will be much more user-friendly in the future (#18).