The only required field in a config was `working_directory`. This is now optional (defaults to `.ytdl-sub-working-directory`) making it so a config is no longer required. This sets the stage to onboard noob users who will solely use prebuilt presets in a single subscription file
A follow-up to the prior commit which fixes some bugs in the 'beautiful subscriptions' implementation. A proper migration guide will be written in the wiki once it's ready
With the recent push to 'beautify subscriptions' (see https://github.com/jmbannon/ytdl-sub/releases/tag/2023.10.02), we need the ability to change subscription names from their legacy form:
```
rick_a:
preset:
- "tv_show"
overrides:
tv_show_name: "Rick A"
url: "https://www.youtube.com/channel/UCuAXFkgsw1L7xaCfnd5JJOw"
```
into:
```
tv_show:
"Rick A": "https://www.youtube.com/channel/UCuAXFkgsw1L7xaCfnd5JJOw"
```
This however has implications from ytdl-sub's legacy download archive naming. By default, we write archives to `.ytdl-sub-{subscription_name}-download-archive.json`. If we change the subscription name, the archive will not be found, causing a complete redownload. This new feature gives us the ability to migrate download archives to a new naming schema.
# Migrating to Beautified Subscriptions
## Step 0
BACK UP ALL CONFIG + SUBSCRIPTION FILES!!!!! If something goes wrong, restore your backup and try again and/or ask for help.
## Step 1
Since we know we'll be changing our `subscripion_name` to the value of `tv_show_name`, we can use that in our newly migrated download archive name by setting this within the `tv_show` preset (or whatever your 'base' preset is).
```
presets:
tv_show:
output_options:
migrated_download_archive_name: ".{tv_show_name_sanitized}-download-archive.json"
```
## Step 2
Perform a download as usual, via `ytdl-sub sub ...`. This will load the old archive, and save it into the new archive. You should see `MIGRATION DETECTED` within the logs. Ensure it completes successfully.
Perform another download invocation and ensure you see the `MIGRATION SUCCESSFUL` within the logs.
## Step 3
Now we can set:
```
presets:
tv_show:
output_options:
# rename migrated_download_archive_name to just download_archive_name
download_archive_name: ".{tv_show_name_sanitized}-download-archive.json"
overrides:
tv_show_name: "{subscription_name}"
url: "{subscription_value}"
```
Our download archives now default to our new format, and we set `tv_show_name` + `url` to use subscription values by default.
## Step 4
We can now beautify our subscription.yaml file to:
```
tv_show:
"Rick A": "https://www.youtube.com/channel/UCuAXFkgsw1L7xaCfnd5JJOw"
```
Makes setting yt-dlp's `format` field easier by giving it its own documented plugin.
Old:
```
my_format_preset:
ytdl_options:
format: "best"
```
New:
```
my_format_preset:
format: "best"
```
The old method will still work, so do not worry about updating configs ASAP. However, the option is available to save a few lines 😉
* [BUGFIX] Do not fail with thumbnail issues
* helper
* refactor
* dry-run
* regen fixture
* download arhcive txt -> json in docs, fixtures
* return if no download thumb path
* better test
* more test coverage
* fix playlist last thumbnail move to copy
* playlist working
* channel working
* working directory getting cleaned
* working dir e2e fixture
* cleanup files in any case
* clean up if pre-split returns None
* actually working
* changed plugin format. Need unit tests, documentation, update e2e to filter
* refactor plugin name, matching now working
* description work in progress, need more intricate testing
* tests looks good, regenerate output
* e2e tested, need more failure tests, no dupe variable names, add docs
* unit tests passing again
* more fail tests
* updated with name, default, better docs
* sanitized vars created
* tests updated with new class
* fix nfo tags, no tests only manual inspection
* txt
* dry-run output tested
* readability
* all tests now have files except rolling recent
* better
* almost fixed
* updated txt files w/indents
* files updated, looking good
* rolling recent dry-run test
* Print in main
* Updated docker to use buildx Buildkit
* Updated github actions to include QEMU/buildx to build multiarch containers
* Removed packages after build is done
* Updated documentation to set the correct default thumbnail file name
* Parallel jobs (#1)
Implemented parallel docker builds on Github Actions
* Updated default config file in Docker container
* Replaced Pillow with ffmpeg for thumbnail converting (and updated documentation to change the default thumbnail to -thumb.jpg) (#2)
* Swapped Pillow for ffmpeg for thumbnail conversion
* Further changed default thumbname to include "-thumb" (Kodi needs this to pick up the thumbnail image)
* Added ffmpeg bitexact for reproducability
* Removed extra packages needed for Pillow
* Updated ref check (#3)
* Fixed formatting on if statement for deploy job
Co-authored-by: Maka0 <bas@oddens.net>
* dry-run huge refactor
* small fixes, all tests passing
* file handler, transaction log, metadata
* dry-run actually dry running
* working tests, still have more
* Everything passing besides channel
* fix youtube channel thumbnail issue, need to fix fixtures on test
* lint 10, channel tests passing. Need dry-run for split and merge video
* all tests passing, lint 10, should test dry-run arg
* begin
* chapters class, trying to add chapter splitting and general ffmpeg metadata creation
* working split video
* so very close
* test passing, is reproducible
* lint fixed
* fix docs
* track no longer used in example
* if, then del
* multiple values
* github hash
* youtube.download_strategy split_video
* almost working, need to deal with track_title and other overwrites
* working test
* ci ffmpeg
* sudo -f
* fix code block
* `ytdl_options` defaults for each download strategy
* fix docker, fix defaults getting added, add debugger test
* test playlist as well
* fix with hack
* fix full channel test
* existing
* youtube video url
* WIP, channel /user/ and /c/ should be valid
* unit tests passing
* soundcloud artist url
* soundcloud username
* break configs with name change
* Update all yamls with urls
* update all docs
* more negative url tests
