eoussama / anusic-api Goto Github PK

There needs to be a constant, predictable object that contains the data and some additional info.
Something like this:

{
    "hasError": false,
    "error": { "name": "", "description": "" },
    "data": {}
}

Errors:

• Invalid route
• Invalid ID (Invalid data type passed as an ID)
• Unknown ID (MAL ID does not match anything cached.

When an Anime title is clicked, a model containing all info and themes should be opened.

• Modal
• Loader feedback
• Anime info
• Themes

[ ] Console logs
[ ] Log dumps

If no requested info was found, attempt scrapping, if anything was found update the cache.

• Add GetAnimeInfoByID
• Added a timestamp for when the entry was last scraped
• Add creation and last update for scraped data

Lyrics should be:

• Scraped from https://www.animesonglyrics.com/
• Cached separately

It appears that the AnimetThemes subreddit has removed their wiki, which has caused the API to stop working.
Fortunately, I had a sizable cache that was designed to help in situations like this. However, as long as we continue to rely on this cache, the content provided by the API will be limited to late 2022 and earlier.

I'm currently exploring other options to restore the full functionality of the API and will keep the community updated on any developments.

Add a franchise endpoint that returns said franchise with linked Anime titles.

Alt names should be comma-separated.

Name the log file after the timestamp.

Description
Certain Anime titles have several theme collections, a common case is dubbed themes.
In some other case, like One Piece, there are 3 different groups of themes (Original Japanese, Funimation, and 4Kids)

Type tabs should be stickied to top.

Request analytics that can either be logged into a file to access from a third-party platform.

• logs route
• logs/:id route
• Add auth
• Log requests

Requests often exceed the timeout time and end up missing on a chunk of scraped data.

Tabs that have nothing to display shouldn't be displayed in the first place.

Describe the bug
Logger always outputs the host file instead of the initiator file.

To Reproduce
Steps to reproduce the behavior:

1. Call Log from any file

Add query params to /anime that allows for searching by name and year.
searching by name should also take alt names into consideration.

Both secondary name and collection name come inside an immediate paragraph after the title.
The only difference is that the collection name is contained inside of a strong tag.

Empty theme name's links should be appended to the last valid theme's

Add theme variant.

A single audio layer should be displayed in the footer of the modal.
This should decrease the strain and requests to the theme's repository.

Consider using maps.

Transform the app into a PWA.

To avoid spamming the scrapping targets, there needs to exist a caching workaround that does not utilize any in-memory caching libraries. Dumping the data in a JSON file would do the trick.

The dump file should be updated frequently, based on a scheduler of a 1-day interval would be preferable.

• Dumping the data into a JSON file
• [ ] Frequent updates to the dump file

Only add the modal when an entry is clicked.

Display the number of themes available for an Anime title given the type and the collection.

With issue #34 the returned data's structure will be changed, the demo app should take the change into consideration.

Logs should be dumped somewhere including all errors and warnings.

Opening/ending players of the same index overlap eachother.

Return something instead of an empty object

Splitting code and asyncrpnizing calls

Description
Raise a warning if the fetched info does not match the total scrapped Anime titles.

Empty collections can be hidden.

Convert altName into a list

Anime titles pre-2000 are scrapped as XXs, thus parsing them into a number type backfires.

Example: http://anusic-api.herokuapp.com/api/v1/anime/21 returns {"id":21,"name":"One Piece","altNames":["Original Japanese Version"],"year":0}

• Parse era
• Emit the year entirely

Treat theme tables that are missing headers as the whole table seems to be skipped if no title is found. e.g Naruto, id 20.

Cache themes, articles, and Anime separately.

Now that the project switched languages, a memory caching solution would be more ideal, preferably Redis.
Rough sketch of the global lifecycle of the API:

Progression:

• Dockerizing the API
• Redis connection
• Caching aware API
• Memory dump

Description
Anime title themes

Structure

• Name

The name of the theme

• MALID

The associated Anime MAL ID

• Type

The type of the theme (Opening or Ending)

• Order

The chronological order of the opening (1, 2, ...)

• Sources #15

List of the theme's link info

• Link

The link to the theme song

• Format

The format of the source (Webm...)

• Resolution

The video resolution (1080, 480...)

• Lyrics

Whether it has lyrics

• Transition

Whether the episode transitions into the ending song

• IsOver

Is the theme laid over the episode

• Spoilers

Whether the theme contains spoilers

• NSFW

Whether the theme has NSFW content

• Episodes

The episodes in which this theme is featured

Add versioning for future proofing.