metadata

license: cc-by-4.0
pretty_name: Chabiko Stream
new_version: WitchesSocialStream/Four-Leaf-Clover

Superseded

see WitchesSocialStream/Four-Leaf-Clover for the newer version!

CURRENTLY OFFLINE

I really didn't like the code I used for 4chan scraping. It was overly complex and prone to failures. So I'm pausing data collection on this for now while I think of a better solution.

Dataset Card for Chabiko Stream

"Rule 34 (Part AI). Anything that can be in a dataset, will be in one eventually given enough time." - KaraKaraWitch

XChan + Chibiko -> Chabiko

chabiko is a XChan scraper. and as such, Chabiko Stream is a daily dump of threads and posts of X chan posts.

ArchiverModes

Broadly, the Archiver archives in 2 modes:

"Archive" [The board supports archives of closed threads.]
- Archives are refreshed at a max of 10 minutes. If boards are moving faster, refreshes happen more often.
"PostStream" [The Board does not support archives. As such, fall back to Streaming as posts. (This is similar to MissingKeys.)]
- Streams are refreshed at a max of 20 seconds. If boards are moving faster, refreshes happen more often.

Supported Sites

4chan
[TBD] Futaba Channel

Formats

Both Archive modes are written to jsonl:

For Archive mode, each json line represents an entire thread dump.
For PostStream, each json represents a single post.

Models

Refer to the following pydantic model for parsing:

class AttachmentData(pydantic.BaseModel):
    attachment_url: str
    filename: str


class Post(pydantic.BaseModel):
    board: str
    thread: int
    pid:int
    name: str
    msg: str
    attachment: Optional[AttachmentData] = None
    posted:int
    
class Thread(pydantic.BaseModel):
    
    title:Optional[str]
    posts:list[Post]

Notices

Archiving only text for now since I need to see how much data images would take up.
Due to potential abuse, we are closing community posts. Contact KaraKaraWitch in other known channels.

License

Apache 2.0.