Files

This section demonstrates common file operations you can perform with the WriftAI Python client. These examples are not exhaustive — check the client reference for all options.

Upload a file

files.upload takes a File, which wraps the file's contents along with an optional mime_type. If provided, mime_type overrides any automatically inferred value; if omitted, the MIME type is inferred from the file's name or extension, falling back to application/octet-stream if it can't be determined.

Large files are automatically split into parts and uploaded concurrently, so you don't need to do anything differently for big files, just pass them the same way.

Upload from a file path

from pathlib import Path
from wriftai import File

upload_result = wriftai.files.upload(File(Path("report.pdf")))

Upload raw bytes

There's no filename to infer from here, so this falls back to application/octet-stream unless mime_type is provided:

from wriftai import File

upload_result = wriftai.files.upload(File(b"raw file contents"))

File-like objects

Binary file objects work as content too:

from wriftai import File

with open("chart.png", "rb") as f:
    upload_result = wriftai.files.upload(File(f))

If the object has a name attribute, as file objects opened with open() do, it's used for MIME type inference the same way a path's extension is.

Override the inferred MIME type

Pass mime_type explicitly to override the inferred or fallback value:

from wriftai import File

upload_result = wriftai.files.upload(
    File(b"raw file contents", mime_type="application/pdf")
)

Track upload progress

Pass on_progress to be notified as the upload proceeds. Each time a chunk is sent, it's called with the number of bytes just sent and the total size of the file.

from wriftai import File


def on_progress(bytes_delta: int, total_bytes: int) -> None:
    print(f"sent {bytes_delta} of {total_bytes} bytes")


upload_result = wriftai.files.upload(
    File(Path("report.pdf")), on_progress=on_progress
)

on_progress may be called concurrently from multiple threads while parts upload together, once per chunk sent. If it updates any shared state, make sure that update is thread-safe.