Canvas langchain document loader

Features:

Indexes Dropbox Files

The following file types are supported: md htm html docx xls xlsx pptx pdf rtf txt

(doc support would require libreoffice, so has not been implemented in this library)

Running locally

You can do the install as described below, or build/run the provider dockerfile.

Docker

Edit dropbox-test.py, fill in the correct auth and one of folder_path, file_path, or file_paths.

Run (this also builds docker):

docker run -it $(docker build -q .)

Installation

Install dependencies:

pip install -r requirements.txt

Usage example:

To use a refresh token, pass in api_key along with api_secret

from dropbox_files import DropboxLoader

auth = {
	"access_token": "ACCESS_TOKEN_FROM_OAUTH",
	"refresh_token": "REFRESH_TOKEN",
	"expires_at": "EXPIRE_TIMESTAMP"
}

loader = DropboxLoader(
	auth = auth,
	folder_path = "/PATH_TO_FOLDER_OR_EMPTY_STRING",
	# file_path = "/PATH/TO_FILE/File.extension",
	# file_paths = [
	# 	"/PATH/TO_FILE/File 1.extension",
	# 	"/PATH/TO_FILE/File 2.extension"
	# ],
	# api_key = "API_KEY",
	# api_secret = "API_SECRET",
)

documents = loader.load()

print("\nDocuments:\n")
print(documents)

print("\nInvalid files:\n")
print(loader.invalid_files)
print("")

print("\nErrors:\n")
print(loader.errors)
print("")

If errors are present, loader.errors will contain one list element per error. It will consist of an error message (key named message) and if the error pertains to a folder issue, a key named folder, for files a file key. Right now the library only catches Dropbox API specific exceptions.