Python tools to work with the Notion API, including retrieving databases as Pandas dataframes.
See DEVELOP.md for information on testing/docs.
pip install notion-toolsDepends on pandas and
notion-client.
from notion_tools import get_notion_token, get_notion_client, database_to_dataframe
database_id = # database id retrievable from the share URL
notion_client = get_notion_client(get_notion_token())
df = database_to_dataframe(notion_client, database_id)To find the database id, take the hex string from the URL for sharing the page:
https://www.notion.so/yourworkspace/{database_id}?v={other_hex_stuff}
notion_tools.get_notion_token() -> str
Gets a Notion API token.
Either from the NOTION_TOKEN environment variable or interactively.
Returns:
The Notion API token.
Return type:
str
notion_tools.get_notion_client(token: Optional[str] = None) -> notion_client.client.Client
Gets a Notion API client.
The Notion API client used is https://github.com/ramnes/notion-sdk-
py
Parameters:
**token** -- The Notion API Integration Token.
Returns:
The Notion API client.
Return type:
NotionClient
notion_tools.nonpaginated(endpoint, wait=0.1)
Create a non-paginated version of a Notion endpoint.
Wraps a paginated Notion endpoint and when called, will return a
response object that concats all the pages. Waits *wait* seconds on
each iteration to the API.
-[ Example ]-
response =
nonpaginated(notion_client.blocks.children.list)(block_id)
>>``<<>>`<<
notion_tools.database_to_dataframe(notion_client: notion_client.client.Client, database_id: str, default_date_handler: str = 'ignore_end', date_handlers: Optional[dict] = None) -> pandas.core.frame.DataFrame
Extracts a Notion Database as a Pandas DataFrame.
Parameters:
* **notion_client** -- The Notion API client.
* **database_id** -- The Notion Database ID. This identifier can
be found in the URL of the database.
* **default_date_handler** (*{"ignore_end"**, **"mangle"**,
**"multiindex"}*) -- The default date handler. See Notes below
on how to use this.
* **date_handlers** -- Specify per-column date handlers.
Returns:
The Notion Database as a Pandas DataFrame.
Return type:
pd.DataFrame
-[ Notes ]-
Notion date properties are represented as 2-tuples with a start and
end timestamps. If there is only a single date in the property, it
is encoded as a start date with a null end date. There are several
options on how to encode this into the resulting dataframe.
"ignore_end":
Keep only the start date of the date object and keep column name
the same as the property name.
"mangle":
For each date property named "foo" in the Notion table, create a
"foo_start" and a "foo_end" column.
"multiindex":
Create a MultiIndex for the columns where the top level contains
the property names and the second level contains "start" and
"end" for date properties.
notion_tools.users_to_dataframe(notion_client: notion_client.client.Client)
Extract all Notion users as a Pandas DataFrame.
Parameters:
**notion_client** -- The Notion API client.
Returns:
The Notion users as a Pandas DataFrame.
Return type:
pd.DataFrame
-[ Notes ]-
If users are deleted from your Notion workspace, they will not be
returned by the API, even if they are still present in Person
properties in your databases.