README.md

notion-to-text as lambda

Serveless version compatible with AWS Lambda. Currently hosted on my AWS at ntt.fpira.com.

This version uses chrome-aws-lambda in place of puppeteer to reduce the size of zip package to deploy.

Known limits

As a beta side-project some rough edges still exist.

• ~15 secs response time, this is because a new puppeteer-core instance is loaded per each invocation
• some invocations may lead to internal server errors

Usage

Just prepend ntt.fpira.com/ in front of a notion.site URL, e.g.

https://pirafrank.notion.site/My-toolbelt-c3e17a2462d64b549e3ec7009e6f3071 -> http://ntt.fpira.com/pirafrank.notion.site/My-toolbelt-c3e17a2462d64b549e3ec7009e6f3071

and you'll get a text/plain response with page title and content.

Available endpoints

Lambda and standalone server share same API format.

Notion page URL	Method	Endpoint	Response type
https://pirafrank.notion.site	GET	https://ntt.fpira.com/json/pirafrank.notion.site	application/json
https://pirafrank.notion.site	GET	http://ntt.fpira.com/raw/pirafrank.notion.site	text/plain
https://pirafrank.notion.site	GET	http://ntt.fpira.com/text/pirafrank.notion.site	text/plain

To get text/plain responses you may omit the /text path.

In case of error, you'll always get an application/json response.

First deploy

# same as --omit=dev
# Docs: https://docs.npmjs.com/cli/v8/using-npm/config#only
npm i --only=production

zip -r function.zip .

aws lambda create-function \
--function-name notion-to-text \
--region eu-west-1 \
--code S3Bucket=notion-to-text,S3Key=function.zip \
--role ARN-ROLE \
--runtime nodejs14.x \
--handler index.handler

aws lambda update-function-configuration \
--function-name notion-to-text \
--timeout 30 \
--memory-size 1536

Update function code

aws s3 cp function.zip s3://notion-to-text/function.zip

aws lambda update-function-code \
--function-name notion-to-text \
--s3-bucket notion-to-text \
--s3-key 'function.zip' \
--publish

License

MIT