Fixes #1266: modifies sanitize.js to allow Telescope to parse img tags with data uris

[chrispinkney]

Issue This PR Addresses

Fixes #1266

Type of Change

• Bugfix: Change which fixes an issue
• New Feature: Change which adds functionality
• Documentation Update: Change which improves documentation
• UI: Change which improves UI

Description

• Modified sanitize.js to allow for img tags to use data uris as src so they can be successfully parsed and indexed by Telescope.
• Adds a test to ensure data uris are sanitized properly.
• Adds two more http and https schema tests for img tag with a data uri src.
• Adds blockquote test and cite sanitation.

Hey everyone!

As per issue #1266, this PR allows images which have a src data uri type to be properly indexed. It passes all the tests on my end thus showing that data uris are being allowed by Telescope.

I tried to ensure that this works as expected by purging my redis db via docker exec -it telescope_redis_1 redis-cli flushall but it seems that once a post is indexed, it is how it is and forever shall be the way it was initially indexed. So instead of going back in time I disallowed all html attributes and waited for a post to come in. Once one arrived I noticed that it had no tags at all, just plaintext, so I reverted back to normal and allowed imgs with a data uri.

While this isn't exactly a perfect sign that everything works as expected, given that npm test sanitize-html runs and returns that all tests pass, and specifically: <img src="data:image/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==" /> returned as expected, I feel that Telescope should now properly parse imgs with a data uri.

Let me know, thanks!

Checklist

• Quality: This PR builds and passes our npm test and works locally
• Tests: This PR includes thorough tests or an explanation of why it does not
• Screenshots: This PR includes screenshots or GIFs of the changes made or an explanation of why it does not (if applicable)
• Documentation: This PR includes updated/added documentation to user exposed functionality or configuration variables are added/changed or an explanation of why it does not(if applicable)

[chrispinkney]

So this is my first solo rebase that I did because I had like 3 commits and figured it'd be too messy. I also pushed them to my origin by accident, so I git rebase HEAD~3 -i and squashed the other two, then force pushed this to my branch and made a PR. I think I did all the steps right so let me know if I didn't!

[humphd]

@chrispinkney can you fix the title of this PR? fixes issue-1266 is not useful, as it doesn't reference what you're working on (the issue number is good, but describe the work too).

[humphd]

Can you also add a test for protocol-less URLs? (i.e., <img src="//www.telescope.com/image.jpg" />)

[chrispinkney]

Can you also add a test for protocol-less URLs? (i.e., <img src="//www.telescope.com/image.jpg" />)

I was going to add that too but there's already one in place that does the same thing here.

Ah nvm, misread. Will do.

[chrispinkney]

@chrispinkney can you fix the title of this PR? fixes issue-1266 is not useful, as it doesn't reference what you're working on (the issue number is good, but describe the work too).

Is this better?

[tonyvugithub]

@chrispinkney @humphd I wonder why the github action for macos failed. Should we retry with another push with a revised single commit and see if all are passed? Right now one check has failed.

This is the error in the macos build
[HPM] Proxy rewrite rule created: "^/admin/build/log" ~> "/log" [ 2020-11-03 23:27:02.081 ] ERROR (3824 on Mac-1604444996117.local): Something went wrong with search indexing

[humphd]

That failure is not actually causing the tests to fail (i.e, the test expects the failure). What's causing the failure is some process hanging and blocking the tests from exiting correctly (see #1284):

A worker process has failed to exit gracefully and has been force exited. This is likely caused by tests leaking due to improper teardown. Try running with --detectOpenHandles to find leaks.

Test Suites: 26 passed, 26 total
Tests:       169 passed, 169 total
Snapshots:   0 total
Time:        15.689 s
Ran all test suites.
Error: The operation was canceled.

[tonyvugithub]

@humphd I noticed that too. So I guess this PR is good to go then if thats the case