HTML entities in the episode names

VortX wrote a month ago: 1

Hi,

some episode names contain HTML entities (e.g. `&`) in their name in the API response of the endpoint `/episodes/:id`. The episode names are correctly displayed in the corresponding episode pages. Not all episodes are affected by this behaviour.

Example 1: BAD

curl -s https://api.tvmaze.com/episodes/47886 | jq -r .name
# Mergers & Acquisitions

https://www.tvmaze.com/episodes/47886
Mergers & Acquisitions

Example 2: BAD

curl -s https://api.tvmaze.com/episodes/58378 | jq -r .name
# Cupid & Psycho

https://www.tvmaze.com/episodes/58378
Cupid & Psycho

Example 3: GOOD

curl -s https://api.tvmaze.com/episodes/109658 | jq -r .name
# Halt & Catch Fire

https://www.tvmaze.com/episodes/109658
Halt & Catch Fire

The name in the API response should not contain these HTML entities but the characters they represent. Maybe the episode names (and other potentially affected attributes) can be corrected in bulk.

Thanks!

LouisWu wrote a month ago: 1

This actually works as expected. Some titles simply do contain HTML entities. So the API correctly returns those unchanged.
When displayed on site, your browser translates them and it looks normal.

Now, whether we should pursue "squashing all the entities" is a question for the admins.

david wrote a month ago: 2

Moving this to data - no (API) bug here, just a few cases of a user literally inputing "&amp" in the episode name.

Thanks for reporting though, we'll go through them and fix manually soon.