API returns truncated inscriptions

E.g

$ curl --compressed -s 'https://openbenches.org/data.json/?bench=8553&format=raw'  | jq .features[].properties.popupContent
"GREATER LIVE HATH<br />\r\nNO MAN OR WOMAN<br />\r\n<br />\r\nTO THE MEMORY OF<br />\r\nCHARLES THOMAS<br />\r\nKENNEDY VC<br />\r\n<br />\r\nHIGHLAND LIGHT INFANTRY<br />\r\nBOER WAR<br />\r\n<br />\r\nBORN …"

compare to what’s shown at https://openbenches.org/bench/8553 Truncation seems to be due to https://github.com/edent/openbenches/blob/baa2fa5dc355b0a65c9739233ddc58b25ddbf919/www/mysql.php#L252 which is used both to get text to display on map, in which context it’s reasonable to truncate, and for API where it doesn’t seem reasonable to truncate.

The API section of the README doesn’t mention that inscription text may be truncated.

The indiscriminate appending of … means inscriptions shown on the map always look like they’ve been truncated regardless of whether they have.

[edent]

I like to keep default behaviour the same wherever possible. That way cached requests don’t break in unexpected ways, and anyone who hasn’t seen the update doesn’t become confused.

[arizonagroovejet]

Ah, OK. I see the README does say as such. In the PR I submitted not truncating was the default and you didn’t mention intention to change it so I was expecting that behaviour. Out of curiosity, why make people have to explicitly request that inscriptions are complete?