Copying a message turns \n into Unicode line separator

opk12 commented 2 years ago

Description

A Unicode line separator character is copied in place of the \nthat was sent.

Steps to reproduce

Select, and copy, a message with json body a\nb.
Paste into the github new issue form, it becomes ab (without newline).
Paste into nano, it becomes a b (space without newline).
Paste into gvim, it becomes a, followed by an arrow symbol, b.

printf '%s' '<paste here>' | hexdump -C shows that e2 80 a8 (which apparently is UTF-8 for the Unicode "LS Line Separator" \u2028) is copied in place of the usual Unix newline 0a ASCII line feed.

Version information

Quaternion 0.0.95.1 on Fedora 35 Quaternion 0.0.9.5 on Debian 11

opk12 commented 2 years ago

For completeness - this is not specifically about HTML messages, it also happens with plain-text messages without formatted_body.

KitsuneRal commented 2 years ago

Thanks; looks like text conversions along the way accidentally de-screen \n into an actual line-break.

quotient-im / Quaternion