The HTML string was being converted to a local 8-bit string prior to parsing with tidy.

According to Qt documentation, this works on Unix systems as toLocal8Bit() is equivalent to toUtf8(), however, it may not on Windows, and may result in character replacement.

Gitlab: #1536
Change-Id: I7aec98cffbd9d72771d1c489f52b38dbb3a1ed9f

Change-Id: I8d5f968fbedbc884c91416246049a0ef4cd652eb

Andreas Traczyk authored 1 year ago and Sébastien Blin committed 1 year ago

Depend on tidy API for attribute extraction rather than regexes.

1. htmlparser methods return nodes instead of pre-parsed strings
2. htmlparser provides some methods to extract text/attr from nodes

Gitlab: #1248
Change-Id: I367d703680938fb0b7c5055ac41e079c1322da30

Andreas Traczyk authored 2 years ago and Sébastien Blin committed 1 year ago

Introduces MessageParser to encapsulate text treatment for raw text messages.

The async parsing sequence is as follows:
- Markdown -> HTML (md4c)
- link coloration (tidy-html5)
- notify UI
- request link preview info from PreviewEngine for the first link
- Preview engine uses QtNetwork instead of QtWebengine
- Linkification is handled by MessageParser instead of linkify.js

QtWebengine is no longer required for message parsing.

Gitlab: #1033
Gitlab: #855
Change-Id: Ief9b91aa291caf284f08230acaf57976f80fa05b