mirrors/gotosocial
1
0
Fork 0
mirror of https://github.com/superseriousbusiness/gotosocial.git synced 2025-11-08 23:37:29 -06:00
Code Issues Projects Releases Packages Wiki Activity Actions

[feature] Parse content warning as HTML, serialize via API to plaintext

Browse source
This commit is contained in:
tobi 2025-03-06 15:36:58 +01:00
commit 22ce924129
47 changed files with 947 additions and 553 deletions
Download patch file Download diff file Expand all files Collapse all files

20
internal/text/sanitize.go
View file

@ -163,27 +163,21 @@ var regular *bluemonday.Policy = func() *bluemonday.Policy {
// Source: https://github.com/microcosm-cc/bluemonday#usage
var strict *bluemonday.Policy = bluemonday.StrictPolicy()
// removeHTML strictly removes *all* recognized
// HTML elements from the given string.
func removeHTML(in string) string {
return strict.Sanitize(in)
}
// SanitizeToHTML sanitizes only risky html elements
// SanitizeHTML sanitizes only risky html elements
// from the given string, allowing safe ones through.
func SanitizeToHTML(in string) string {
func SanitizeHTML(in string) string {
return regular.Sanitize(in)
}
// SanitizeToPlaintext runs text through basic sanitization.
// This removes any html elements that were in the string,
// and returns clean plaintext.
func SanitizeToPlaintext(in string) string {
// RemoveHTML runs text through strict sanitization.
// This removes any html elements that were in the
// string, and returns pruned plaintext.
func RemoveHTML(in string) string {
// Unescape first to catch any tricky critters.
content := html.UnescapeString(in)
// Remove all detected HTML.
content = removeHTML(content)
content = strict.Sanitize(content)
// Unescape again to return plaintext.
content = html.UnescapeString(content)