From 9477fd7eba9bda6813b65c6c54380904892ca35e Mon Sep 17 00:00:00 2001
From: tobi <31960611+tsmethurst@users.noreply.github.com>
Date: Sun, 8 Dec 2024 16:03:00 +0100
Subject: [PATCH] [feature] Allow partial-word hashtags using non-breaking
spaces (#3606)
* [feature] Allow partial-word hashtags using non-breaking spaces
* update docs
---
docs/user_guide/posts.md | 3 +++
internal/text/plain_test.go | 13 +++++++++++++
internal/text/util.go | 30 ++++++++++++++++++++++++++++--
3 files changed, 44 insertions(+), 2 deletions(-)
diff --git a/docs/user_guide/posts.md b/docs/user_guide/posts.md
index 1f718cfae..c45ad4bcb 100644
--- a/docs/user_guide/posts.md
+++ b/docs/user_guide/posts.md
@@ -285,6 +285,9 @@ For accessibility reasons, it is considerate to use upper camel case when you're
You can include as many hashtags as you like within a GoToSocial post, and each hashtag has a length limit of 100 characters.
+!!! tip
+ To end a hashtag, you can simply use a space, for example in the text `this #soup rules`, the hashtag is terminated by a space so `#soup` becomes the hashtag. However, you can also use a pipe character `|`, or the unicode characters `\u200B` (zero-width no-break space) or `\uFEFF` (zero-width space), to create "partial-word" hashtags. For example, with input text `this #so|up rules`, only the `#so` part becomes the hashtag. Likewise, with the input text `this #soup rules`, which contains an invisible zero-width space after the o and before the u, only the `#so` part becomes the hashtag. See here for more information on zero-width spaces: https://en.wikipedia.org/wiki/Zero-width_space.
+
## Input Sanitization
In order not to spread scripts, vulnerabilities, and glitchy HTML all over the place, GoToSocial performs the following types of input sanitization:
diff --git a/internal/text/plain_test.go b/internal/text/plain_test.go
index 48280bb44..fac54a38e 100644
--- a/internal/text/plain_test.go
+++ b/internal/text/plain_test.go
@@ -36,6 +36,8 @@
moreComplexExpected = "Another test @foss_satan
#Hashtag
Text
:rainbow:
"
withUTF8Link = "here's a link with utf-8 characters in it: https://example.org/söme_url"
withUTF8LinkExpected = "here's a link with utf-8 characters in it: https://example.org/söme_url
"
+ withFunkyTags = "#hashtag1 pee #hashtag2\u200Bpee #hashtag3|poo #hashtag4\uFEFFpoo"
+ withFunkyTagsExpected = "#hashtag1 pee #hashtag2\u200bpee #hashtag3|poo #hashtag4\ufeffpoo
"
)
type PlainTestSuite struct {
@@ -136,6 +138,17 @@ func (suite *PlainTestSuite) TestDeriveHashtagsOK() {
suite.Equal("올빼미", tags[0].Name)
}
+func (suite *PlainTestSuite) TestFunkyTags() {
+ formatted := suite.FromPlain(withFunkyTags)
+ suite.Equal(withFunkyTagsExpected, formatted.HTML)
+
+ tags := formatted.Tags
+ suite.Equal("hashtag1", tags[0].Name)
+ suite.Equal("hashtag2", tags[1].Name)
+ suite.Equal("hashtag3", tags[2].Name)
+ suite.Equal("hashtag4", tags[3].Name)
+}
+
func (suite *PlainTestSuite) TestDeriveMultiple() {
statusText := `Another test @foss_satan@fossbros-anonymous.io
diff --git a/internal/text/util.go b/internal/text/util.go
index 204c64838..af45cfaf0 100644
--- a/internal/text/util.go
+++ b/internal/text/util.go
@@ -38,8 +38,34 @@ func isPermittedInHashtag(r rune) bool {
// is a recognized break character for before
// or after a #hashtag.
func isHashtagBoundary(r rune) bool {
- return unicode.IsSpace(r) ||
- (unicode.IsPunct(r) && r != '_')
+ switch {
+
+ // Zero width space.
+ case r == '\u200B':
+ return true
+
+ // Zero width no-break space.
+ case r == '\uFEFF':
+ return true
+
+ // Pipe character sometimes
+ // used as workaround.
+ case r == '|':
+ return true
+
+ // Standard Unicode white space.
+ case unicode.IsSpace(r):
+ return true
+
+ // Non-underscore punctuation.
+ case unicode.IsPunct(r) && r != '_':
+ return true
+
+ // Not recognized
+ // hashtag boundary.
+ default:
+ return false
+ }
}
// isMentionBoundary returns true if rune r