From 5fb5f087d6a4aa6c3d5bdb30326598dff30c10d6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ond=C5=99ej=20Hru=C5=A1ka?= <ondra@ondrovo.com>
Date: Tue, 12 Oct 2021 00:36:08 +0200
Subject: [PATCH] hashtag bug fix

---
 CHANGELOG.md   | 11 +++++++++--
 src/command.rs | 10 ++++++----
 2 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 759d546..fa464a0 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,10 +1,17 @@
 # Changelog
 
+## v0.4.2
+- Fix URL fragment detected as hashtag
+
 ## v0.4.1
-- Add a translation system using the `locales` folder and `messages.json`
-- Add more trace logging
+- "en" translation fixes
+- add `messages.json`
 - Config files are now parsed as JSON5 = comments are allowed
 
+## v0.4.0
+- Add a translation system using the `locales` folder
+- Add more trace logging
+
 ## v0.3.0
 - Changed config/storage format to directory-based, removed shared config mutex
 - Made more options configurable (timeouts, catch-up limits, etc)
diff --git a/src/command.rs b/src/command.rs
index 9bcb6f9..b0036df 100644
--- a/src/command.rs
+++ b/src/command.rs
@@ -135,13 +135,13 @@ static RE_PING: once_cell::sync::Lazy<Regex> = Lazy::new(|| command!(r"ping"));
 static RE_ANNOUNCE: once_cell::sync::Lazy<Regex> =
     Lazy::new(|| Regex::new(r"(?:^|\s|>|\n)[\\/]announce\s+(.*)$").unwrap());
 
-static RE_A_HASHTAG: once_cell::sync::Lazy<Regex> = Lazy::new(|| Regex::new(r"(?:^|\b|\s|>|\n)#(\w+)").unwrap());
+static RE_A_HASHTAG: once_cell::sync::Lazy<Regex> = Lazy::new(|| Regex::new(r"(?:^|\s|>|\n)#(\w+)").unwrap());
 
 pub static RE_NOBOT_TAG: once_cell::sync::Lazy<Regex> =
-    Lazy::new(|| Regex::new(r"(?:^|\b|\s|>|\n)#nobot(?:\b|$)").unwrap());
+    Lazy::new(|| Regex::new(r"(?:^|\s|>|\n)#nobot(?:\b|$)").unwrap());
 
 pub static RE_HASHTAG_TRIGGERING_PLEROMA_BUG: once_cell::sync::Lazy<Regex> =
-    Lazy::new(|| Regex::new(r"(?:^|\b|\s|>|\n)#\w+[^\s]*$").unwrap());
+    Lazy::new(|| Regex::new(r"(?:^|\s|>|\n)#\w+[^\s]*$").unwrap());
 
 pub fn parse_status_tags(content: &str) -> Vec<String> {
     debug!("Raw content: {}", content);
@@ -563,6 +563,8 @@ mod test {
         assert!(RE_A_HASHTAG.is_match("#городДляЛюдей"));
         assert!(RE_A_HASHTAG.is_match("foo #banana gfdfgd"));
         assert!(RE_A_HASHTAG.is_match("foo #городДляЛюдей aaa"));
+        assert!(!RE_A_HASHTAG.is_match("foo https://google.com/#banana gfdfgd"));
+        assert!(!RE_A_HASHTAG.is_match("foo https://google.com/?foo#banana gfdfgd"));
 
         for (i, c) in RE_A_HASHTAG.captures_iter("foo #banana #χαλβάς #ласточка").enumerate() {
             if i == 0 {
@@ -590,7 +592,7 @@ mod test {
         assert!(!RE_NOBOT_TAG.is_match("banana #tag sdfsd"));
         assert!(!RE_NOBOT_TAG.is_match("banana #nobotanicals sdfsd"));
         assert!(RE_NOBOT_TAG.is_match("#nobot"));
-        assert!(RE_NOBOT_TAG.is_match("aaa#nobot"));
+        assert!(RE_NOBOT_TAG.is_match("aaa\n#nobot"));
         assert!(RE_NOBOT_TAG.is_match("aaa #nobot"));
         assert!(RE_NOBOT_TAG.is_match("#nobot xxx"));
         assert!(RE_NOBOT_TAG.is_match("#nobot\nxxx"));