From 2401da8e5f416b2bd7716a4f24518229e3a8150e Mon Sep 17 00:00:00 2001
From: Guillem Hernandez Sola <guillem.hernandez.sola@gmail.com>
Date: Thu, 23 Apr 2026 20:01:57 +0200
Subject: [PATCH] Fixed some truncation

---
 jenkins/3catTw         |  3 ++-
 twitter2bsky_daemon.py | 13 +++++++++----
 2 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/jenkins/3catTw b/jenkins/3catTw
index e8bae75..a38bbdc 100644
--- a/jenkins/3catTw
+++ b/jenkins/3catTw
@@ -46,7 +46,8 @@ pipeline {
                         fastfeedparser \
                         beautifulsoup4 \
                         charset-normalizer \
-                        Pillow
+                        Pillow \
+                        grapheme
 
                     # Verify required imports
                     "${VENV_DIR}/bin/python" -c "import fastfeedparser; print('fastfeedparser OK')"
diff --git a/twitter2bsky_daemon.py b/twitter2bsky_daemon.py
index f3d2ca9..699f69c 100644
--- a/twitter2bsky_daemon.py
+++ b/twitter2bsky_daemon.py
@@ -19,6 +19,7 @@ from playwright.sync_api import sync_playwright
 from moviepy import VideoFileClip
 from bs4 import BeautifulSoup
 from PIL import Image
+import grapheme  # add to imports at top
 
 # --- Configuration ---
 LOG_PATH = "twitter2bsky.log"
@@ -26,7 +27,7 @@ STATE_PATH = "twitter2bsky_state.json"
 SCRAPE_TWEET_LIMIT = 30
 DEDUPE_BSKY_LIMIT = 30
 TWEET_MAX_AGE_DAYS = 3
-BSKY_TEXT_MAX_LENGTH = 275
+BSKY_TEXT_MAX_LENGTH = 300
 DEFAULT_BSKY_LANGS = ["ca"]
 
 VIDEO_MAX_DURATION_SECONDS = 179
@@ -108,6 +109,9 @@ _cache = _RunCache()
 def reset_caches():
     _cache.clear()
 
+def grapheme_len(text):
+    """Return the grapheme cluster count, matching Bluesky's character counting."""
+    return grapheme.length(text)
 
 # --- Custom Classes ---
 class ScrapedMedia:
@@ -1160,16 +1164,17 @@ def find_tail_preservation_start(text, primary_non_x_url):
 
 
 def truncate_text_safely(text, max_length=BSKY_TEXT_MAX_LENGTH):
-    if len(text) <= max_length:
+    if grapheme_len(text) <= max_length:
         return text
 
-    truncated = text[: max_length - 3]
+    # Truncate by grapheme clusters
+    clusters = list(grapheme.graphemes(text))
+    truncated = "".join(clusters[: max_length - 3])
     last_space = truncated.rfind(" ")
     if last_space > TRUNCATE_MIN_PREFIX_CHARS:
         return truncated[:last_space] + "..."
     return truncated + "..."
 
-
 def truncate_text_preserving_tail(text, tail_start, max_length=BSKY_TEXT_MAX_LENGTH):
     if (
         not text