From 13a00ca66708f257f0e1d5664147c159c02d519c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Andr=C3=A9=20Gl=C3=BCpker?= <git@wgmd.de>
Date: Sat, 25 May 2024 13:08:53 +0200
Subject: Build diff for changes

---
 urldelta.py | 30 +++++++++++++++++++++++++++---
 1 file changed, 27 insertions(+), 3 deletions(-)

(limited to 'urldelta.py')

diff --git a/urldelta.py b/urldelta.py
index 569e201..5584966 100755
--- a/urldelta.py
+++ b/urldelta.py
@@ -1,6 +1,7 @@
 #!/usr/bin/env python3
 
 import datetime
+import difflib
 import sqlite3
 import time
 import requests
@@ -19,6 +20,28 @@ def extract_text(content) -> str:
     return soup.get_text(separator="\n", strip=True)
 
 
+def compare_to_html(
+    text_old: str,
+    text_new: str,
+) -> str:
+    if not text_old or not text_new:
+        return "N/A"
+
+    output = ["<pre>"]
+    for line in difflib.Differ().compare(text_old.splitlines(), text_new.splitlines()):
+        if line.startswith("+"):
+            output.append(f"🟢 {line.strip()[2:]}")
+        elif line.startswith("-"):
+            output.append(f"🔴 {line.strip()[2:]}")
+        elif line.startswith("?"):
+            output.append(f"🔵 {line.strip()[2:]}")
+        else:
+            output.append(f"<small>🔵 {line.strip()}</small>")
+    output.append("</pre>")
+
+    return "\n".join(output)
+
+
 def get_page_delta(url):
     conn = sqlite3.connect(os.path.join(os.path.dirname(__file__), "database", "website_data.db"))
     cursor = conn.cursor()
@@ -87,16 +110,17 @@ def get_page_delta(url):
             )
             conn.commit()
 
-    cursor.execute("SELECT headers, content, fetch_date FROM deltas WHERE website_id = ?", (id,))
+    cursor.execute("SELECT extracted_old, extracted_new, fetch_date FROM deltas WHERE website_id = ?", (id,))
 
     updates = []
     for update in cursor.fetchall():
+        extracted_old, extracted_new, fetch_date = update
         updates.append(
             RSSItem(
                 title=f"Change on {url}",
                 url=url,
-                content=f"Headers: {update[0]}\n\nContent: {extract_text(update[1])}".replace("\n", "<br>"),
-                date=datetime.datetime.fromtimestamp(update[2], tz=datetime.UTC),
+                content=compare_to_html(extracted_old, extracted_new),
+                date=datetime.datetime.fromtimestamp(fetch_date, tz=datetime.UTC),
                 enclosures=[],
                 guid=update[2],
             )
-- 
cgit v1.2.3