From 3d8d3efd25b6a0ba14cf1794da92f2c40b13a040 Mon Sep 17 00:00:00 2001
From: Thomas Sileo <t@a4.io>
Date: Sun, 5 Aug 2018 14:45:44 +0200
Subject: [PATCH] More Open Graph tweaks

---
 utils/opengraph.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/utils/opengraph.py b/utils/opengraph.py
index 5401131..318b1bc 100644
--- a/utils/opengraph.py
+++ b/utils/opengraph.py
@@ -44,12 +44,15 @@ def fetch_og_metadata(user_agent, links):
 
         r = requests.get(l, headers={"User-Agent": user_agent}, timeout=15)
         r.raise_for_status()
+        if not r.headers.get("content-type").startswith("text/html"):
+            logger.debug(f"skipping {l}")
+            continue
 
         html = r.text
         try:
             data = dict(opengraph.OpenGraph(html=html))
         except Exception:
-            logger.exception("failed to parse {l}")
+            logger.exception(f"failed to parse {l}")
             continue
         if data.get("url"):
             res.append(data)