Fix #87: do not process hrefs pointing outside iFixit

openzim · Oct 20, 2022 · 0f84bae · 0f84bae
1 parent 996489a
commit 0f84bae
Show file tree

Hide file tree

Showing 3 changed files with 5 additions and 21 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,3 +1,7 @@
+# 0.2.3
+
+- Do not process unrecognized href, i.e. pointing outside iFixit
+
 # 0.2.2
 
 - Fixed URL normalization on articles redirecting outside domain (help.ifixit.com)

diff --git a/ifixit2zim/VERSION b/ifixit2zim/VERSION
@@ -1 +1 @@
-0.2.1
+0.2.3
diff --git a/ifixit2zim/shared.py b/ifixit2zim/shared.py
@@ -241,26 +241,6 @@ def _process_external_url(url, rel_prefix):
 
     @staticmethod
     def _process_unrecognized_href(url, rel_prefix):
-        if not url.startswith("https://") and not url.startswith("http://"):
-            return Global._process_external_url(url, rel_prefix)
-        try:
-            resp = requests.head(url, timeout=5)
-            headers = resp.headers
-        except requests.exceptions.ConnectionError:
-            logger.debug(f"Unable to HEAD unrecognized href (ConnectionError): {url}")
-            return Global._process_external_url(url, rel_prefix)
-        except requests.exceptions.ReadTimeout:
-            logger.debug(f"Unable to HEAD unrecognized href (ReadTimeout): {url}")
-            return Global._process_external_url(url, rel_prefix)
-        except Exception as exc:
-            logger.warning(f"Unable to HEAD unrecognized href: {url}")
-            logger.exception(exc)
-            return Global._process_external_url(url, rel_prefix)
-
-        contentType = headers.get("Content-Type")
-        if contentType and contentType.startswith("image/"):
-            return f"{rel_prefix}{Global.get_image_path(url)}"
-
         return Global._process_external_url(url, rel_prefix)
 
     def _process_href_regex_dynamics(href, rel_prefix):