Merge pull request #425 from openzim/double_slash_main

Stop checking main entry processability when it is already found
openzim · Nov 27, 2024 · 5040eee · 5040eee
2 parents 8733200 + 1e6367c
commit 5040eee
Show file tree

Hide file tree

Showing 6 changed files with 45 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -12,6 +12,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Upgrade to wombat 3.8.6 (#334)
 - Fix wombat setup settings (especially `isSW`) (#293)
 
+### Fixed
+
+- Stop checking main entry processability when it is already found (#424)
+
 ## [2.1.3] - 2024-11-01
 
 ### Changed

diff --git a/src/warc2zim/converter.py b/src/warc2zim/converter.py
@@ -478,7 +478,11 @@ def gather_information_from_warc(self):
 
             status_code = get_status_code(record)
             if not can_process_status_code(status_code):
-                if record.rec_type == "response" and self.main_path == zim_path:
+                if (
+                    not main_page_found
+                    and record.rec_type == "response"
+                    and self.main_path == zim_path
+                ):
                     raise UnprocessableWarcError(
                         f"Main URL returned an unprocessable HTTP code: {status_code}"
                     )

diff --git a/test-website/Caddyfile b/test-website/Caddyfile
@@ -73,6 +73,14 @@
 
     respond /502-response 502
 
+    respond // "Hello you" 400
+
+    respond /double-slash/test1 "Hello you" 200
+    respond /double-slash//test1 400
+
+    respond /double-slash/test2 "Hello you v1" 200
+    respond /double-slash//test2 "Hello you v2" 200
+
     redir /301-internal-redirect-ok /internal_redirect_target.html 301
     redir /301-external-redirect-ok https://www.example.com 301
     redir /302-internal-redirect-ok /internal_redirect_target.html 302

diff --git a/test-website/Dockerfile b/test-website/Dockerfile
@@ -1,5 +1,5 @@
 FROM caddy:2.6.1-alpine
-LABEL org.opencontainers.image.source https://github.com/openzim/warc2zim
+LABEL org.opencontainers.image.source=https://github.com/openzim/warc2zim
 
 COPY Caddyfile /etc/caddy/Caddyfile
 

diff --git a/test-website/content/double-slash.html b/test-website/content/double-slash.html
@@ -0,0 +1,26 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+  <meta charset="utf-8">
+  <title>Test website</title>
+  <link rel="apple-touch-icon" sizes="180x180" href="./icons/apple-touch-icon.png">
+  <link rel="icon" type="image/png" sizes="32x32" href="./icons/favicon-32x32.png">
+  <link rel="icon" type="image/png" sizes="16x16" href="./icons/favicon-16x16.png">
+  <link rel="manifest" href="./icons/site.webmanifest">
+  <link rel="shortcut icon" href="./icons/favicon.ico">
+</head>
+
+<body>
+
+  <h2>Double slash in URLs</h2>
+
+  <a href=".//">.//</a>
+  <a href="./double-slash//test1">./double-slash//test1</a>
+  <a href="./double-slash/test1">./double-slash/test1</a>
+  <a href="./double-slash/test2">./double-slash/test2</a>
+  <a href="./double-slash//test2">./double-slash//test2</a>
+
+</body>
+
+</html>
diff --git a/test-website/content/index.html b/test-website/content/index.html
@@ -52,6 +52,7 @@
         <li><a href="./http-equiv-redirect.html">Redirect with http-equiv meta directive</a></li>
         <li><a href="./image-srcset.html">Image with srcset</a></li>
         <li><a href="./form-get.html">Form GET</a></li>
+        <li><a href="./double-slash.html">Double Slash</a></li>
     </ul>
 </body>