fixing bad url handling

2 files changed, 21 insertions, 12 deletions

diff --git a/core/views/caves.py b/core/views/caves.py
index d469a9d..75971bb 100644
--- a/core/views/caves.py
+++ b/core/views/caves.py
@@ -413,6 +413,9 @@ def cavepage(request, karea=None, subpath=None):
         cave = caves[0]
         return rendercave(request, cave, cave.slug())
 
+    if len(subpath) > 100: # overlong URL
+           return redirect(f"/caves")
+
 
     # HORRIBLE HACK, to be removed..
     subpath = subpath.strip("//")
@@ -420,6 +423,7 @@ def cavepage(request, karea=None, subpath=None):
     parts = subpath.strip("/").split("/")
     if len(parts) > 5:
         # recursive loop. break out of it.
+        print(karea,subpath)
         subparts = parts[0].split(".")
         caveid = subparts[0] 
         slug = f"{karea}-{caveid}"
@@ -427,34 +431,39 @@ def cavepage(request, karea=None, subpath=None):
             return redirect(f"/{cave.url}")
         else:
             return redirect(f"/caves")
-
-
+           
     # epath = karea  + subpath # e.g. 1623  /204
     # return expo.expopage(request, epath)
     
-    # BUGGER the real problem is the the cave descript has embedded in it images like
+    
+    # BUGGER the real problem is the the cave description has embedded in it images like
     # src="110/entrance.jpeg and since the cave url is now /1623/110/110.html
     # the images try to load from /1623/110/110/entrance.jpeg and of course fail.
     # THIS IS A HORRIBLE HACK
     if len(parts) == 1: 
-        # simple filename, no folders in path, 
-        # either need to insert caveid OR leave as relative link as we are already "in" /1623/nn/
+         # either need to insert caveid OR leave as relative link as we are already "in" /1623/nn/
+        # if we are doing a cave description file
+        print("SIMPLE", subpath, parts)
         subparts = parts[0].split(".")
-        caveid = subparts[0] # e.g. 204.htm
+        caveid = subparts[0] # e.g. 204.htm => "204"
         k2path = karea +"/"+ caveid + subpath
-        return redirect(f"/{k2path}") # infinite loop
+        return redirect(f"/{k2path}") # potential infinite loop
     elif len(parts) >2:
         # e.g. i/204.jpg, but that's ok as we are already "in" /1623/nn/
-        if parts[0] == parts[1]: # double caveid
+        if parts[0] == parts[1]: # double caveid, e.g. /161/161/france.html
             epath = karea 
             for i in parts[1:]:
                 epath += "/" + i               
-            #print(f"{subpath=}\n  {epath=}")
+            # print(f"{subpath=}\n  {epath=}")
             return expo.expopage(request, epath)
 
     # if either the first two parts are not /caveid/caveid/
     # or the number of parts == 2, 
-    # print(f"2 {subpath=}")
+    # This is the bit that MOST images in descriptions get to.
+    # simple filename, no folders in path, e.g. href="94_ent_close.jpg"
+    # inside a cave or entrance description, this "just works" as the href is just 
+    # added to the context of the current page, so there are >> 1 "parts" to the URL.
+    print(f"2 {subpath=}")
     epath = karea + "/" + subpath
     return expo.expopage(request, epath)
 
diff --git a/core/views/expo.py b/core/views/expo.py
index ac5c522..d492820 100644
--- a/core/views/expo.py
+++ b/core/views/expo.py
@@ -171,9 +171,9 @@ def expowebpage(request, expowebpath, path):
     if path == "" or path == "index.htm":
         return indexpage(request)
         
-    if not os.path.isfile(expowebpath / path):
+    if not Path(expowebpath / path).is_file():
         # Should not get here if the path has suffix "_edit"
-        print(f" - 404 error in expowebpage() {path}")
+        print(f" - 404 error in expowebpage() {Path(expowebpath / path)}")
         return render(request, "pagenotfound.html", {"path": path}, status=404)
 
     # print(f' - {sys_getfilesystemencoding()=}')