[xenforo] fix/improve download URL handling

gallery_dl/extractor/xenforo.py

@@ -32,8 +32,8 @@ class XenforoExtractor(BaseExtractor):
         extract_urls = text.re(
             r'(?s)(?:'
             r'<video (.*?\ssrc="[^"]+".*?)</video>'
-            r'|<a [^>]*?href="[^"]*?'
-            r'(/(?:index\.php\?)?attachments/[^"]+".*?)</a>'
+            r'|<a [^>]*?'
+            r'href="([^"]*?/(?:index\.php\?)?attachments/[^"]+".*?)</a>'
             r'|<div class="bb(?:Image|Media)Wrapper[^>]*?'
             r'data-src="([^"]+".*?) />'
             r'|(?:<a [^>]*?href="|<iframe [^>]*?src="|'
@@ -41,7 +41,8 @@ class XenforoExtractor(BaseExtractor):
             r')'
         ).findall
 
-        root_media = self.config_instance("root-media") or self.root
+        root = self.root
+        base = root if (pos := root.find("/", 8)) < 0 else root[:pos]
         for post in self.posts():
             urls = extract_urls(post["content"])
             if post["attachments"]:
@@ -76,12 +77,12 @@ class XenforoExtractor(BaseExtractor):
                     data["id"] = text.parse_int(
                         data["filename"].partition("-")[0])
                     if url[0] == "/":
-                        url = root_media + url
+                        url = base + url
                     yield Message.Url, url, data
 
                 elif (inline := bb or inl):
-                    path = inline[:inline.find('"')]
-                    name, _, id = path[path.rfind("/", 0, -1):].strip(
+                    url = inline[:inline.find('"')]
+                    name, _, id = url[url.rfind("/", 0, -1):].strip(
                         "/").rpartition(".")
                     data["id"] = id = text.parse_int(id)
                     if id:
@@ -100,13 +101,11 @@ class XenforoExtractor(BaseExtractor):
                     data["num"] += 1
                     data["num_internal"] += 1
                     data["type"] = "inline"
-
-                    url = self.root + path if path[0] == "/" else path
+                    if url[0] == "/":
+                        url = base + url
                     yield Message.Url, url, data
 
     def items_media(self, path, pnum):
-        self.root_media = self.config_instance("root-media") or self.root
-
         if (order := self.config("order-posts")) and \
                 order[0] in ("d", "r"):
             pages = self._pagination_reverse(path, pnum)
@@ -115,13 +114,15 @@ class XenforoExtractor(BaseExtractor):
             pages = self._pagination(path, pnum)
             reverse = False
 
-        if meta := self.config("metadata"):
+        if self.config("metadata"):
             extr_media = self._extract_media_ex
             meta = True
         else:
             extr_media = self._extract_media
             meta = False
 
+        root = self.root
+        base = root if (pos := root.find("/", 8)) < 0 else root[:pos]
         for page in pages:
             posts = page.split(
                 '<div class="itemList-item js-inlineModContainer')
@@ -134,8 +135,8 @@ class XenforoExtractor(BaseExtractor):
                 href, pos = text.extract(html, 'href="', '"')
                 name, pos = text.extract(html, "alt='", "'", pos)
 
-                href = href[:-1]
-                url, media = extr_media(href, href.rpartition("/")[2])
+                url, media = extr_media(
+                    base + href, href[href.rfind("/", 0, -1)+1:-1])
                 if not meta and name:
                     text.nameext_from_name(text.unescape(name), media)
 
@@ -294,14 +295,14 @@ class XenforoExtractor(BaseExtractor):
 
         return post
 
-    def _extract_media(self, path, file):
+    def _extract_media(self, url, file):
         media = {}
         name, _, media["id"] = file.rpartition(".")
         media["filename"], _, media["extension"] = name.rpartition("-")
-        return f"{self.root_media}{path}/full", media
+        return url + "full", media
 
-    def _extract_media_ex(self, path, file):
-        page = self.request(f"{self.root_media}{path}/").text
+    def _extract_media_ex(self, url, file):
+        page = self.request(url).text
 
         schema = self._extract_jsonld(page)
         main = schema["mainEntity"]
@@ -359,7 +360,6 @@ BASE_PATTERN = XenforoExtractor.update({
     },
     "atfforum": {
         "root": "https://allthefallen.moe/forum",
-        "root-media": "https://allthefallen.moe",
         "pattern": r"(?:www\.)?allthefallen\.moe/forum",
         "cookies": ("xf_user",),
     },
@@ -485,8 +485,8 @@ class XenforoMediaItemExtractor(XenforoExtractor):
     example = "https://simpcity.cr/media/NAME.123/"
 
     def items(self):
-        self.root_media = self.root
+        url = f"{self.root}{self.groups[-2]}/"
         url, media = (self._extract_media_ex if self.config("metadata") else
-                      self._extract_media)(self.groups[-2], self.groups[-1])
+                      self._extract_media)(url, self.groups[-1])
         yield Message.Directory, "", media
         yield Message.Url, url, media