[bcy] fix partial image URLs (#613)

Images from new posts can have incomplete/partial URLs (1)
without any filename extension when fetching their data from
'/apiv3/user/selfPosts', so now all data gets taken from
'/item/detail/ID' pages.

It is currently unknown how to get the non-watermarked original version
of these images, or if that is possible at all. (2)
Images with a watermark will have their 'filter' metadata field set to
"watermark". For original images this field is an empty string "".

Enabling the 'noop' option will, in addition to the watermarked version,
yield the the '~noop.image' filter version (3),
where 'filter' is set to "noop".

(1) "https://img-bcy-qn.pstatp.com/banciyuan/3ccdff22479c4060aadc86718209b281"
(2) "https://p1-bcy.byteimg.com/img/banciyuan/3ccdff22479c4060aadc86718209b281~tplv-banciyuan-logo-v3:wqnpnLLlhZLlpKfprZTnjotfCuWNiuasoeWFgyAtIEFDR-eIseWlveiAheekvuWMug==.image"
(3) "https://p1-bcy.byteimg.com/img/banciyuan/3ccdff22479c4060aadc86718209b281~noop.image"

gallery_dl/extractor/bcy.py

@@ -29,41 +29,72 @@ class BcyExtractor(Extractor):
     def items(self):
         sub = re.compile(r"^https?://p\d+-bcy\.byteimg\.com/img/banciyuan").sub
         iroot = "https://img-bcy-qn.pstatp.com"
+        noop = self.config("noop")
 
-        for post in self.posts():
+        for post_id in self.posts():
-            if not post["image_list"]:
+            post = self._parse_post(post_id)
+            if not post:
                 continue
 
-            data = {
+            yield Message.Directory, post
-                "user": {
+            for post["num"], image in enumerate(post["_multi"], 1):
-                    "id"     : post["uid"],
+                post["id"] = image["mid"]
-                    "name"   : post["uname"],
+                post["width"] = image["w"]
-                    "avatar" : sub(iroot, post["avatar"].partition("~")[0]),
+                post["height"] = image["h"]
-                },
+
-                "post": {
+                url = image["path"].partition("~")[0]
-                    "id"     : text.parse_int(post["item_id"]),
+                text.nameext_from_url(url, post)
-                    "tags"   : [t["tag_name"] for t in post["post_tags"]],
+
-                    "date"   : text.parse_timestamp(post["ctime"]),
+                if post["extension"]:
-                    "parody" : post["work"],
+                    if not url.startswith(iroot):
-                    "content": post["plain"],
+                        url = sub(iroot, url)
-                    "likes"  : post["like_count"],
+                    post["filter"] = ""
-                    "shares" : post["share_count"],
+                    yield Message.Url, url, post
-                    "replies": post["reply_count"],
+
-                },
+                else:
-            }
+                    post["filter"] = "watermark"
-
+                    yield Message.Url, image["origin"], post
-            yield Message.Directory, data
+
-            for data["num"], image in enumerate(post["image_list"], 1):
+                    if noop:
-                data["id"] = image["mid"]
+                        post["extension"] = ""
-                data["width"] = image["w"]
+                        post["filter"] = "noop"
-                data["height"] = image["h"]
+                        yield Message.Url, image["original_path"], post
-
+
-                url = image["path"]
+    def _parse_post(self, post_id):
-                if not url.startswith(iroot):
+        url = "{}/item/detail/{}".format(self.root, post_id)
-                    url = sub(iroot, url.partition("~")[0])
+        response = self.request(url)
-                data["url"] = url
+        if response.status_code >= 400:
-
+            return None
-                yield Message.Url, url, text.nameext_from_url(url, data)
+
+        data = json.loads(
+            text.extract(response.text, 'JSON.parse("', '");')[0]
+            .replace('\\\\u002F', '/')
+            .replace('\\"', '"')
+        )["detail"]
+
+        post = data["post_data"]
+        if not post["multi"]:
+            return None
+        user = data["detail_user"]
+
+        return {
+            "user": {
+                "id"     : user["uid"],
+                "name"   : user["uname"],
+                "avatar" : user["avatar"],
+            },
+            "post": {
+                "id"     : text.parse_int(post["item_id"]),
+                "tags"   : [t["tag_name"] for t in post["post_tags"]],
+                "date"   : text.parse_timestamp(post["ctime"]),
+                "parody" : text.parse_unicode_escapes(post["work"]),
+                "content": post["plain"],
+                "likes"  : post["like_count"],
+                "shares" : post["share_count"],
+                "replies": post["reply_count"],
+            },
+            "_multi": post["multi"],
+        }
 
 
 class BcyUserExtractor(BcyExtractor):
@@ -88,7 +119,8 @@ class BcyUserExtractor(BcyExtractor):
 
             item = None
             for item in data["data"]["items"]:
-                yield item["item_detail"]
+                if item["item_detail"]["multi"]:
+                    yield item["item_detail"]["item_id"]
 
             if not item:
                 return
@@ -106,7 +138,7 @@ class BcyPostExtractor(BcyExtractor):
             "user": {
                 "id"     : 1933712,
                 "name"   : "wukloo",
-                "avatar" : "re:https://img-bcy-qn.pstatp.com/Public/Upload/",
+                "avatar" : str,
             },
             "post": {
                 "id"     : 6355835481002893070,
@@ -128,17 +160,4 @@ class BcyPostExtractor(BcyExtractor):
     })
 
     def posts(self):
-        url = self.root + "/item/detail/" + self.item_id
+        return (self.item_id,)
-        page = self.request(url).text
-
-        data = json.loads(
-            text.extract(page, 'JSON.parse("', '");')[0]
-            .replace('\\\\u002F', '/')
-            .replace('\\"', '"')
-        )["detail"]
-
-        post = data["post_data"]
-        post["image_list"] = post["multi"]
-        post["plain"] = text.parse_unicode_escapes(post["plain"])
-        post.update(data["detail_user"])
-        return (post,)