diff --git a/gallery_dl/extractor/pururin.py b/gallery_dl/extractor/pururin.py
index 32567f63..e56dbcd6 100644
--- a/gallery_dl/extractor/pururin.py
+++ b/gallery_dl/extractor/pururin.py
@@ -10,7 +10,6 @@
 
 from .common import GalleryExtractor
 from .. import text, util
-import binascii
 
 
 class PururinGalleryExtractor(GalleryExtractor):
@@ -19,12 +18,11 @@ class PururinGalleryExtractor(GalleryExtractor):
     pattern = r"(?:https?://)?(?:www\.)?pururin\.[ti]o/(?:gallery|read)/(\d+)"
     test = (
         ("https://pururin.to/gallery/38661/iowant-2", {
-            "pattern": r"https://cdn.pururin.[ti]o/\w+"
-                       r"/images/data/\d+/\d+\.jpg",
+            "pattern": r"https://i\.pururin\.to/38661/\d+\.jpg",
             "keyword": {
                 "title"     : "re:I ?owant 2!!",
                 "title_en"  : "re:I ?owant 2!!",
-                "title_jp"  : "",
+                "title_ja"  : "",
                 "gallery_id": 38661,
                 "count"     : 19,
                 "artist"    : ["Shoda Norihiro"],
@@ -33,11 +31,11 @@ class PururinGalleryExtractor(GalleryExtractor):
                 "characters": ["Iowa", "Teitoku"],
                 "tags"      : list,
                 "type"      : "Doujinshi",
-                "collection": "I owant you!",
-                "convention": "C92",
+                "collection": ["I owant you!"],
+                "convention": ["C92"],
                 "rating"    : float,
                 "uploader"  : "demo",
-                "scanlator" : "mrwayne",
+                "scanlator" : ["mrwayne", "The Lost Light"],
                 "lang"      : "en",
                 "language"  : "English",
             }
@@ -54,51 +52,49 @@ class PururinGalleryExtractor(GalleryExtractor):
         url = "{}/gallery/{}/x".format(self.root, self.gallery_id)
         GalleryExtractor.__init__(self, match, url)
 
-        self._ext = ""
-        self._cnt = 0
-
     def metadata(self, page):
         extr = text.extract_from(page)
 
-        def _lst(key, e=extr):
-            return [
-                text.unescape(item)
-                for item in text.extract_iter(e(key, "</td>"), 'title="', '"')
-            ]
+        def _lst(e=extr):
+            v = text.unescape(e('value="', '"'))
+            return [item["name"] for item in util.json_loads(v)] if v else ()
 
         def _str(key, e=extr):
-            return text.unescape(text.extract(
-                e(key, "</td>"), 'title="', '"')[0] or "")
+            return text.unescape(text.extr(
+                e(key, "</td>"), 'title="', '"')).partition(" / ")[0]
 
-        url = "{}/read/{}/01/x".format(self.root, self.gallery_id)
-        page = self.request(url).text
-        info = util.json_loads(binascii.a2b_base64(text.extr(
-            page, '<gallery-read encoded="', '"')).decode())
-        self._ext = info["image_extension"]
-        self._cnt = info["total_pages"]
+        title = text.unescape(extr('<h1><span itemprop="name">', '<'))
+        title_en, _, title_ja = title.partition(" / ")
 
         data = {
             "gallery_id": text.parse_int(self.gallery_id),
-            "title"     : info["title"] or info.get("j_title") or "",
-            "title_en"  : info["title"],
-            "title_jp"  : info.get("j_title") or "",
-            "artist"    : _lst("<td>Artist</td>"),
-            "group"     : _lst("<td>Circle</td>"),
-            "parody"    : _lst("<td>Parody</td>"),
-            "tags"      : _lst("<td>Contents</td>"),
-            "type"      : _str("<td>Category</td>"),
-            "characters": _lst("<td>Character</td>"),
-            "collection": _str("<td>Collection</td>"),
+            "title"     : title_en or title_ja,
+            "title_en"  : title_en,
+            "title_ja"  : title_ja,
             "language"  : _str("<td>Language</td>"),
-            "scanlator" : _str("<td>Scanlator</td>"),
-            "convention": _str("<td>Convention</td>"),
+            "type"      : _str("<td>Category</td>"),
             "uploader"  : text.remove_html(extr("<td>Uploader</td>", "</td>")),
-            "rating"    : text.parse_float(extr(" :rating='"       , "'")),
+            "rating"    : text.parse_float(extr(
+                'itemprop="ratingValue" content="', '"')),
+            "artist"    : extr('name="artist_tags"', '') or _lst(),
+            "group"     : _lst(),
+            "parody"    : _lst(),
+            "tags"      : _lst(),
+            "characters": _lst(),
+            "scanlator" : _lst(),
+            "convention": _lst(),
+            "collection": _lst(),
         }
         data["lang"] = util.language_to_code(data["language"])
         return data
 
     def images(self, _):
-        ufmt = "https://cdn.pururin.to/assets/images/data/{}/{{}}.{}".format(
-            self.gallery_id, self._ext)
-        return [(ufmt.format(num), None) for num in range(1, self._cnt + 1)]
+        url = "{}/read/{}/01/x".format(self.root, self.gallery_id)
+        page = self.request(url).text
+
+        svr, pos = text.extract(page, 'data-svr="', '"')
+        img, pos = text.extract(page, 'data-img="', '"', pos)
+        data = util.json_loads(text.unescape(img))
+
+        base = "{}/{}/".format(svr, data["directory"])
+        return [(base + i["filename"], None) for i in data["images"]]