commit 07005ec3c6588f674620988c14030e5b0c87bee5
Author: Bùi Thành Nhân <hi@imnhan.com>
Date:   Mon Jan 4 09:04:21 2021 +0700

    check manga image url from db instead of regex

diff --git a/pyproject.toml b/pyproject.toml
index 05d43d6..0c7b27f 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pytaku"
-version = "0.3.29"
+version = "0.3.30"
 description = "Self-hostable web-based manga reader"
 authors = ["Bùi Thành Nhân <hi@imnhan.com>"]
 license = "AGPL-3.0-only"
diff --git a/src/pytaku/main.py b/src/pytaku/main.py
index dd50161..51c7238 100644
--- a/src/pytaku/main.py
+++ b/src/pytaku/main.py
@@ -19,6 +19,7 @@
     get_prev_next_chapters,
     get_username,
     import_follows,
+    is_manga_page_url,
     load_chapter,
     load_title,
     read,
@@ -73,15 +74,14 @@ def _decode_proxy_url(b64_url):
 
 def _is_manga_img_url(
     url,
-    pattern=re.compile(
-        r"^https://("
-        r"([\w-]+\.)?(mangadex\.org/(data|images)|(mangabeast\d{0,4}.com|hot\.granpulse\.us|fan-official\.eorzea\.us|fan-complete\.ivalice\.us)/manga)/"
-        r"|"
-        r"([\w-]+\.)+mangadex\.network:\d{2,6}/[\w-]+/data/"
-        r")"
-    ),
+    cover_pattern=re.compile(r"^https://([\w-]+\.)?mangadex\.org/images"),
 ):
-    return pattern.match(url)
+    """
+    Check if either a cover or page img url.
+    """
+    # Mangasee has god knows how many domains for manga page images, so we can't just
+    # do a regex check, but need to query our db instead.
+    return cover_pattern.match(url) or is_manga_page_url(url)
 
 
 def proxied(url) -> str:
diff --git a/src/pytaku/persistence.py b/src/pytaku/persistence.py
index bb2a9af..fa78bd9 100644
--- a/src/pytaku/persistence.py
+++ b/src/pytaku/persistence.py
@@ -390,3 +390,17 @@ def delete_expired_tokens():
         return_num_affected=True,
     )
     return num_deleted
+
+
+def is_manga_page_url(url):
+    """
+    Checks if url exists in db as a page image.
+    This is currently used to avoid abuse of our /proxy/ endpoint.
+    """
+    result = run_sql(
+        """
+        SELECT 1 FROM chapter, json_each(pages) WHERE value = ? LIMIT 1;
+        """,
+        (url,),
+    )
+    return len(result) == 1