diff --git a/hosts/tsuki/services/scrapers/nhk-easy-news/default.nix b/hosts/tsuki/services/scrapers/nhk-easy-news/default.nix
index 2d2478b..a784928 100644
--- a/hosts/tsuki/services/scrapers/nhk-easy-news/default.nix
+++ b/hosts/tsuki/services/scrapers/nhk-easy-news/default.nix
@@ -6,7 +6,7 @@
 in {
   systemd.services.scrape-nhk-easy-news = {
     after = [ "network.target" ];
-    serviceConfig = {
+    serviceConfig = rec {
       Type = "oneshot";
       ExecStart = script;
       DynamicUser = true;
@@ -15,6 +15,10 @@ in {
       ProtectProc = "invisible";
       ProtectSystem = "strict";
       WorkingDirectory = "/data/scrapers/nhk-easy-news";
+      BindPaths = [ WorkingDirectory ];
+      ReadWritePaths = [ WorkingDirectory ];
+      StateDirectory = "nhk-easy-news-scraper";
+      StateDirectoryMode = "0755";
     };
   };
 
diff --git a/hosts/tsuki/services/scrapers/nhk-easy-news/script.py b/hosts/tsuki/services/scrapers/nhk-easy-news/script.py
index ce7b2fc..ce4d108 100644
--- a/hosts/tsuki/services/scrapers/nhk-easy-news/script.py
+++ b/hosts/tsuki/services/scrapers/nhk-easy-news/script.py
@@ -1,38 +1,104 @@
 from pathlib import Path
+from urllib.parse import urlparse
+from textwrap import dedent, indent
 import os
+import json
 
 import requests
 import wget
 
 
+BASE_URL = "https://www3.nhk.or.jp/news/easy"
+
+
+def url_filename(url) -> str:
+    parsed_url = urlparse(url)
+    path = parsed_url.path
+    return Path(path).name
+
+
+def try_download(url, path) -> str | None:
+    try:
+        wget.download(url, out=str(path))
+    except Exception as err:
+        if path.exists():
+            os.remove(path)
+        return err
+
+
+def download_article(article, dir_path):
+    news_id = article['news_id']
+    article_path = dir_path / news_id
+    if not article_path.exists():
+        print(f"New article with ID: {news_id}")
+        os.mkdir(article_path)
+
+    index_path = article_path / 'index.html'
+    if not index_path.exists():
+        print("  Downloading article")
+        err = try_download(article['news_web_url'], index_path)
+        if err is not None:
+            print(dedent(f'''
+            Failed to download article {news_id}:
+              {article['news_web_url']}
+            {indent(str(err), '  ')}
+            '''))
+            return
+
+    info_path = article_path / 'info.json'
+    if not info_path.exists():
+        print("  Exporting metadata")
+        with open(info_path, 'w') as file:
+            json.dump(article, file)
+
+    for toggle, url_attr in (
+        ('has_news_web_image', 'news_web_image_uri'),
+        # ('has_news_web_movie', 'news_web_movie_uri'),
+        # ('has_news_easy_image', 'news_easy_image_uri'),
+        # ('has_news_easy_movie', 'news_easy_movie_uri'),
+        # ('has_news_easy_voice', 'news_easy_voice_uri'),
+    ):
+        if not article[toggle]:
+            continue
+
+        url = article[url_attr]
+        # if not url.startswith('http'):
+        #     url = BASE_URL + '/' + url
+
+        path = article_path / url_filename(url)
+        if path.exists():
+            continue
+
+        print(f'  Downloading supplementary material: {url_filename(url)}')
+
+        err = try_download(url, path)
+        if err is not None:
+            print(dedent(f'''
+            Failed to download supplementary material for article {news_id}:
+              {url}
+            {indent(str(err), '  ')}
+            '''))
+
+
 def main():
-    nhkjson = requests.get(
-        'http://www3.nhk.or.jp/news/easy/news-list.json').json()
+    print("Starting nhk easy news scraper")
+    print()
+    print("Fetching article index...")
+    nhkjson = requests.get(BASE_URL + '/news-list.json').json()
     base_dir = Path(".").resolve()
+    print('Got article index')
 
     if not (base_dir / 'articles').exists():
         os.mkdir(base_dir / 'articles')
 
-    for key, value in nhkjson[0].items():
-        for x in value:
-            news_id = x['news_id']
-            path = base_dir / f'articles/nhkeasy_{news_id}.html'
+    for date, articlelist in nhkjson[0].items():
+        date_dir = base_dir / 'articles' / date
+        if not date_dir.exists():
+            print(f"Found new articles for {date}")
+            os.mkdir(date_dir)
 
-            if path.exists():
-                # This means that the article has already been downloaded.
-                # Skip and continue
-                continue
-
-            print(f"New article with ID: {news_id}")
-            try:
-                nhkurl = x['news_web_url']
-                wget.download(nhkurl, out=str(path))
-                print("Successful download of article ID: " + x['news_id'])
-            except Exception as err:
-                if path.exists():
-                    os.remove(path)
-                print("Failed to download article ID: " + x['news_id'])
-                print(err)
+        for article in articlelist:
+            download_article(article, date_dir)
 
 
 if __name__ == '__main__':