Text Content Caching for 1 hr to reduce load and response times

2 files changed, 68 insertions, 18 deletions

diff --git a/middleware/contentCachingMiddleware.py b/middleware/contentCachingMiddleware.py
new file mode 100644
index 00000000..d5f23f88
--- /dev/null
+++ b/middleware/contentCachingMiddleware.py
@@ -0,0 +1,43 @@
+from django.http import HttpResponse
+import redis
+
+
+r = redis.Redis(host='localhost', port=6379, db=0)
+
+class ContentCachingMiddleware(object):
+    # We will cache all text/html responses for 1 hour
+    CACHE_TTL = 60 * 60
+
+    def __init__(self, get_response):
+        self.get_response = get_response
+
+    def __call__(self, request):
+        lang_cookie = request.COOKIES.get('lang', '')
+
+        # Don't cache POST requests
+        if request.method == 'POST' or lang_cookie == 'ja':
+            return self.get_response(request)
+
+        # Try to get cached response
+        path = request.get_full_path()
+        cache_key = f'{lang_cookie}:{path}'
+
+        # clear cache if clear_cache is in query params
+        if request.GET.get('cc', '') == 'true':
+            r.delete(cache_key)
+
+        cached_response = r.get(cache_key)
+        if cached_response:
+            print('cached', cache_key)
+            return HttpResponse(cached_response)
+        
+        # Get response from view
+        response = self.get_response(request)
+
+        # Cache response if content-type is text/html
+        if response.status_code == 200 and 'text/html' in response.get('Content-Type', ''):
+            print(cache_key, 'cached')
+            r.set(cache_key, response.content.decode('utf-8'))
+            r.expire(cache_key, self.CACHE_TTL)
+
+        return response
\ No newline at end of file
diff --git a/middleware/translationMiddleware.py b/middleware/translationMiddleware.py
index ba1eda1f..4687984d 100644
--- a/middleware/translationMiddleware.py
+++ b/middleware/translationMiddleware.py
@@ -1,8 +1,11 @@
 import os
+from django.http import HttpResponse
+import redis
+
+
+r = redis.Redis(host='localhost', port=6379, db=0)
 
-from bs4 import BeautifulSoup
 from django.conf import settings
-from django.core.cache import cache
 from google.cloud import translate_v2 as translate
 
 cred_path = os.path.join(settings.BASE_DIR, 'credentials-translate.json')
@@ -11,36 +14,40 @@ os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = cred_path
 class TranslationMiddleware:
     def __init__(self, get_response):
         self.get_response = get_response
+        self.CACHE_TTL = 60 * 60
 
     def __call__(self, request):
-        response = self.get_response(request)
-
         # translate only if lang cookie is set to ja
-        if request.COOKIES.get('lang') != 'ja':
+        response = self.get_response(request)
+        lang_cookie = request.COOKIES.get('lang', '')
+        
+        if lang_cookie != 'ja':
             return response
 
         content_type = response.get('Content-Type', '').lower()
         if 'text' not in content_type:
             return response
         
-        HTML_content =response.content.decode('utf-8').strip()
+        # Try to get cached response
+        path = request.get_full_path()
+        cache_key = cache_key = f'{lang_cookie}:{path}'
+        print(f'Japanese translation for {path} requested')
+
+        cache_response = r.get(cache_key)
+        if cache_response:
+            print('cached', cache_key)
+            # r.delete(cache_key)
+            response = HttpResponse(cache_response)
+            return response
 
-        # add no translate class to the 'highlight' class
-        soup = BeautifulSoup(HTML_content, 'html.parser')
-        for tag in soup.find_all(class_='highlight'):
-            tag['class'].append('notranslate')
-        HTML_content = str(soup)
+        print('Cache not found, translating... for key: ', cache_key)
+        HTML_content =response.content.decode('utf-8')
         
         HTML_content = HTML_content.replace(
             "That Computer Scientist",
             "ザットコンピュータサイエンティスト"
         )
 
-        # check if the content is already in the cache
-        if cache.get(HTML_content):
-            response.content = cache.get(HTML_content)
-            return response
-
         translate_client = translate.Client()
         target_language = 'ja'
         translated_content = translate_client.translate(
@@ -48,8 +55,8 @@ class TranslationMiddleware:
             target_language=target_language,
         )['translatedText']
 
-        # set the content in the cache for 7 days
-        cache.set(HTML_content, translated_content, 60 * 60 * 24 * 7)
+        r.set(cache_key, translated_content)
+        r.expire(cache_key, self.CACHE_TTL)
 
         response.content = translated_content.encode('utf-8')
         return response