增加ua解析记录

Author: liangliangyy

blog/documents.py

@@ -51,13 +51,35 @@ class GeoIp(InnerDoc):
     location = GeoPoint()
 
 
+class UserAgentBrowser(InnerDoc):
+    Family = Keyword()
+    Version = Keyword()
+
+
+class UserAgentOS(UserAgentBrowser):
+    pass
+
+
+class UserAgentDevice(InnerDoc):
+    Family = Keyword()
+    Brand = Keyword()
+    Model = Keyword()
+
+
+class UserAgent(InnerDoc):
+    browser = Object(UserAgentBrowser, required=False)
+    os = Object(UserAgentOS, required=False)
+    device = Object(UserAgentDevice, required=False)
+    string = Text()
+
+
 class ElapsedTimeDocument(Document):
     url = Keyword()
     time_taken = Long()
     log_datetime = Date()
-    useragent = Text(analyzer='ik_max_word', search_analyzer='ik_smart')
     ip = Keyword()
     geoip = Object(GeoIp, required=False)
+    useragent = Object(UserAgent, required=False)
 
     class Index:
         name = 'performance'
@@ -80,11 +102,25 @@ def delete_index():
     @staticmethod
     def create(url, time_taken, log_datetime, useragent, ip):
         from elasticsearch import Elasticsearch
-        es = Elasticsearch(settings.ELASTICSEARCH_DSL['default']['hosts'])
+        client = Elasticsearch(settings.ELASTICSEARCH_DSL['default']['hosts'])
 
-        res = es.indices.exists(index="performance")
+        res = client.indices.exists(index="performance")
         if not res:
             ElapsedTimeDocument.init()
+        ua = UserAgent()
+        ua.browser = UserAgentBrowser()
+        ua.browser.Family = useragent.browser.family
+        ua.browser.Version = useragent.browser.version_string
+
+        ua.browser = UserAgentOS()
+        ua.browser.Family = useragent.os.family
+        ua.browser.Version = useragent.os.version_string
+
+        ua.device = UserAgentDevice()
+        ua.device.Family = useragent.device.family
+        ua.device.Brand = useragent.device.brand
+        ua.device.Model = useragent.device.model
+        ua.string = useragent.ua_string
 
         doc = ElapsedTimeDocument(
             meta={
@@ -96,7 +132,7 @@ def create(url, time_taken, log_datetime, useragent, ip):
             url=url,
             time_taken=time_taken,
             log_datetime=log_datetime,
-            useragent=useragent, ip=ip)
+            useragent=ua, ip=ip)
         doc.save(pipeline="geoip")
 
 

blog/middleware.py

@@ -16,6 +16,7 @@
 import time
 import logging
 from ipware import get_client_ip
+from user_agents import parse
 from blog.documents import ELASTICSEARCH_ENABLED, ElaspedTimeDocumentManager
 
 logger = logging.getLogger(__name__)
@@ -32,22 +33,23 @@ def __call__(self, request):
         response = self.get_response(request)
         http_user_agent = request.META.get('HTTP_USER_AGENT', '')
         ip, _ = get_client_ip(request)
-        if 'spider'.upper() not in http_user_agent.upper():
-            try:
-                cast_time = time.time() - start_time
-                if ELASTICSEARCH_ENABLED:
-                    time_taken = round((cast_time) * 1000, 2)
-                    url = request.path
-                    from django.utils import timezone
-
-                    ElaspedTimeDocumentManager.create(
-                        url=url,
-                        time_taken=time_taken,
-                        log_datetime=timezone.now(),
-                        useragent=http_user_agent,
-                        ip=ip)
-                response.content = response.content.replace(
-                    b'<!!LOAD_TIMES!!>', str.encode(str(cast_time)[:5]))
-            except Exception as e:
-                logger.error("Error OnlineMiddleware: %s" % e)
+        user_agent = parse(http_user_agent)
+
+        try:
+            cast_time = time.time() - start_time
+            if ELASTICSEARCH_ENABLED:
+                time_taken = round((cast_time) * 1000, 2)
+                url = request.path
+                from django.utils import timezone
+                ElaspedTimeDocumentManager.create(
+                    url=url,
+                    time_taken=time_taken,
+                    log_datetime=timezone.now(),
+                    useragent=user_agent,
+                    ip=ip)
+            response.content = response.content.replace(
+                b'<!!LOAD_TIMES!!>', str.encode(str(cast_time)[:5]))
+        except Exception as e:
+            logger.error("Error OnlineMiddleware: %s" % e)
+
         return response

blog/views.py

@@ -189,7 +189,8 @@ class AuthorDetailView(ArticleListView):
     page_type = '作者文章归档'
 
     def get_queryset_cache_key(self):
-        author_name = self.kwargs['author_name']
+        from uuslug import slugify
+        author_name = slugify(self.kwargs['author_name'])
         cache_key = 'author_{author_name}_{page}'.format(
             author_name=author_name, page=self.page_number)
         return cache_key

docs/docker.md

@@ -7,7 +7,8 @@
 ## docker镜像方式
 本项目已经支持了docker部署，如果你已经有了`mysql`，那么直接使用基础镜像即可，启动命令如下所示：
 ```shell
- docker run -d  -p 8000:8000 -e DJANGO_MYSQL_HOST=mysqlhost -e DJANGO_MYSQL_PASSWORD=mysqlrootpassword -e DJANGO_MYSQL_USER=root -e DJANGO_MYSQL_DATABASE=djangoblog --name djangoblog liangliangyy/djangoblog
+docker pull liangliangyy/djangoblog:latest
+docker run -d  -p 8000:8000 -e DJANGO_MYSQL_HOST=mysqlhost -e DJANGO_MYSQL_PASSWORD=mysqlrootpassword -e DJANGO_MYSQL_USER=root -e DJANGO_MYSQL_DATABASE=djangoblog --name djangoblog liangliangyy/djangoblog:latest
 ```
 启动完成后，访问 http://127.0.0.1:8000 
 ## 使用docker-compose

requirements.txt

@@ -20,8 +20,9 @@ python-slugify==5.0.2
 pytz==2021.1
 raven==6.10.0
 rcssmin==1.0.6
-requests==2.25.1
+requests==2.26.0
 rjsmin==1.1.0
 urllib3==1.26.6
 WeRoBot==1.13.1
-Whoosh==2.7.4
+Whoosh==2.7.4
+user-agents==2.2.0