Fix gunicorn related typo

To only load the code once (and start autotrim once)

Dockerfile

@@ -5,4 +5,4 @@ RUN apk add python3 py3-lxml py3-gunicorn py3-pip git
 ADD . /app
 RUN pip3 install /app
 
-CMD gunicorn --bind 0.0.0.0:8080 -w 4 morss
+CMD gunicorn --bind 0.0.0.0:8080 -w 4 --preload morss

README.md

@@ -186,7 +186,7 @@ uwsgi --http :8080 --plugin python --wsgi-file main.py
 #### Using Gunicorn
 
 ```shell
-gunicorn morss
+gunicorn --preload morss
 ```
 
 #### Using docker
@@ -316,6 +316,13 @@ will be cleared every time the program is run). Path can be defined with
 - `CACHE=mysql`: MySQL cache. Connection can be defined with the following
 environment variables: `MYSQL_USER`, `MYSQL_PWD`, `MYSQL_DB`, `MYSQL_HOST`
 
+To limit the siz of the cache:
+' `CACHE_SIZE` sets the target number of items in the cache (further items will
+be deleted but the cache migth be temporarily bigger than that). Defaults to 10k
+entries.
+- `CACHE_LIFESPAN` sets how often the cache must be trimmed (i.e. cut down to
+the number of items set in `CACHE_SIZE`). Defaults to 1hr.
+
 ## Configuration
 ### Length limitation
 

morss/__main__.py

@@ -25,36 +25,15 @@ from . import cli
 
 from .morss import MorssException
 
-import wsgiref.simple_server
-import wsgiref.handlers
-
-
-PORT = int(os.getenv('PORT', 8080))
-
 
 def main():
     if 'REQUEST_URI' in os.environ:
         # mod_cgi (w/o file handler)
-
-        app = wsgi.cgi_app
-        app = wsgi.cgi_dispatcher(app)
-        app = wsgi.cgi_error_handler(app)
-        app = wsgi.cgi_encode(app)
-
-        wsgiref.handlers.CGIHandler().run(app)
+        wsgi.cgi_handle_request()
 
     elif len(sys.argv) <= 1:
         # start internal (basic) http server (w/ file handler)
-
-        app = wsgi.cgi_app
-        app = wsgi.cgi_file_handler(app)
-        app = wsgi.cgi_dispatcher(app)
-        app = wsgi.cgi_error_handler(app)
-        app = wsgi.cgi_encode(app)
-
-        print('Serving http://localhost:%s/' % PORT)
-        httpd = wsgiref.simple_server.make_server('', PORT, app)
-        httpd.serve_forever()
+        wsgi.cgi_start_server()
 
     else:
         # as a CLI app

morss/crawler.py

@@ -25,7 +25,9 @@ import chardet
 from cgi import parse_header
 import lxml.html
 import time
+import threading
 import random
+from collections import OrderedDict
 
 try:
     # python 2
@@ -48,6 +50,10 @@ except NameError:
     basestring = unicode = str
 
 
+CACHE_SIZE = int(os.getenv('CACHE_SIZE', 10000)) # max number of items in cache (default: 10k items)
+CACHE_LIFESPAN = int(os.getenv('CACHE_LIFESPAN', 60*60)) # how often to auto-clear the cache (default: 1hr)
+
+
 # uncomment the lines below to ignore SSL certs
 #import ssl
 #ssl._create_default_https_context = ssl._create_unverified_context
@@ -605,6 +611,18 @@ class CacheHandler(BaseHandler):
 class BaseCache:
     """ Subclasses must behave like a dict """
 
+    def trim(self):
+        pass
+
+    def autotrim(self, delay=CACHE_LIFESPAN):
+        # trim the cache every so often
+
+        self.trim()
+
+        t = threading.Timer(delay, self.autotrim)
+        t.daemon = True
+        t.start()
+
     def __contains__(self, url):
         try:
             self[url]
@@ -627,9 +645,15 @@ class SQLiteCache(BaseCache):
             self.con.execute('CREATE TABLE IF NOT EXISTS data (url UNICODE PRIMARY KEY, code INT, msg UNICODE, headers UNICODE, data BLOB, timestamp INT)')
             self.con.execute('pragma journal_mode=WAL')
 
+        self.trim()
+
     def __del__(self):
         self.con.close()
 
+    def trim(self):
+        with self.con:
+            self.con.execute('DELETE FROM data WHERE timestamp <= ( SELECT timestamp FROM ( SELECT timestamp FROM data ORDER BY timestamp DESC LIMIT 1 OFFSET ? ) foo )', (CACHE_SIZE,))
+
     def __getitem__(self, url):
         row = self.con.execute('SELECT * FROM data WHERE url=?', (url,)).fetchone()
 
@@ -660,9 +684,15 @@ class MySQLCacheHandler(BaseCache):
         with self.cursor() as cursor:
             cursor.execute('CREATE TABLE IF NOT EXISTS data (url VARCHAR(255) NOT NULL PRIMARY KEY, code INT, msg TEXT, headers TEXT, data BLOB, timestamp INT)')
 
+        self.trim()
+
     def cursor(self):
         return pymysql.connect(host=self.host, user=self.user, password=self.password, database=self.database, charset='utf8', autocommit=True).cursor()
 
+    def trim(self):
+        with self.cursor() as cursor:
+            cursor.execute('DELETE FROM data WHERE timestamp <= ( SELECT timestamp FROM ( SELECT timestamp FROM data ORDER BY timestamp DESC LIMIT 1 OFFSET %s ) foo )', (CACHE_SIZE,))
+
     def __getitem__(self, url):
         cursor = self.cursor()
         cursor.execute('SELECT * FROM data WHERE url=%s', (url,))
@@ -679,20 +709,39 @@ class MySQLCacheHandler(BaseCache):
                 (url,) + value + value)
 
 
+class CappedDict(OrderedDict, BaseCache):
+    def trim(self):
+        if CACHE_SIZE >= 0:
+            for i in range( max( len(self) - CACHE_SIZE , 0 )):
+                self.popitem(False)
+
+    def __setitem__(self, key, value):
+        # https://docs.python.org/2/library/collections.html#ordereddict-examples-and-recipes
+        if key in self:
+            del self[key]
+        OrderedDict.__setitem__(self, key, value)
+
+
 if 'CACHE' in os.environ:
     if os.environ['CACHE'] == 'mysql':
         default_cache = MySQLCacheHandler(
             user = os.getenv('MYSQL_USER'),
             password = os.getenv('MYSQL_PWD'),
             database = os.getenv('MYSQL_DB'),
-            host = os.getenv('MYSQL_HOST')
+            host = os.getenv('MYSQL_HOST', 'localhost')
         )
 
     elif os.environ['CACHE'] == 'sqlite':
-        default_cache = SQLiteCache(os.getenv('SQLITE_PATH', ':memory:'))
+        if 'SQLITE_PATH' in os.environ:
+            path = os.getenv('SQLITE_PATH') + '/morss-cache.db'
 
         else:
-        default_cache = {}
+            path = ':memory:'
+
+        default_cache = SQLiteCache(path)
+
+else:
+        default_cache = CappedDict()
 
 
 if __name__ == '__main__':

morss/wsgi.py

@@ -22,6 +22,8 @@ import lxml.etree
 
 import cgitb
 import wsgiref.util
+import wsgiref.simple_server
+import wsgiref.handlers
 import mimetypes
 
 try:
@@ -37,6 +39,9 @@ from .morss import FeedFetch, FeedGather, FeedFormat
 from .morss import Options, log, TIMEOUT, DELAY, MorssException
 
 
+PORT = int(os.getenv('PORT', 8080))
+
+
 def parse_options(options):
     """ Turns ['md=True'] into {'md':True} """
     out = {}
@@ -267,3 +272,24 @@ application = cgi_file_handler(application)
 application = cgi_dispatcher(application)
 application = cgi_error_handler(application)
 application = cgi_encode(application)
+
+
+def cgi_handle_request():
+    app = cgi_app
+    app = cgi_dispatcher(app)
+    app = cgi_error_handler(app)
+    app = cgi_encode(app)
+
+    wsgiref.handlers.CGIHandler().run(app)
+
+
+def cgi_start_server():
+    crawler.default_cache.autotrim()
+
+    print('Serving http://localhost:%s/' % PORT)
+    httpd = wsgiref.simple_server.make_server('', PORT, application)
+    httpd.serve_forever()
+
+
+if 'gunicorn' in os.getenv('SERVER_SOFTWARE', ''):
+    crawler.default_cache.autotrim()