Cache Parse() results

mattrobenolt · mattrobenolt · commit 54e000407a63 · 2015-10-11T02:57:58.000-07:00
Each parse takes ~2ms on my machine, and it's pretty common throughout the life of a running
process, to parse identical user-agent strings. This adds a very primitive cache similar in vein
to the cache inside the `urlparse` package.

Before:
```
$ python -m timeit -s 'from ua_parser.user_agent_parser import Parse' 'Parse("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/46.0.2490.52 Safari/537.36")'
100 loops, best of 3: 2.14 msec per loop
```

After:
```
$ python -m timeit -s 'from ua_parser.user_agent_parser import Parse' 'Parse("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/46.0.2490.52 Safari/537.36")'
1000000 loops, best of 3: 0.956 usec per loop
```
diff --git a/Makefile b/Makefile
@@ -1,11 +1,9 @@
-PWD = $(shell pwd)
-
-all: prep test 
+all: prep test
 
 prep:
 	#git submodule update --init
 	#sudo apt-get install python-yaml
-	
+
 test:
 	@#test ! -d tmp && mkdir tmp
 	@export PYTHONPATH=tmp && python setup.py develop -d tmp
@@ -21,4 +19,4 @@ clean:
 	@rm -rf tmp\
 	   ua_parser.egg-info
 
-.PHONY: all clean
+.PHONY: all prep test clean
diff --git a/ua_parser/user_agent_parser.py b/ua_parser/user_agent_parser.py
@@ -196,6 +196,10 @@ def Parse(self, user_agent_string):
         return device, brand, model
 
 
+MAX_CACHE_SIZE = 20
+_parse_cache = {}
+
+
 def Parse(user_agent_string, **jsParseBits):
     """ Parse all the things
     Args:
@@ -205,12 +209,20 @@ def Parse(user_agent_string, **jsParseBits):
       A dictionary containing all parsed bits
     """
     jsParseBits = jsParseBits or {}
-    return {
+    key = (user_agent_string, repr(jsParseBits))
+    cached = _parse_cache.get(key)
+    if cached is not None:
+        return cached
+    if len(_parse_cache) > MAX_CACHE_SIZE:
+        _parse_cache.clear()
+    v = {
         'user_agent': ParseUserAgent(user_agent_string, **jsParseBits),
         'os': ParseOS(user_agent_string, **jsParseBits),
         'device': ParseDevice(user_agent_string, **jsParseBits),
         'string': user_agent_string
     }
+    _parse_cache[key] = v
+    return v
 
 
 def ParseUserAgent(user_agent_string, **jsParseBits):
