From 3d0d38f8da5dac18f53fbc8df7cd87ce12e544f4 Mon Sep 17 00:00:00 2001
From: Al <albarrentine@gmail.com>
Date: Wed, 18 May 2016 18:09:26 -0400
Subject: [PATCH] [osm] Adding OSM definitions to quickly determine if a set of
 properties meets one of the libpostal definitions (valid amenity, etc.)

---
 scripts/geodata/osm/definitions.py | 61 ++++++++++++++++++++++++++++++
 1 file changed, 61 insertions(+)
 create mode 100644 scripts/geodata/osm/definitions.py

diff --git a/scripts/geodata/osm/definitions.py b/scripts/geodata/osm/definitions.py
new file mode 100644
index 00000000..6899f7a0
--- /dev/null
+++ b/scripts/geodata/osm/definitions.py
@@ -0,0 +1,61 @@
+import os
+import re
+import six
+
+from collections import defaultdict
+
+this_dir = os.path.realpath(os.path.dirname(__file__))
+
+DEFAULT_SCRIPT_PATH = os.path.join(this_dir, 'fetch_osm_address_data.sh')
+
+valid_key_regex = re.compile('VALID_(.*?)_KEYS="(.*)"')
+kv_regex = re.compile('([^\s]*)=([^\s]*)')
+
+
+class OSMDefinitions(object):
+    ALL = '*'
+
+    ADMIN_BORDER = 'admin_border'
+    AEROWAY = 'aeroway'
+    AMENITY = 'amenity'
+    BUILDING = 'building'
+    HISTORIC = 'historic'
+    LANDUSE = 'landuse'
+    NATURAL = 'natural'
+    NEIGHBORHOOD = 'neighborhood'
+    OFFICE = 'office'
+    PLACE = 'place'
+    SHOP = 'shop'
+    TOURISM = 'tourism'
+    WATERWAY = 'waterway'
+
+    def __init__(self, filename=DEFAULT_SCRIPT_PATH):
+        script = open(filename).read()
+
+        definitions = {}
+
+        for definition, text in valid_key_regex.findall(script):
+            definition = definition.lower()
+
+            kvs = defaultdict(set)
+
+            for k, v in kv_regex.findall(text):
+                if v != '':
+                    kvs[k].add(v)
+                else:
+                    kvs[k].add(self.ALL)
+
+            definitions[definition] = kvs
+
+        self.definitions = definitions
+
+    def meets_definition(self, props, category):
+        defs = self.definitions.get(category, {})
+        if not defs:
+            return False
+        for k, v in six.iteritems(props):
+            if v in defs.get(k, set()):
+                return True
+        return False
+
+osm_definitions = OSMDefinitions()