[mwm.py] Process metadata

2 files changed, 23 insertions, 4 deletions

diff --git a/tools/python/mwm/find_feature.py b/tools/python/mwm/find_feature.py
index 87b88066d5..2115848f4b 100755
--- a/tools/python/mwm/find_feature.py
+++ b/tools/python/mwm/find_feature.py
@@ -10,6 +10,7 @@ if len(sys.argv) < 4:
     print('  t for inside types ("t hwtag" will find all hwtags-*)')
     print('  et for exact type ("et shop" won\'t find shop-chemist)')
     print('  n for names, case-sensitive ("n Starbucks" for all starbucks)')
+    print('  m for metadata keys ("m flats" for features with flats)')
     sys.exit(1)
 
 typ = sys.argv[2].lower()
@@ -18,7 +19,7 @@ find = sys.argv[3].decode('utf-8')
 mwm = MWM(open(sys.argv[1], 'rb'))
 mwm.read_header()
 mwm.read_types(os.path.join(os.path.dirname(sys.argv[0]), '..', '..', '..', 'data', 'types.txt'))
-for feature in mwm.iter_features():
+for feature in mwm.iter_features(metadata=True):
     found = False
     if typ == 'n' and 'name' in feature['header']:
         for value in feature['header']['name'].values():
@@ -30,5 +31,8 @@ for feature in mwm.iter_features():
                 found = True
             elif typ == 't' and find in t:
                 found = True
+    elif typ == 'm' and 'metadata' in feature:
+        if find in feature['metadata']:
+            found = True
     if found:
-        print(json.dumps(feature, ensure_ascii=False))
+        print(json.dumps(feature, ensure_ascii=False).encode('utf-8'))
diff --git a/tools/python/mwm/mwm.py b/tools/python/mwm/mwm.py
index 99addccf8d..bb4c941690 100644
--- a/tools/python/mwm/mwm.py
+++ b/tools/python/mwm/mwm.py
@@ -12,7 +12,9 @@ from datetime import datetime
 # - Find why polygon geometry is incorrect in iter_features()
 # - Find feature ids in the 'dat' section, or find a way to read the 'offs' section
 
+
 class MWM:
+    # coding/multilang_utf8_string.cpp
     languages = ["default",
                  "en", "ja", "fr", "ko_rm", "ar", "de", "int_name", "ru", "sv", "zh", "fi", "be", "ka", "ko",
                  "he", "nl", "ga", "ja_rm", "el", "it", "es", "zh_pinyin", "th", "cy", "sr", "uk", "ca", "hu",
@@ -20,12 +22,14 @@ class MWM:
                  "af", "ja_kana", "lb", "pt", "hr", "fur", "vi", "tr", "bg", "eo", "lt", "la", "kk", "gsw",
                  "et", "ku", "mn", "mk", "lv", "hi"]
 
+    # indexer/feature_meta.hpp
     metadata = ["0",
                 "cuisine", "open_hours", "phone_number", "fax_number", "stars",
                 "operator", "url", "website", "internet", "ele",
                 "turn_lanes", "turn_lanes_forward", "turn_lanes_backward", "email", "postcode",
                 "wikipedia", "maxspeed", "flats", "height", "min_height",
-                "denomination", "building_levels"]
+                "denomination", "building_levels", "test_id", "ref:sponsored", "price_rate",
+                "rating", "fuel", "routes"]
 
     def __init__(self, f):
         self.f = f
@@ -137,6 +141,8 @@ class MWM:
                         t = self.read_varuint()
                         t = self.metadata[t] if t < len(self.metadata) else str(t)
                         fields[t] = self.read_string()
+                        if t == 'fuel':
+                            fields[t] = fields[t].split('\x01')
             else:
                 while True:
                     t = self.read_uint(1)
@@ -204,13 +210,18 @@ class MWM:
         RELATION = 0xC000000000000000
         RESET = ~(NODE | WAY | RELATION)
 
-    def iter_features(self):
+    def iter_features(self, metadata=False):
         """Reads 'dat' section."""
         if not self.has_tag('dat'):
             return
         # TODO: read 'offs'?
+        md = {}
+        if metadata:
+            md = self.read_metadata()
         self.seek_tag('dat')
+        ftid = -1
         while self.inside_tag('dat'):
+            ftid += 1
             feature = {}
             feature_size = self.read_varuint()
             next_feature = self.f.tell() + feature_size
@@ -245,6 +256,10 @@ class MWM:
                     header['house'] = self.read_numeric_string()
             feature['header'] = header
 
+            # Metadata
+            if ftid in md:
+                feature['metadata'] = md[ftid]
+
             # Geometry
             geometry = {}
             if geom_type == MWM.GeomType.POINT or geom_type == MWM.GeomType.POINT_EX: