fix: infer travel_mode from timelinePath speed when activity type is UNKNOWN

work_trace/wizards/wt_import_timeline_wizard.py

@@ -38,18 +38,32 @@ def _distance_meters(lat1, lon1, lat2, lon2):
 
 
 def _activity_type_to_travel_mode(activity_type):
-    """Map Google activity type string to Odoo travel_mode selection value."""
     if not activity_type:
-        return 'unknown'
+        return None
     t = activity_type.upper()
-    if t in VEHICLE_ACTIVITIES:
-        return 'driving'
+    if 'UNKNOWN' in t:
+        return None
     if t == 'IN_RAIL_VEHICLE':
         return 'transit'
+    if t in VEHICLE_ACTIVITIES:
+        return 'driving'
     if t in WALKING_ACTIVITIES:
         return 'walking'
     if t in CYCLING_ACTIVITIES:
         return 'cycling'
+    return None
+
+
+def _speed_to_travel_mode(max_speed_mph):
+    """Infer travel mode from maximum speed observed along a path."""
+    if max_speed_mph > 150:
+        return 'transit'   # flight
+    if max_speed_mph > 15:
+        return 'driving'
+    if max_speed_mph > 3:
+        return 'cycling'
+    if max_speed_mph > 0:
+        return 'walking'
     return 'unknown'
 
 
@@ -69,6 +83,36 @@ def _parse_ts(ts_str):
         return None
 
 
+def _max_speed_from_path(timeline_path):
+    """Calculate the maximum speed (mph) from a timelinePath segment's points."""
+    points = []
+    for pt in timeline_path:
+        latlng = pt.get('point', '')
+        ts_str = pt.get('time', '')
+        lat, lng = _parse_latlng(latlng)
+        ts = _parse_ts(ts_str)
+        if lat is not None and ts is not None:
+            points.append((ts, lat, lng))
+
+    if len(points) < 2:
+        return 0.0
+
+    points.sort(key=lambda x: x[0])
+    max_speed = 0.0
+    for i in range(1, len(points)):
+        t0, lat0, lng0 = points[i - 1]
+        t1, lat1, lng1 = points[i]
+        dt_hours = (t1 - t0).total_seconds() / 3600.0
+        if dt_hours <= 0:
+            continue
+        dist_miles = _haversine_miles(lat0, lng0, lat1, lng1)
+        speed = dist_miles / dt_hours
+        if speed > max_speed:
+            max_speed = speed
+
+    return max_speed
+
+
 class WtImportTimelineWizard(models.TransientModel):
     _name = 'wt.import.timeline.wizard'
     _description = 'Import Google Timeline'
@@ -102,7 +146,6 @@ class WtImportTimelineWizard(models.TransientModel):
     def action_import(self):
         self.ensure_one()
 
-        # Load data from file upload or server path
         if self.import_mode == 'server_path':
             if not self.server_file_path:
                 raise UserError(_('Please enter the server file path.'))
@@ -123,7 +166,6 @@ class WtImportTimelineWizard(models.TransientModel):
             except Exception as e:
                 raise UserError(_('Invalid JSON file: %s') % str(e))
 
-        # Auto-detect format
         if 'semanticSegments' in data:
             stops = self._parse_semantic_timeline(data)
         elif 'timelineEdits' in data:
@@ -134,7 +176,6 @@ class WtImportTimelineWizard(models.TransientModel):
         if not stops:
             raise UserError(_('No location stops found in the file.'))
 
-        # Filter by minimum stop duration
         min_secs = self.min_stop_minutes * 60
         stops = [s for s in stops
                  if s.get('arrived_at') and s.get('departed_at')
@@ -145,7 +186,6 @@ class WtImportTimelineWizard(models.TransientModel):
 
         stops.sort(key=lambda s: s['arrived_at'])
 
-        # Compute distances and travel times
         for i, stop in enumerate(stops):
             if i > 0:
                 prev = stops[i - 1]
@@ -159,7 +199,6 @@ class WtImportTimelineWizard(models.TransientModel):
                 stop['distance_from_previous'] = 0.0
                 stop['travel_time_from_previous'] = 0.0
 
-        # Skip existing records (deduplicate by arrived_at)
         LocationLog = self.env['wt.location.log']
         existing = set(
             r.arrived_at.strftime('%Y-%m-%d %H:%M:%S')
@@ -207,59 +246,77 @@ class WtImportTimelineWizard(models.TransientModel):
         }
 
     def _parse_semantic_timeline(self, data):
-        """Parse Timeline.json (semanticSegments format from Android export).
+        """Parse Timeline.json (semanticSegments format).
 
-        Iterates segments in order. Activity segments (travel between stops)
-        are captured to provide travel_mode for the following visit segment.
+        Travel mode is determined in priority order:
+          1. activity segment topCandidate.type (if not UNKNOWN)
+          2. max speed calculated from preceding timelinePath points
+          3. 'unknown' as fallback
         """
         stops = []
         segments = data.get('semanticSegments', [])
-        pending_travel_mode = 'unknown'
+
+        # We scan segments in order, accumulating travel signals between visits
+        pending_mode = 'unknown'
+        pending_path_points = []   # all timelinePath points seen since last visit
 
         for seg in segments:
-            # --- Activity / travel segment ---
-            activity = seg.get('activity')
-            if activity:
+            # --- timelinePath: collect GPS path points for speed inference ---
+            if 'timelinePath' in seg:
+                pending_path_points.extend(seg['timelinePath'])
+                continue
+
+            # --- activity segment: try to get explicit travel mode ---
+            if 'activity' in seg:
+                activity = seg['activity']
                 top = activity.get('topCandidate', {})
-                activity_type = top.get('type', '')
-                mode = _activity_type_to_travel_mode(activity_type)
-                if mode != 'unknown':
-                    pending_travel_mode = mode
+                mode = _activity_type_to_travel_mode(top.get('type', ''))
+                if mode:
+                    pending_mode = mode
                 continue
 
-            # --- Visit / stop segment ---
-            visit = seg.get('visit')
-            if not visit:
-                # timelinePath or unknown — ignore but keep pending travel mode
+            # --- visit segment: the stop we care about ---
+            if 'visit' in seg:
+                start_ts = _parse_ts(seg.get('startTime'))
+                end_ts = _parse_ts(seg.get('endTime'))
+                if not start_ts or not end_ts:
+                    pending_mode = 'unknown'
+                    pending_path_points = []
+                    continue
+
+                candidate = seg['visit'].get('topCandidate', {})
+                semantic_type = candidate.get('semanticType', '')
+                latlng_str = candidate.get('placeLocation', {}).get('latLng', '')
+                lat, lng = _parse_latlng(latlng_str) if latlng_str else (None, None)
+
+                category = SEMANTIC_TYPE_CATEGORY.get(semantic_type, '')
+                if not category and semantic_type:
+                    category = semantic_type.replace('_', ' ').title()
+
+                # Determine travel mode: explicit activity type wins,
+                # otherwise infer from max speed along the preceding path
+                travel_mode = pending_mode
+                if travel_mode == 'unknown' and pending_path_points:
+                    max_speed = _max_speed_from_path(pending_path_points)
+                    travel_mode = _speed_to_travel_mode(max_speed)
+
+                stops.append({
+                    'arrived_at': start_ts,
+                    'departed_at': end_ts,
+                    'lat': lat,
+                    'lng': lng,
+                    'place_name': '',
+                    'category': category,
+                    'travel_mode': travel_mode,
+                })
+
+                # Reset for next inter-visit gap
+                pending_mode = 'unknown'
+                pending_path_points = []
                 continue
 
-            start_ts = _parse_ts(seg.get('startTime'))
-            end_ts = _parse_ts(seg.get('endTime'))
-            if not start_ts or not end_ts:
-                pending_travel_mode = 'unknown'
-                continue
-
-            candidate = visit.get('topCandidate', {})
-            semantic_type = candidate.get('semanticType', '')
-            latlng_str = candidate.get('placeLocation', {}).get('latLng', '')
-            lat, lng = _parse_latlng(latlng_str) if latlng_str else (None, None)
-
-            category = SEMANTIC_TYPE_CATEGORY.get(semantic_type, '')
-            if not category and semantic_type:
-                category = semantic_type.replace('_', ' ').title()
-
-            stops.append({
-                'arrived_at': start_ts,
-                'departed_at': end_ts,
-                'lat': lat,
-                'lng': lng,
-                'place_name': '',
-                'category': category,
-                'travel_mode': pending_travel_mode,
-            })
-
-            # Reset after consuming for the next visit
-            pending_travel_mode = 'unknown'
+            # timelineMemory or other unknown segment types — ignore
+            pass
 
         return stops
 
@@ -298,7 +355,8 @@ class WtImportTimelineWizard(models.TransientModel):
             if not window:
                 return 'unknown'
             counts = Counter(a['type'] for a in window)
-            return _activity_type_to_travel_mode(counts.most_common(1)[0][0])
+            top_type = counts.most_common(1)[0][0]
+            return _activity_type_to_travel_mode(top_type) or 'unknown'
 
         stops = []
         current_cluster = [positions[0]]