Ingestion 3d pointclouds

sdk/diffgram/file/file_3d.py

@@ -0,0 +1,112 @@
+from diffgram.core.core import Project
+import json
+import os
+import requests
+import io
+import math
+import uuid
+
+
+class File3D:
+    point_list: list
+    client: Project
+
+    def __init__(self, client, name, point_list = []):
+        self.point_list = []
+        self.client = client
+        self.original_filename = name
+        self.point_list = point_list
+
+    def add_point(self,
+                  x: float,
+                  y: float,
+                  z: float,
+                  intensity = None,
+                  device_id = None,
+                  timestamp = None,
+                  is_ground = False):
+
+        if intensity is not None:
+            if intensity > 1.0 or intensity < 0:
+                raise Exception('Intensity point must be between 0 and 1. Value is: {}'.format(intensity))
+
+        self.point_list.append({
+            'x': x,
+            'y': y,
+            'z': z,
+            'intensity': intensity,
+            'device_id': device_id,
+            'timestamp': timestamp,
+            'is_ground': is_ground,
+        })
+        return self.point_list
+
+    def upload(self, dataset_name = None, chunk_size = 5000000):
+        """
+            Builds a JSON file from current point list and uploads it to
+            Diffgram.
+        :param dataset_name:
+        :param chunk_size: Size of each chunk of the JSON file. Default is 5MB
+        :return:
+        """
+
+        if len(self.point_list) == 0:
+            print('At least 1 point should be provided in the point list. Please add point using add_point()')
+            return False
+        file_data = {
+            'point_list': self.point_list
+        }
+        json_data = json.dumps(file_data)
+
+        with open('data.json', 'w') as outfile:
+            json.dump(json_data, outfile)
+        endpoint = "/api/walrus/project/{}/upload/large".format(
+            self.client.project_string_id
+        )
+        chunk_size = 5000000  # 5 MB chunks
+        dataset_id = self.client.default_directory['id']
+        if dataset_name is not None:
+            dataset_id = self.client.directory.get(dataset_name).id
+
+        with io.StringIO(json_data) as s:
+            pos = s.tell()
+            s.seek(0, os.SEEK_END)
+            file_size = s.tell()
+            s.seek(pos)
+            num_chunks = int(math.ceil(file_size / chunk_size))
+            last_chunk_size = -1
+            if file_size % chunk_size != 0:
+                last_chunk_size = file_size % chunk_size
+
+            uid_upload = str(uuid.uuid4())
+            for i in range(0, num_chunks):
+
+                payload = {
+                    'dzuuid': uid_upload,
+                    'original_filename': self.original_filename,
+                    'dzchunkindex': i,
+                    'dztotalfilesize': file_size,
+                    'dzchunksize': chunk_size,
+                    'dztotalchunkcount': num_chunks,
+                    'dzchunkbyteoffset': i * chunk_size,
+                    'directory_id': dataset_id,
+                    'source': 'from_sensor_fusion_json',
+                }
+                # Adjust final chunk size
+                if i == (num_chunks - 1) and last_chunk_size != -1:
+                    # Read last chunk size here...
+                    payload['dzchunksize'] = last_chunk_size
+
+                # Read file Chunk
+                s.seek(payload['dzchunkbyteoffset'])
+                file_chunk = s.read(payload['dzchunksize'])
+                files = {'file': ('{}_sf.json'.format(self.original_filename), file_chunk)}
+                # Make request to server
+                url = self.client.host + endpoint
+                response = self.client.session.post(url,
+                                                    data = payload,
+                                                    files = files)
+
+                self.client.handle_errors(response)
+
+            return True