Merge pull request #100 from mabel-dev/branching

Branching

tarchia/__version__.py

@@ -1,4 +1,4 @@
-__build__ = 134
+__build__ = 135
 
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.

tarchia/exceptions.py

@@ -86,6 +86,10 @@ def __init__(self, root: str, commit: str):
         super().__init__(message)
 
 
+class UnableToReadBlobError(Exception):
+    """Can't find a blob when trying to add to manifest"""
+
+
 class AmbiguousTableError(Exception):  # pragma: no cover
     def __init__(self, table: str):
         self.table = table

tarchia/interfaces/storage/google_cloud_storage.py

@@ -45,4 +45,6 @@ def read_blob(self, location: str, bucket_in_path: bool = False) -> bytes:
             bucket_name = self.bucket_name
         bucket = self.client.get_bucket(bucket_name)
         blob = bucket.get_blob(location)
-        return blob.download_as_bytes()
+        if blob:
+            return blob.download_as_bytes()
+        return None

tarchia/metadata/manifests/__init__.py

@@ -116,6 +116,11 @@ def build_manifest_entry(path: str, expected_schema: Schema) -> ManifestEntry:
 
     # Read the file bytes and initialize the Parquet file object
     file_bytes = storage_provider.read_blob(blob_path, bucket_in_path=True)
+    if file_bytes is None:
+        from tarchia.exceptions import UnableToReadBlobError
+
+        raise UnableToReadBlobError(f"Unable to read {blob_path}.")
+
     new_manifest_entry.file_size = len(file_bytes)
     new_manifest_entry.sha256_checksum = sha256(file_bytes).hexdigest()
 

tarchia/models/metadata_models.py

@@ -1,3 +1,4 @@
+import time
 from enum import Enum
 from typing import Any
 from typing import List
@@ -149,6 +150,7 @@ class EventTypes(Enum):
     format_version: int = Field(default=1)
     disposition: TableDisposition = Field(default=TableDisposition.SNAPSHOT)
     metadata: dict = Field(default_factory=dict)
+    created_at: int = int(time.time_ns() / 1e6)
 
     def is_valid(self):
         # only columns in the schema can be encrypted
@@ -167,6 +169,7 @@ class OwnerEntry(Eventable, TarchiaBaseModel):
         type (OwnerType): The type of the owner.
         user (str): The name of the user/group that owns this group.
         memberships (List(str)): Identifiers to automatically map users to Owners
+        created_at (int): Timestamp this record was creted.
     """
 
     class EventTypes(Enum):
@@ -180,6 +183,7 @@ class EventTypes(Enum):
     type: OwnerType
     steward: str
     memberships: List[str]
+    created_at: int = int(time.time_ns() / 1e6)
 
     def is_valid(self):
         if not self.name.isidentifier():