acryl-datahub 1.1.0rc4__py3-none-any.whl → 1.1.0.1rc6__py3-none-any.whl

datahub/ingestion/source/openapi_parser.py

@@ -59,17 +59,21 @@ def request_call(
     username: Optional[str] = None,
     password: Optional[str] = None,
     proxies: Optional[dict] = None,
+    verify_ssl: bool = True,
 ) -> requests.Response:
     headers = {"accept": "application/json"}
     if username is not None and password is not None:
         return requests.get(
-            url, headers=headers, auth=HTTPBasicAuth(username, password)
+            url,
+            headers=headers,
+            auth=HTTPBasicAuth(username, password),
+            verify=verify_ssl,
         )
     elif token is not None:
         headers["Authorization"] = f"{token}"
-        return requests.get(url, proxies=proxies, headers=headers)
+        return requests.get(url, proxies=proxies, headers=headers, verify=verify_ssl)
     else:
-        return requests.get(url, headers=headers)
+        return requests.get(url, headers=headers, verify=verify_ssl)
 
 
 def get_swag_json(
@@ -79,10 +83,16 @@ def get_swag_json(
     password: Optional[str] = None,
     swagger_file: str = "",
     proxies: Optional[dict] = None,
+    verify_ssl: bool = True,
 ) -> Dict:
     tot_url = url + swagger_file
     response = request_call(
-        url=tot_url, token=token, username=username, password=password, proxies=proxies
+        url=tot_url,
+        token=token,
+        username=username,
+        password=password,
+        proxies=proxies,
+        verify_ssl=verify_ssl,
     )
 
     if response.status_code != 200:
@@ -127,37 +137,45 @@ def get_endpoints(sw_dict: dict) -> dict:
     check_sw_version(sw_dict)
 
     for p_k, p_o in sw_dict["paths"].items():
-        method = list(p_o)[0]
-        if "200" in p_o[method]["responses"]:
-            base_res = p_o[method]["responses"]["200"]
-        elif 200 in p_o[method]["responses"]:
-            # if you read a plain yml file the 200 will be an integer
-            base_res = p_o[method]["responses"][200]
-        else:
-            # the endpoint does not have a 200 response
-            continue
-
-        if "description" in p_o[method]:
-            desc = p_o[method]["description"]
-        elif "summary" in p_o[method]:
-            desc = p_o[method]["summary"]
-        else:  # still testing
-            desc = ""
-
-        try:
-            tags = p_o[method]["tags"]
-        except KeyError:
-            tags = []
-
-        url_details[p_k] = {"description": desc, "tags": tags, "method": method}
-
-        example_data = check_for_api_example_data(base_res, p_k)
-        if example_data:
-            url_details[p_k]["data"] = example_data
-
-        # checking whether there are defined parameters to execute the call...
-        if "parameters" in p_o[method]:
-            url_details[p_k]["parameters"] = p_o[method]["parameters"]
+        for method, method_spec in p_o.items():
+            # skip non-method keys like "parameters"
+            if method.lower() not in [
+                "get",
+                "post",
+                "put",
+                "delete",
+                "patch",
+                "options",
+                "head",
+            ]:
+                continue
+
+            responses = method_spec.get("responses", {})
+            base_res = responses.get("200") or responses.get(200)
+            if not base_res:
+                # if there is no 200 response, we skip this method
+                continue
+
+            # if the description is not present, we will use the summary
+            # if both are not present, we will use an empty string
+            desc = method_spec.get("description") or method_spec.get("summary", "")
+
+            # if the tags are not present, we will use an empty list
+            tags = method_spec.get("tags", [])
+
+            url_details[p_k] = {
+                "description": desc,
+                "tags": tags,
+                "method": method.upper(),
+            }
+
+            example_data = check_for_api_example_data(base_res, p_k)
+            if example_data:
+                url_details[p_k]["data"] = example_data
+
+            # checking whether there are defined parameters to execute the call...
+            if "parameters" in p_o[method]:
+                url_details[p_k]["parameters"] = p_o[method]["parameters"]
 
     return dict(sorted(url_details.items()))
 
@@ -358,6 +376,7 @@ def get_tok(
     tok_url: str = "",
     method: str = "post",
     proxies: Optional[dict] = None,
+    verify_ssl: bool = True,
 ) -> str:
     """
     Trying to post username/password to get auth.
@@ -368,7 +387,7 @@ def get_tok(
         # this will make a POST call with username and password
         data = {"username": username, "password": password, "maxDuration": True}
         # url2post = url + "api/authenticate/"
-        response = requests.post(url4req, proxies=proxies, json=data)
+        response = requests.post(url4req, proxies=proxies, json=data, verify=verify_ssl)
         if response.status_code == 200:
             cont = json.loads(response.content)
             if "token" in cont:  # other authentication scheme
@@ -377,7 +396,7 @@ def get_tok(
                 token = f"Bearer {cont['tokens']['access']}"
     elif method == "get":
         # this will make a GET call with username and password
-        response = requests.get(url4req)
+        response = requests.get(url4req, verify=verify_ssl)
         if response.status_code == 200:
             cont = json.loads(response.content)
             token = cont["token"]

datahub/ingestion/source/s3/source.py

@@ -7,7 +7,7 @@ import re
 import time
 from datetime import datetime
 from pathlib import PurePath
-from typing import TYPE_CHECKING, Dict, Iterable, List, Optional, Tuple
+from typing import TYPE_CHECKING, Any, Dict, Iterable, List, Optional, Tuple
 from urllib.parse import urlparse
 
 import smart_open.compression as so_compression
@@ -43,6 +43,9 @@ from datahub.ingestion.source.aws.s3_util import (
     strip_s3_prefix,
 )
 from datahub.ingestion.source.data_lake_common.data_lake_utils import ContainerWUCreator
+from datahub.ingestion.source.data_lake_common.object_store import (
+    create_object_store_adapter,
+)
 from datahub.ingestion.source.data_lake_common.path_spec import FolderTraversalMethod
 from datahub.ingestion.source.s3.config import DataLakeSourceConfig, PathSpec
 from datahub.ingestion.source.s3.report import DataLakeSourceReport
@@ -197,12 +200,59 @@ class S3Source(StatefulIngestionSourceBase):
     report: DataLakeSourceReport
     profiling_times_taken: List[float]
     container_WU_creator: ContainerWUCreator
+    object_store_adapter: Any
 
     def __init__(self, config: DataLakeSourceConfig, ctx: PipelineContext):
         super().__init__(config, ctx)
         self.source_config = config
         self.report = DataLakeSourceReport()
         self.profiling_times_taken = []
+        self.container_WU_creator = ContainerWUCreator(
+            self.source_config.platform,
+            self.source_config.platform_instance,
+            self.source_config.env,
+        )
+
+        # Create an object store adapter for handling external URLs and paths
+        if self.is_s3_platform():
+            # Get the AWS region from config, if available
+            aws_region = None
+            if self.source_config.aws_config:
+                aws_region = self.source_config.aws_config.aws_region
+
+                # For backward compatibility with tests: if we're using a test endpoint, use us-east-1
+                if self.source_config.aws_config.aws_endpoint_url and (
+                    "localstack"
+                    in self.source_config.aws_config.aws_endpoint_url.lower()
+                    or "storage.googleapis.com"
+                    in self.source_config.aws_config.aws_endpoint_url.lower()
+                ):
+                    aws_region = "us-east-1"
+
+            # Create an S3 adapter with the configured region
+            self.object_store_adapter = create_object_store_adapter(
+                "s3", aws_region=aws_region
+            )
+
+            # Special handling for GCS via S3 (via boto compatibility layer)
+            if (
+                self.source_config.aws_config
+                and self.source_config.aws_config.aws_endpoint_url
+                and "storage.googleapis.com"
+                in self.source_config.aws_config.aws_endpoint_url.lower()
+            ):
+                # We need to preserve the S3-style paths but use GCS external URL generation
+                self.object_store_adapter = create_object_store_adapter("gcs")
+                # Override create_s3_path to maintain S3 compatibility
+                self.object_store_adapter.register_customization(
+                    "create_s3_path", lambda bucket, key: f"s3://{bucket}/{key}"
+                )
+        else:
+            # For local files, create a default adapter
+            self.object_store_adapter = create_object_store_adapter(
+                self.source_config.platform or "file"
+            )
+
         config_report = {
             config_option: config.dict().get(config_option)
             for config_option in config_options_to_report
@@ -605,6 +655,19 @@ class S3Source(StatefulIngestionSourceBase):
             maxPartition=max_partition_summary, minPartition=min_partition_summary
         )
 
+    def get_external_url(self, table_data: TableData) -> Optional[str]:
+        """
+        Get the external URL for a table using the configured object store adapter.
+
+        Args:
+            table_data: Table data containing path information
+
+        Returns:
+            An external URL or None if not applicable
+        """
+        # The adapter handles all the URL generation with proper region handling
+        return self.object_store_adapter.get_external_url(table_data)
+
     def ingest_table(
         self, table_data: TableData, path_spec: PathSpec
     ) -> Iterable[MetadataWorkUnit]:
@@ -674,6 +737,7 @@ class S3Source(StatefulIngestionSourceBase):
                 if max_partition
                 else None
             ),
+            externalUrl=self.get_external_url(table_data),
         )
         aspects.append(dataset_properties)
         if table_data.size_in_bytes > 0:
@@ -1082,11 +1146,6 @@ class S3Source(StatefulIngestionSourceBase):
                     )
 
     def get_workunits_internal(self) -> Iterable[MetadataWorkUnit]:
-        self.container_WU_creator = ContainerWUCreator(
-            self.source_config.platform,
-            self.source_config.platform_instance,
-            self.source_config.env,
-        )
         with PerfTimer() as timer:
             assert self.source_config.path_specs
             for path_spec in self.source_config.path_specs:

datahub/ingestion/source/snowflake/snowflake_config.py

@@ -22,6 +22,7 @@ from datahub.ingestion.api.incremental_properties_helper import (
 from datahub.ingestion.glossary.classification_mixin import (
     ClassificationSourceConfigMixin,
 )
+from datahub.ingestion.source.snowflake.constants import SnowflakeEdition
 from datahub.ingestion.source.snowflake.snowflake_connection import (
     SnowflakeConnectionConfig,
 )
@@ -326,6 +327,18 @@ class SnowflakeV2Config(
         " Map of share name -> details of share.",
     )
 
+    known_snowflake_edition: Optional[SnowflakeEdition] = Field(
+        default=None,
+        description="Explicitly specify the Snowflake edition (STANDARD or ENTERPRISE). If unset, the edition will be inferred automatically using 'SHOW TAGS'.",
+    )
+
+    # Allows empty containers to be ingested before datasets are added, avoiding permission errors
+    warn_no_datasets: bool = Field(
+        hidden_from_docs=True,
+        default=False,
+        description="If True, warns when no datasets are found during ingestion. If False, ingestion fails when no datasets are found.",
+    )
+
     include_assertion_results: bool = Field(
         default=False,
         description="Whether to ingest assertion run results for assertions created using Datahub"

datahub/ingestion/source/snowflake/snowflake_queries.py

@@ -127,6 +127,8 @@ class SnowflakeQueriesExtractorReport(Report):
     sql_aggregator: Optional[SqlAggregatorReport] = None
 
     num_ddl_queries_dropped: int = 0
+    num_stream_queries_observed: int = 0
+    num_create_temp_view_queries_observed: int = 0
     num_users: int = 0
 
 
@@ -373,6 +375,13 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
                     if entry:
                         yield entry
 
+    @classmethod
+    def _has_temp_keyword(cls, query_text: str) -> bool:
+        return (
+            re.search(r"\bTEMP\b", query_text, re.IGNORECASE) is not None
+            or re.search(r"\bTEMPORARY\b", query_text, re.IGNORECASE) is not None
+        )
+
     def _parse_audit_log_row(
         self, row: Dict[str, Any], users: UsersMapping
     ) -> Optional[Union[TableRename, TableSwap, PreparsedQuery, ObservedQuery]]:
@@ -389,6 +398,15 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
             key = key.lower()
             res[key] = value
 
+        timestamp: datetime = res["query_start_time"]
+        timestamp = timestamp.astimezone(timezone.utc)
+
+        # TODO need to map snowflake query types to ours
+        query_text: str = res["query_text"]
+        query_type: QueryType = SNOWFLAKE_QUERY_TYPE_MAPPING.get(
+            res["query_type"], QueryType.UNKNOWN
+        )
+
         direct_objects_accessed = res["direct_objects_accessed"]
         objects_modified = res["objects_modified"]
         object_modified_by_ddl = res["object_modified_by_ddl"]
@@ -399,9 +417,9 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
                 "Error fetching ddl lineage from Snowflake"
             ):
                 known_ddl_entry = self.parse_ddl_query(
-                    res["query_text"],
+                    query_text,
                     res["session_id"],
-                    res["query_start_time"],
+                    timestamp,
                     object_modified_by_ddl,
                     res["query_type"],
                 )
@@ -419,24 +437,38 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
             )
         )
 
-        # Use direct_objects_accessed instead objects_modified
-        # objects_modified returns $SYS_VIEW_X with no mapping
+        # There are a couple cases when we'd want to prefer our own SQL parsing
+        # over Snowflake's metadata.
+        # 1. For queries that use a stream, objects_modified returns $SYS_VIEW_X with no mapping.
+        #    We can check direct_objects_accessed to see if there is a stream used, and if so,
+        #    prefer doing SQL parsing over Snowflake's metadata.
+        # 2. For queries that create a view, objects_modified is empty and object_modified_by_ddl
+        #    contains the view name and columns. Because `object_modified_by_ddl` doesn't contain
+        #    source columns e.g. lineage information, we must do our own SQL parsing. We're mainly
+        #    focused on temporary views. It's fine if we parse a couple extra views, but in general
+        #    we want view definitions to come from Snowflake's schema metadata and not from query logs.
+
         has_stream_objects = any(
             obj.get("objectDomain") == "Stream" for obj in direct_objects_accessed
         )
+        is_create_view = query_type == QueryType.CREATE_VIEW
+        is_create_temp_view = is_create_view and self._has_temp_keyword(query_text)
+
+        if has_stream_objects or is_create_temp_view:
+            if has_stream_objects:
+                self.report.num_stream_queries_observed += 1
+            elif is_create_temp_view:
+                self.report.num_create_temp_view_queries_observed += 1
 
-        # If a stream is used, default to query parsing.
-        if has_stream_objects:
-            logger.debug("Found matching stream object")
             return ObservedQuery(
-                query=res["query_text"],
+                query=query_text,
                 session_id=res["session_id"],
-                timestamp=res["query_start_time"].astimezone(timezone.utc),
+                timestamp=timestamp,
                 user=user,
                 default_db=res["default_db"],
                 default_schema=res["default_schema"],
                 query_hash=get_query_fingerprint(
-                    res["query_text"], self.identifiers.platform, fast=True
+                    query_text, self.identifiers.platform, fast=True
                 ),
             )
 
@@ -502,25 +534,17 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
                     )
                 )
 
-        timestamp: datetime = res["query_start_time"]
-        timestamp = timestamp.astimezone(timezone.utc)
-
-        # TODO need to map snowflake query types to ours
-        query_type = SNOWFLAKE_QUERY_TYPE_MAPPING.get(
-            res["query_type"], QueryType.UNKNOWN
-        )
-
         entry = PreparsedQuery(
             # Despite having Snowflake's fingerprints available, our own fingerprinting logic does a better
             # job at eliminating redundant / repetitive queries. As such, we include the fast fingerprint
             # here
             query_id=get_query_fingerprint(
-                res["query_text"],
+                query_text,
                 self.identifiers.platform,
                 fast=True,
                 secondary_id=res["query_secondary_fingerprint"],
             ),
-            query_text=res["query_text"],
+            query_text=query_text,
             upstreams=upstreams,
             downstream=downstream,
             column_lineage=column_lineage,
@@ -543,7 +567,6 @@ class SnowflakeQueriesExtractor(SnowflakeStructuredReportMixin, Closeable):
         object_modified_by_ddl: dict,
         query_type: str,
     ) -> Optional[Union[TableRename, TableSwap]]:
-        timestamp = timestamp.astimezone(timezone.utc)
         if (
             object_modified_by_ddl["operationType"] == "ALTER"
             and query_type == "RENAME_TABLE"

datahub/ingestion/source/snowflake/snowflake_query.py

@@ -43,13 +43,6 @@ class SnowflakeQuery:
     ACCESS_HISTORY_TABLE_VIEW_DOMAINS_FILTER = "({})".format(
         ",".join(f"'{domain}'" for domain in ACCESS_HISTORY_TABLE_VIEW_DOMAINS)
     )
-    ACCESS_HISTORY_TABLE_DOMAINS_FILTER = (
-        "("
-        f"'{SnowflakeObjectDomain.TABLE.capitalize()}',"
-        f"'{SnowflakeObjectDomain.VIEW.capitalize()}',"
-        f"'{SnowflakeObjectDomain.STREAM.capitalize()}',"
-        ")"
-    )
 
     @staticmethod
     def current_account() -> str:

datahub/ingestion/source/snowflake/snowflake_v2.py

@@ -9,6 +9,7 @@ import re
 from dataclasses import dataclass
 from typing import Dict, Iterable, List, Optional, Union
 
+from datahub.configuration.time_window_config import BaseTimeWindowConfig
 from datahub.ingestion.api.common import PipelineContext
 from datahub.ingestion.api.decorators import (
     SupportStatus,
@@ -551,11 +552,15 @@ class SnowflakeV2Source(
             and len(discovered_views) == 0
             and len(discovered_streams) == 0
         ):
-            self.structured_reporter.failure(
-                GENERIC_PERMISSION_ERROR_KEY,
-                "No tables/views/streams found. Please check permissions.",
-            )
-            return
+            if self.config.warn_no_datasets:
+                self.structured_reporter.warning(
+                    "No tables/views/streams found. Verify dataset permissions if Snowflake source is not empty.",
+                )
+            else:
+                self.structured_reporter.failure(
+                    GENERIC_PERMISSION_ERROR_KEY,
+                    "No tables/views/streams found. Verify dataset permissions in Snowflake.",
+                )
 
         self.discovered_datasets = (
             discovered_tables + discovered_views + discovered_streams
@@ -571,7 +576,11 @@ class SnowflakeV2Source(
                 queries_extractor = SnowflakeQueriesExtractor(
                     connection=self.connection,
                     config=SnowflakeQueriesExtractorConfig(
-                        window=self.config,
+                        window=BaseTimeWindowConfig(
+                            start_time=self.config.start_time,
+                            end_time=self.config.end_time,
+                            bucket_duration=self.config.bucket_duration,
+                        ),
                         temporary_tables_pattern=self.config.temporary_tables_pattern,
                         include_lineage=self.config.include_table_lineage,
                         include_usage_statistics=self.config.include_usage_stats,
@@ -732,6 +741,8 @@ class SnowflakeV2Source(
             return None
 
     def is_standard_edition(self) -> bool:
+        if self.config.known_snowflake_edition is not None:
+            return self.config.known_snowflake_edition == SnowflakeEdition.STANDARD
         try:
             self.connection.query(SnowflakeQuery.show_tags())
             return False

datahub/ingestion/source/sql/athena.py

@@ -323,6 +323,7 @@ class Partitionitem:
     "Optionally enabled via configuration. Profiling uses sql queries on whole table which can be expensive operation.",
 )
 @capability(SourceCapability.LINEAGE_COARSE, "Supported for S3 tables")
+@capability(SourceCapability.LINEAGE_FINE, "Supported for S3 tables")
 @capability(SourceCapability.DESCRIPTIONS, "Enabled by default")
 class AthenaSource(SQLAlchemySource):
     """

datahub/ingestion/source/sql/hive.py

@@ -139,7 +139,7 @@ class StoragePathParser:
                 path = f"{parsed.netloc}/{parsed.path.lstrip('/')}"
 
             elif platform == StoragePlatform.AZURE:
-                if scheme in ("abfs", "abfss"):
+                if scheme in ("abfs", "abfss", "wasbs"):
                     # Format: abfss://container@account.dfs.core.windows.net/path
                     container = parsed.netloc.split("@")[0]
                     path = f"{container}/{parsed.path.lstrip('/')}"
@@ -153,7 +153,7 @@ class StoragePathParser:
 
             elif platform == StoragePlatform.DBFS:
                 # For DBFS, use path as-is
-                path = parsed.path.lstrip("/")
+                path = "/" + parsed.path.lstrip("/")
 
             elif platform == StoragePlatform.LOCAL:
                 # For local files, use full path
@@ -169,7 +169,6 @@ class StoragePathParser:
             # Clean up the path
             path = path.rstrip("/")  # Remove trailing slashes
             path = re.sub(r"/+", "/", path)  # Normalize multiple slashes
-            path = f"/{path}"
 
             return platform, path