tinybird 4.6.11.dev0__tar.gz → 4.6.12__tar.gz

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: tinybird
-Version: 4.6.11.dev0
+Version: 4.6.12
 Summary: Tinybird Command Line Tool
 Home-page: https://www.tinybird.co/docs/forward/commands
 Author: Tinybird
@@ -52,6 +52,18 @@ The Tinybird command-line tool allows you to use all the Tinybird functionality
 Changelog
 ----------
 
+4.6.12
+********
+
+- `Added` Forward sink pipes support ``{{ tb_secret(...) }}`` in ``EXPORT_*`` metadata (for example ``EXPORT_BUCKET_URI``). Secrets are resolved at deploy time and baked into the sink settings; changing a secret requires a redeploy.
+- `Fixed` Forward deploy now raises an error when a secret used in sink metadata resolves to an empty value.
+- `Fixed` Forward deploy now rejects ``{{ tb_secret(...) }}`` in ``EXPORT_SCHEDULE`` and ``COPY_SCHEDULE``. Use a literal cron expression (for example ``0 * * * *``) or ``@on-demand``.
+
+4.6.11
+********
+
+- `Added` `tb datasource sample` support for DynamoDB connectors.
+
 4.6.10
 ********
 

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/datafile/common.py

@@ -401,12 +401,15 @@ class Datafile:
         if node.get("mode") and node["mode"] not in ["append", "replace"]:
             raise DatafileValidationError("COPY node mode must be append or replace")
         # copy schedule must be @on-demand or a cron-expression
-        if (
-            node.get("copy_schedule")
-            and node["copy_schedule"] != ON_DEMAND
-            and not croniter.is_valid(node["copy_schedule"])
-        ):
-            raise DatafileValidationError("COPY node schedule must be @on-demand or a valid cron expression.")
+        copy_schedule = node.get("copy_schedule")
+        if copy_schedule and copy_schedule != ON_DEMAND:
+            if isinstance(copy_schedule, str) and "tb_secret" in copy_schedule:
+                raise DatafileValidationError(
+                    "COPY_SCHEDULE does not support { tb_secret(...) }. "
+                    'Use a literal cron expression (for example "0 * * * *") or @on-demand.'
+                )
+            if not croniter.is_valid(copy_schedule):
+                raise DatafileValidationError("COPY node schedule must be @on-demand or a valid cron expression.")
         for key in node.keys():
             if key not in CopyParameters.valid_params():
                 raise DatafileValidationError(
@@ -503,10 +506,16 @@ class Datafile:
 
         # Validate schedule format (common for both Kafka and S3/GCS)
         export_schedule = node.get("export_schedule")
-        if export_schedule and export_schedule != ON_DEMAND and not croniter.is_valid(export_schedule):
-            raise DatafileValidationError(
-                f"Sink node {repr(node['name'])} has invalid export_schedule '{export_schedule}'. Must be @on-demand or a valid cron expression."
-            )
+        if export_schedule and export_schedule != ON_DEMAND:
+            if isinstance(export_schedule, str) and "tb_secret" in export_schedule:
+                raise DatafileValidationError(
+                    f"Sink node {repr(node['name'])}: EXPORT_SCHEDULE does not support {{ tb_secret(...) }}. "
+                    'Use a literal cron expression (for example "0 * * * *") or @on-demand.'
+                )
+            if not croniter.is_valid(export_schedule):
+                raise DatafileValidationError(
+                    f"Sink node {repr(node['name'])} has invalid export_schedule '{export_schedule}'. Must be @on-demand or a valid cron expression."
+                )
 
     def validate(self):
         if self.kind == DatafileKind.pipe:

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/sql_template.py

@@ -2911,13 +2911,21 @@ def extract_variables_from_sql(sql: str, params: List[Dict[str, Any]]) -> Dict[s
     return defaults
 
 
-def render_template_with_secrets(name: str, content: str, secrets: Optional[Dict[str, str]] = None) -> str:
+def render_template_with_secrets(
+    name: str,
+    content: str,
+    secrets: Optional[Dict[str, str]] = None,
+    empty_secret_raises: bool = False,
+) -> str:
     """Renders a template with secrets, allowing for default values.
 
     Args:
         name: The name of the template
         content: The template content
         secrets: A dictionary mapping secret names to their values
+        empty_secret_raises: When True, empty secret values or empty defaults raise
+            instead of returning '""'. Used for sink metadata (Forward deploy) where
+            an empty bucket path must not be baked as s3://"".
 
     Returns:
         The rendered template
@@ -2959,6 +2967,26 @@ def render_template_with_secrets(name: str, content: str, secrets: Optional[Dict
         Traceback (most recent call last):
         ...
         tinybird.sql_template.SQLTemplateException: Template Syntax Error: Cannot access secret 'MISSING_SECRET'. Check the secret exists in the Workspace and the token has the required scope.
+
+        >>> render_template_with_secrets(
+        ...     "sink_bucket",
+        ...     "s3://{{ tb_secret('MISSING_SECRET', '') }}",
+        ...     secrets = {},
+        ...     empty_secret_raises=True,
+        ... )
+        Traceback (most recent call last):
+        ...
+        tinybird.sql_template.SQLTemplateException: Template Syntax Error: Secret 'MISSING_SECRET' resolves to an empty value and cannot be used in sink metadata.
+
+        >>> render_template_with_secrets(
+        ...     "sink_bucket",
+        ...     "s3://{{ tb_secret('EMPTY_SECRET', 'fallback') }}",
+        ...     secrets = {'EMPTY_SECRET': ''},
+        ...     empty_secret_raises=True,
+        ... )
+        Traceback (most recent call last):
+        ...
+        tinybird.sql_template.SQLTemplateException: Template Syntax Error: Secret 'EMPTY_SECRET' resolves to an empty value and cannot be used in sink metadata.
     """
     if not secrets:
         secrets = {}
@@ -2974,15 +3002,24 @@ def render_template_with_secrets(name: str, content: str, secrets: Optional[Dict
             The secret value or default
 
         Raises:
-            SQLTemplateException: If the secret is not found and no default is provided
+            SQLTemplateException: If the secret is not found and no default is provided,
+                or if empty_secret_raises is True and the resolved value is empty
         """
         if secret_name in secrets:
             value = secrets[secret_name]
             if isinstance(value, str) and len(value) == 0:
+                if empty_secret_raises:
+                    raise SQLTemplateException(
+                        f"Secret '{secret_name}' resolves to an empty value and cannot be used in sink metadata."
+                    )
                 return '""'
             return value
         elif default is not None:
             if isinstance(default, str) and len(default) == 0:
+                if empty_secret_raises:
+                    raise SQLTemplateException(
+                        f"Secret '{secret_name}' resolves to an empty value and cannot be used in sink metadata."
+                    )
                 return '""'
             return default
         else:

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/sql_toolset.py

@@ -158,7 +158,20 @@ def has_unoptimized_join(sql: str, left_table: Optional[Union[Tuple[str, str], T
             raise UnoptimizedJoinException(sql)
 
 
-def format_where_for_mutation_command(where_clause: str, lightweight: bool = False) -> str:
+def _format_fragment_for_mutation_command(fragment: str) -> str:
+    """Normalize a SQL fragment the way CH stores it in the
+    `system.mutations.command` column: format the expression with CH's own
+    formatter, then escape it as it appears inside the serialized command
+    (string literals keep backslash-escaped quotes)."""
+    formatted = chquery.format(f"""SELECT {fragment}""").split("SELECT ")[1]
+    formatted = formatted.replace("\\", "\\\\").replace("'", "''")
+    quoted = chquery.format(f"SELECT '{formatted}'").split("SELECT ")[1]
+    return quoted[1:-1]
+
+
+def format_where_for_mutation_command(
+    where_clause: str, lightweight: bool = False, partition: Optional[str] = None
+) -> str:
     """
     >>> format_where_for_mutation_command("numnights = 99")
     'DELETE WHERE numnights = 99'
@@ -172,12 +185,18 @@ def format_where_for_mutation_command(where_clause: str, lightweight: bool = Fal
     "DELETE WHERE reservationid = \\\\'\\\\\\\\\\\\'foo\\\\'"
     >>> format_where_for_mutation_command("number < 3", lightweight=True)
     'UPDATE _row_exists = 0 WHERE number < 3'
+    >>> format_where_for_mutation_command("number < 3", lightweight=True, partition="201901")
+    'UPDATE _row_exists = 0 IN PARTITION 201901 WHERE number < 3'
+    >>> format_where_for_mutation_command("number < 3", lightweight=True, partition="'2019-01-01'")
+    "UPDATE _row_exists = 0 IN PARTITION \\\\'2019-01-01\\\\' WHERE number < 3"
     """
-    formatted_condition = chquery.format(f"""SELECT {where_clause}""").split("SELECT ")[1]
-    formatted_condition = formatted_condition.replace("\\", "\\\\").replace("'", "''")
-    quoted_condition = chquery.format(f"SELECT '{formatted_condition}'").split("SELECT ")[1]
-    prefix = "UPDATE _row_exists = 0 WHERE" if lightweight else "DELETE WHERE"
-    return f"{prefix} {quoted_condition[1:-1]}"
+    quoted_condition = _format_fragment_for_mutation_command(where_clause)
+    prefix = "UPDATE _row_exists = 0" if lightweight else "DELETE"
+    # Partition-scoped mutations store the partition expression in the
+    # command (`… IN PARTITION <expr> WHERE <cond>`), normalized by the same
+    # CH formatter, so it must be part of the match.
+    partition_clause = f" IN PARTITION {_format_fragment_for_mutation_command(partition)}" if partition else ""
+    return f"{prefix}{partition_clause} WHERE {quoted_condition}"
 
 
 # Functions that take table/dictionary names as string literal arguments.

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/__cli__.py

@@ -4,5 +4,5 @@ __description__ = 'Tinybird Command Line Tool'
 __url__ = 'https://www.tinybird.co/docs/forward/commands'
 __author__ = 'Tinybird'
 __author_email__ = 'support@tinybird.co'
-__version__ = '4.6.11.dev0'
-__revision__ = 'd98671a'
+__version__ = '4.6.12'
+__revision__ = '9f77ad2'

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/client.py

@@ -546,20 +546,36 @@ class TinyB:
     def datasource_sync(self, datasource_id: str):
         return self._req(f"/v0/datasources/{datasource_id}/scheduling/runs", method="POST", data="")
 
-    def datasource_sample(self, datasource_name: str, max_files: int = 1) -> Dict[str, Any]:
-        """Start a sample import job for an S3/GCS connected datasource.
+    def datasource_sample(
+        self,
+        datasource_name: str,
+        max_files: int = 1,
+        rows: Optional[int] = None,
+        max_bytes: Optional[str] = None,
+        full_export: bool = False,
+    ) -> Dict[str, Any]:
+        """Start a sample import job for an S3/GCS/DynamoDB connected datasource.
 
         Args:
             datasource_name: Name of the datasource to import sample data into
-            max_files: Maximum number of files to import (default 1, max 10)
+            max_files: Maximum number of files to import for blob storage connectors (default 1, max 10)
+            rows: For DynamoDB, the maximum number of rows to scan and import (mutually exclusive with max_bytes)
+            max_bytes: For DynamoDB, the maximum approximate JSONEachRow bytes to import, capped by the server's
+                workspace limit (mutually exclusive with rows)
+            full_export: For DynamoDB, trigger a full PITR export instead of a bounded scan
 
         Returns:
             dict with job info including id, job_id, job_url, job, status
         """
+        payload: Dict[str, Any] = {"max_files": max_files, "full_export": full_export}
+        if rows is not None:
+            payload["rows"] = rows
+        if max_bytes is not None:
+            payload["max_bytes"] = max_bytes
         return self._req(
             f"/v0/datasources/{datasource_name}/sample",
             method="POST",
-            data=json.dumps({"max_files": max_files}),
+            data=json.dumps(payload),
         )
 
     def datasource_scheduling_state(self, datasource_id: str):

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/modules/common.py

@@ -1614,6 +1614,7 @@ def run_aws_iamrole_connection_flow(
     connection_name: str,
     policy: str,
     local_unavailable: bool = False,
+    env: str = "local",
 ) -> Tuple[str, str, Optional[TinyB], Optional[TinyB]]:
     """
     Run the interactive AWS IAM Role connection flow for S3 or DynamoDB.
@@ -1628,6 +1629,9 @@ def run_aws_iamrole_connection_flow(
         connection_name: The name for the connection being created.
         policy: The access policy type ('read' or 'write').
         local_unavailable: If True, local environment is unavailable (e.g., missing AWS credentials).
+        env: The active target environment ('local' or 'cloud'). When not local, the
+            passed-in `client` is reused as the cloud client so secrets land in the
+            workspace the command targets (including branches).
 
     Returns:
         A tuple containing:
@@ -1736,12 +1740,15 @@ def run_aws_iamrole_connection_flow(
 
     if use_cloud:
         try:
-            cloud_client = TinyB(
-                token=config.get("token", ""),
-                host=config.get("host", ""),
-                staging=False,
-                request_from=getattr(client, "request_from", None),
-            )
+            if env != "local":
+                cloud_client = client
+            else:
+                cloud_client = TinyB(
+                    token=config.get("token", ""),
+                    host=config.get("host", ""),
+                    staging=False,
+                    request_from=getattr(client, "request_from", None),
+                )
         except Exception as e:
             click.echo(FeedbackManager.warning(message=f"Failed to initialize cloud client: {e}"))
             click.echo(FeedbackManager.warning(message="Continuing without cloud environment."))

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/modules/connection_dynamodb.py

@@ -158,6 +158,7 @@ def connection_create_dynamodb(
         connection_name=connection_name,
         policy="read",
         local_unavailable=local_aws_unavailable,
+        env=env,
     )
 
     unique_suffix = uuid.uuid4().hex[:8]

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/modules/connection_s3.py

@@ -363,6 +363,7 @@ def connection_create_s3(
         connection_name=connection_name,
         policy=access_type.lower(),
         local_unavailable=local_aws_unavailable,
+        env=env,
     )
     unique_suffix = uuid.uuid4().hex[:8]  # Use first 8 chars of a UUID for brevity
     secret_name = f"s3_role_arn_{connection_name}_{unique_suffix}"

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird/tb/modules/datasource.py

@@ -862,20 +862,50 @@ def datasource_sync(ctx: Context, datasource_name: str, yes: bool):
     default=False,
     help="Wait for the import job to finish",
 )
+@click.option(
+    "--rows",
+    default=None,
+    type=int,
+    help="For DynamoDB, the maximum number of rows to scan and import (default 1500; mutually exclusive with --max-bytes)",
+)
+@click.option(
+    "--max-bytes",
+    default=None,
+    type=str,
+    help="For DynamoDB, the maximum approximate JSONEachRow bytes to import, e.g. 500MB (capped at 10GB by default; mutually exclusive with --rows)",
+)
+@click.option(
+    "--full-export",
+    is_flag=True,
+    default=False,
+    help="For DynamoDB, trigger a full PITR export instead of a bounded sample (mutually exclusive with --rows and --max-bytes)",
+)
 @click.pass_context
-def datasource_sample(ctx: Context, datasource_name: str, max_files: int, wait: bool) -> None:
-    """Import sample data from a datasource connected to an S3 or GCS bucket.
-
-    This command only works with datasources that have IMPORT_CONNECTION_NAME and
-    IMPORT_BUCKET_URI configured (i.e., datasources ingesting from S3 or GCS).
-    It imports a limited number of files from the bucket URI pattern, useful for
-    testing data pipelines in branches without importing the entire dataset.
+def datasource_sample(
+    ctx: Context,
+    datasource_name: str,
+    max_files: int,
+    wait: bool,
+    rows: Optional[int],
+    max_bytes: Optional[str],
+    full_export: bool,
+) -> None:
+    """Import sample data from a datasource connected to S3, GCS, or DynamoDB.
+
+    For S3 and GCS, this imports a limited number of files from the bucket URI
+    pattern. For DynamoDB, this scans and imports a bounded sample limited by
+    either --rows or --max-bytes (defaulting to 1500 rows), or --full-export to
+    trigger a PITR export of the whole table.
 
     By default, returns immediately with job info. Use --wait to block until complete.
 
     Examples:
         tb --branch=my_branch datasource sample my_s3_ds
         tb --branch=my_branch datasource sample my_s3_ds --max-files 3 --wait
+        tb --branch=my_branch datasource sample my_dynamodb_ds --wait
+        tb --branch=my_branch datasource sample my_dynamodb_ds --rows 100000 --wait
+        tb --branch=my_branch datasource sample my_dynamodb_ds --max-bytes 1GB --wait
+        tb --branch=my_branch datasource sample my_dynamodb_ds --full-export --wait
     """
     from tinybird.tb.modules.common import wait_job
     from tinybird.tb.modules.job_common import echo_job_url
@@ -884,10 +914,32 @@ def datasource_sample(ctx: Context, datasource_name: str, max_files: int, wait:
         client: TinyB = ctx.obj["client"]
         config = ctx.obj.get("config", {})
 
+        if rows is not None and max_bytes is not None:
+            raise CLIDatasourceException(
+                FeedbackManager.error(message="--rows and --max-bytes are mutually exclusive; pass only one.")
+            )
+
+        if full_export and (rows is not None or max_bytes is not None):
+            raise CLIDatasourceException(
+                FeedbackManager.error(message="--full-export cannot be combined with --rows or --max-bytes.")
+            )
+
+        if full_export:
+            click.echo(
+                FeedbackManager.warning(
+                    message=(
+                        "DynamoDB full export samples will import the whole table. "
+                        "Use --rows or --max-bytes without --full-export for a bounded sample."
+                    )
+                )
+            )
+
         click.echo(FeedbackManager.info(message=f"Starting sample import for {datasource_name}..."))
 
         # Start the job
-        result = client.datasource_sample(datasource_name, max_files=max_files)
+        result = client.datasource_sample(
+            datasource_name, max_files=max_files, rows=rows, max_bytes=max_bytes, full_export=full_export
+        )
 
         job_id = result.get("job_id") or result.get("id")
         if not job_id:

{tinybird-4.6.11.dev0 → tinybird-4.6.12}/tinybird.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: tinybird
-Version: 4.6.11.dev0
+Version: 4.6.12
 Summary: Tinybird Command Line Tool
 Home-page: https://www.tinybird.co/docs/forward/commands
 Author: Tinybird
@@ -52,6 +52,18 @@ The Tinybird command-line tool allows you to use all the Tinybird functionality
 Changelog
 ----------
 
+4.6.12
+********
+
+- `Added` Forward sink pipes support ``{{ tb_secret(...) }}`` in ``EXPORT_*`` metadata (for example ``EXPORT_BUCKET_URI``). Secrets are resolved at deploy time and baked into the sink settings; changing a secret requires a redeploy.
+- `Fixed` Forward deploy now raises an error when a secret used in sink metadata resolves to an empty value.
+- `Fixed` Forward deploy now rejects ``{{ tb_secret(...) }}`` in ``EXPORT_SCHEDULE`` and ``COPY_SCHEDULE``. Use a literal cron expression (for example ``0 * * * *``) or ``@on-demand``.
+
+4.6.11
+********
+
+- `Added` `tb datasource sample` support for DynamoDB connectors.
+
 4.6.10
 ********