Feature Request: Comprehensive Metadata Discovery for Data Development Tools

[mullinsms]

Summary

The ADBC Snowflake driver currently supports discovery of databases, schemas, tables, views, columns, and key constraints. However, Snowflake has a rich object model that data development tools need to surface — functions, procedures, stages, tasks, warehouses, and many more. This issue requests expanding metadata discovery to cover the full breadth of Snowflake objects.

We build a collaborative data operations platform and currently use the Snowflake JDBC driver supplemented by SHOW and DESCRIBE commands to extract metadata. We are evaluating the ADBC driver as a modern Arrow-native replacement but cannot adopt it without coverage for the object categories listed below.

Each section describes what we need, why it matters, what metadata properties are required, and how the data can be obtained from Snowflake.

---

Table of Contents

1. Functions (Scalar & Table-Valued)
2. Function & Procedure Parameters
3. System / Built-in Functions
4. Stored Procedures
5. Sequences
6. Stages (Internal & External)
7. File Formats
8. Tasks
9. Streams
10. Pipes
11. Warehouses
12. Users & Roles
13. Shares (Inbound & Outbound)
14. Storage Integrations
15. View Definitions (SQL Text)
16. Extended Table Types
17. Check Constraints
18. Table Statistics

---

1. Functions (Scalar & Table-Valued)

Motivation

User-defined functions are first-class schema objects in Snowflake. Data development tools need to enumerate them alongside tables and views for autocomplete, documentation, and catalog browsing. Snowflake supports function overloading (same name, different signatures), which requires careful handling.

Required Metadata

Property	Description	Source Column
Database	Containing database	catalog_name
Schema	Containing schema	schema_name
Function name	Function identifier	name from SHOW / split from arguments
Is table function	Scalar vs table-valued	is_table_function
Language	SQL, JavaScript, Python, Java, Scala	language
Return type	Return data type (scalar) or result columns (table)	Parsed from arguments / DESCRIBE output
Volatility	VOLATILE, IMMUTABLE, STABLE	is_memoizable and DESCRIBE output
Comment	User-provided description	description
Created time	Creation timestamp	created_on
Is secure	Whether the function is secure	is_secure
Is external	Whether it's an external function	is_external_function
Arguments signature	Full type signature for overload resolution	arguments

How to Obtain

-- List all user functions in a schema
SHOW USER FUNCTIONS IN SCHEMA "db"."schema";

-- List all user functions in a schema (single function)
SHOW USER FUNCTIONS LIKE 'func_name' IN SCHEMA "db"."schema";

-- Get detailed metadata including parameter names, body, language
DESCRIBE FUNCTION "db"."schema"."func_name"(VARCHAR, NUMBER);

SHOW USER FUNCTIONS returns: created_on, name, schema_name, is_builtin, is_aggregate, is_ansi, min_num_arguments, max_num_arguments, arguments, description, catalog_name, is_table_function, valid_for_clustering, is_secure, is_external_function, language, is_memoizable.

The arguments column contains the full signature, e.g. MY_FUNC(VARCHAR, NUMBER) RETURN VARCHAR. This must be parsed to separate input types from return type and to construct the signature needed for DESCRIBE FUNCTION.

DESCRIBE FUNCTION returns rows with property and value columns including: signature, returns, language, body, null handling, volatility, is_external, external_access_integrations, secrets, imports, handler, packages.

Overload Handling

Multiple functions can share the same name with different parameter lists. The driver should:

1. Group results from SHOW USER FUNCTIONS by name
2. Expose each overload as a separate entry (or as children of a parent function)
3. Include the full type signature so callers can distinguish overloads

---

2. Function & Procedure Parameters

Motivation

Parameter metadata enables signature help, call template generation, and documentation. Without parameter names and types, tools can only show the function name.

Required Metadata

Property	Description
Parameter name	e.g., input_date, threshold
Ordinal position	1-based position
Data type	Snowflake type name
Parameter mode	IN, OUT, INOUT (procedures only)
Default value	Default expression if optional
Comment	Parameter description if available

For table functions, the result column schema:

Property	Description
Column name	Result column name
Data type	Result column type
Ordinal position	Position in result set

How to Obtain

-- Returns rows including 'signature' and 'returns' properties
DESCRIBE FUNCTION "db"."schema"."func_name"(VARCHAR, NUMBER);
DESCRIBE PROCEDURE "db"."schema"."proc_name"(VARCHAR);

The signature property contains parameter names and types: (input_date VARCHAR, threshold NUMBER).
The returns property contains the return type (scalar) or TABLE(col1 TYPE, col2 TYPE) (table function).

---

3. System / Built-in Functions

Motivation

SQL editors provide autocomplete and inline documentation for built-in functions like COALESCE, DATE_TRUNC, ARRAY_AGG, etc. A hardcoded list drifts out of date across Snowflake releases.

Required Metadata

Property	Description	Source Column
Function name	e.g., ABS, CONCAT	Parsed from arguments
Category	Aggregate, string, date, etc.	Not directly available
Description	What the function does	description
Arguments signature	Input types	arguments
Min/max arguments	Arity range	min_num_arguments, max_num_arguments
Is aggregate	Whether it's an aggregate function	is_aggregate

How to Obtain

-- Returns all built-in functions (no schema qualifier)
SHOW FUNCTIONS;

Returns the same columns as SHOW USER FUNCTIONS but for system-provided functions. Filter where is_builtin = 'Y'.

---

4. Stored Procedures

Motivation

Stored procedures are a core part of Snowflake's programmability model. They support SQL, JavaScript, Python, Java, and Scala. Like functions, they support overloading.

Required Metadata

Property	Description	Source Column
Database	Containing database	catalog_name
Schema	Containing schema	schema_name
Procedure name	Procedure identifier	Parsed from arguments
Language	SQL, JavaScript, Python, Java, Scala	language (from DESCRIBE)
Return type	Return data type	Parsed from arguments
Comment	Description	description
Created time	Creation timestamp	created_on
Arguments signature	Full type signature	arguments
Execute as	OWNER or CALLER	execute as

How to Obtain

SHOW PROCEDURES IN SCHEMA "db"."schema";
SHOW PROCEDURES LIKE 'proc_name' IN SCHEMA "db"."schema";
DESCRIBE PROCEDURE "db"."schema"."proc_name"(VARCHAR, NUMBER);

SHOW PROCEDURES returns: created_on, name, schema_name, is_builtin, is_aggregate, is_ansi, min_num_arguments, max_num_arguments, arguments, description, catalog_name, is_table_function, valid_for_clustering, is_secure.

DESCRIBE PROCEDURE returns: signature, returns, language, execute as, body, null handling, volatility, external_access_integrations, secrets, imports, handler, packages.

---

5. Sequences

Motivation

Sequences are used for generating unique identifiers. Tools need to display them in the schema browser and show their current state.

Required Metadata

Property	Description	Source Column
Database	Containing database	database_name
Schema	Containing schema	schema_name
Sequence name	Sequence identifier	name
Next value	Next value to be issued	next_value
Interval	Increment step	interval
Comment	Description	comment
Owner	Owning role	owner
Created time	Creation timestamp	created_on

How to Obtain

SHOW SEQUENCES IN SCHEMA "db"."schema";

Returns: created_on, name, schema_name, database_name, next_value, interval, owner, owner_role_type, comment.

---

6. Stages (Internal & External)

Motivation

Stages are the primary mechanism for loading and unloading data in Snowflake. Tools that support data pipeline authoring or data loading workflows need to enumerate stages and understand their configuration.

Required Metadata — Internal Stages

Property	Description	Source Column
Stage name	Stage identifier	name
Region	Storage region	region
Comment	Description	comment
Owner	Owning role	owner
Type	INTERNAL	type
Created time	Creation timestamp	created_on

Required Metadata — External Stages

Property	Description	Source Column
Stage name	Stage identifier	name
URL	External location URI	url
Cloud	AWS, Azure, GCP	cloud
Region	Storage region	region
Storage integration	Associated integration name	storage_integration
Has credentials	Whether credentials are configured	has_credentials
Has encryption key	Whether encryption is configured	has_encryption_key
Comment	Description	comment
Owner	Owning role	owner
Type	EXTERNAL	type
Created time	Creation timestamp	created_on

How to Obtain

SHOW STAGES IN SCHEMA "db"."schema";

Returns: created_on, name, database_name, schema_name, url, has_credentials, has_encryption_key, owner, comment, region, type (INTERNAL/EXTERNAL), cloud, storage_integration, owner_role_type, directory_enabled.

Differentiate internal vs external by the type column.

---

7. File Formats

Motivation

File formats define how staged data is parsed during COPY operations. They are referenced by name in COPY INTO statements and need to be browsable.

Required Metadata

Property	Description	Source Column
Format name	File format identifier	name
Format type	CSV, JSON, PARQUET, AVRO, ORC, XML	type
Comment	Description	comment
Owner	Owning role	owner
Created time	Creation timestamp	created_on

How to Obtain

SHOW FILE FORMATS IN SCHEMA "db"."schema";

Returns: created_on, name, database_name, schema_name, type, owner, comment, owner_role_type.

---

8. Tasks

Motivation

Tasks are Snowflake's native scheduling mechanism for SQL and stored procedure execution. Tools need to display task definitions, schedules, states, and dependency chains (DAGs).

Required Metadata

Property	Description	Source Column
Task name	Task identifier	name
Schedule	Cron or interval expression	schedule
State	STARTED, SUSPENDED, etc.	state
Definition	The SQL statement executed	definition
Condition	WHEN expression (optional)	condition
Predecessors	Parent tasks in DAG	predecessors
Warehouse	Execution warehouse	warehouse
Comment	Description	comment
Owner	Owning role	owner
Created time	Creation timestamp	created_on

How to Obtain

SHOW TASKS IN SCHEMA "db"."schema";

Returns: created_on, name, id, database_name, schema_name, owner, comment, warehouse, schedule, predecessors, state, definition, condition, allow_overlapping_execution, error_integration, last_committed_on, last_suspended_on, owner_role_type, config, budget, last_suspended_reason.

---

9. Streams

Motivation

Streams track DML changes (inserts, updates, deletes) on tables, providing change data capture (CDC) functionality. They are central to Snowflake's continuous data pipeline patterns and are referenced in task definitions.

Required Metadata

Property	Description	Source Column
Stream name	Stream identifier	name
Source object	Table, view, or directory table being tracked	table_name
Source type	Type of source object	source_type
Stream type	STANDARD, APPEND_ONLY, INSERT_ONLY	type
Mode	DEFAULT, APPEND_ONLY, INSERT_ONLY	mode
Is stale	Whether the stream has fallen behind	stale
Stale after	Timestamp when stream becomes stale	stale_after
Comment	Description	comment
Owner	Owning role	owner
Created time	Creation timestamp	created_on
Database	Containing database	database_name
Schema	Containing schema	schema_name
Base tables	Underlying base tables	base_tables

How to Obtain

SHOW STREAMS IN SCHEMA "db"."schema";

Returns: created_on, name, database_name, schema_name, owner, comment, table_name, source_type, base_tables, type, stale, mode, stale_after, invalid_reason, owner_role_type.

---

10. Pipes

Motivation

Pipes provide continuous, serverless data ingestion from stages into tables (Snowpipe). Tools need to display pipe configurations and monitor their status.

Required Metadata

Property	Description	Source Column
Pipe name	Pipe identifier	name
Definition	The COPY INTO statement	definition
Is auto-ingest	Whether auto-ingest is enabled	notification_channel presence
Notification channel	SQS/SNS/Event Grid endpoint	notification_channel
Comment	Description	comment
Owner	Owning role	owner
Created time	Creation timestamp	created_on
Integration	Notification integration name	integration
Pattern	File pattern filter	Embedded in definition
Database	Containing database	database_name
Schema	Containing schema	schema_name

How to Obtain

SHOW PIPES IN SCHEMA "db"."schema";

Returns: created_on, name, database_name, schema_name, definition, owner, notification_channel, comment, integration, pattern, error_integration, owner_role_type, invalid_reason, budget.

---

11. Warehouses

Motivation

Warehouses are Snowflake's compute resources. Tools that manage connections or display environment context need to enumerate available warehouses and their state.

Required Metadata

Property	Description	Source Column
Warehouse name	Warehouse identifier	name
Type	STANDARD, SNOWPARK_OPTIMIZED	type
State	STARTED, SUSPENDED, RESIZING	state
Size	XS, S, M, L, XL, etc.	size
Auto-suspend (seconds)	Idle timeout before suspend	auto_suspend
Auto-resume	Whether auto-resume is enabled	auto_resume
Comment	Description	comment
Owner	Owning role	owner
Running queries	Active query count	running
Queued queries	Queued query count	queued
Is default	Whether it's the session default	is_default
Created time	Creation timestamp	created_on
Resumed time	Last resume timestamp	resumed_on
Updated time	Last config change	updated_on
Resource monitor	Associated resource monitor	resource_monitor

How to Obtain

SHOW WAREHOUSES;

Returns: name, state, type, size, min_cluster_count, max_cluster_count, started_clusters, running, queued, is_default, is_current, auto_suspend, auto_resume, available, provisioning, quiescing, other, created_on, resumed_on, updated_on, owner, comment, enable_query_acceleration, query_acceleration_max_scale_factor, resource_monitor, owner_role_type, budget.

---

12. Users & Roles

Motivation

Tools that manage access, display connection context, or support administrative workflows need to discover users and roles.

Required Metadata — Users

Property	Description	Source Column
User name	User identifier	name
Email	User email address	email
Default role	Default role at login	default_role
Default warehouse	Default warehouse at login	default_warehouse
Comment	Description	comment
Owner	Owning role	owner
Disabled	Whether the user is disabled	disabled
Locked until	Lock expiration timestamp	locked_until_time
Created time	Creation timestamp	created_on

Required Metadata — Roles

Property	Description	Source Column
Role name	Role identifier	name
Comment	Description	comment
Owner	Owning role	owner
Is default	Whether it's the user's default role	is_default
Created time	Creation timestamp	created_on
Granted roles	Number of granted child roles	granted_roles
Granted to roles	Number of parent roles	granted_to_roles

How to Obtain

SHOW USERS;
SHOW ROLES;
SELECT CURRENT_AVAILABLE_ROLES();  -- roles accessible to current user

---

13. Shares (Inbound & Outbound)

Motivation

Snowflake Secure Data Sharing is a core platform feature. Tools need to display both outbound shares (data this account provides) and inbound shares (data consumed from other accounts).

Required Metadata

Property	Description	Source Column
Share name	Share identifier	name
Kind	OUTBOUND or INBOUND	kind
Database name	Database associated with the share	database_name
Comment	Description	comment
Owner	Owning role (outbound only)	owner
Owner account	Account that owns the share	owner_account
To (consumers)	List of consumer accounts (outbound)	to
Created time	Creation timestamp	created_on

How to Obtain

SHOW SHARES;
SELECT current_organization_name() || '.' || current_account_name() AS current_account_name;

The kind column distinguishes OUTBOUND (this account shares data) from INBOUND (this account consumes shared data).

---

14. Storage Integrations

Motivation

Storage integrations configure secure access to external cloud storage (S3, GCS, Azure Blob). They are referenced by external stages and external tables.

Required Metadata

Property	Description	Source Column
Integration name	Integration identifier	name
Type	EXTERNAL_STAGE	type
Enabled	Whether the integration is active	enabled
Comment	Description	comment
Created time	Creation timestamp	created_on

How to Obtain

SHOW STORAGE INTEGRATIONS;

Returns: name, type, category, enabled, comment, created_on.

---

15. View Definitions (SQL Text)

Motivation

Users inspect, copy, and refactor view definitions. View SQL is essential for understanding data lineage, migration between environments, and documentation.

Required Metadata

Property	Description
View database	Containing database
View schema	Containing schema
View name	View identifier
View definition	The SELECT statement defining the view
Is secure	Whether the view is secure (definition hidden from non-owners)
Is materialized	Whether it's a materialized view

How to Obtain

-- Via INFORMATION_SCHEMA
SELECT TABLE_CATALOG, TABLE_SCHEMA, TABLE_NAME, VIEW_DEFINITION,
       CHECK_OPTION, IS_UPDATABLE, IS_SECURE, IS_MATERIALIZED
FROM "db".INFORMATION_SCHEMA.VIEWS
WHERE TABLE_SCHEMA = 'schema_name';

-- Via DDL generation
SELECT GET_DDL('VIEW', '"db"."schema"."view_name"');

Note: Secure view definitions are only visible to the view owner role.

---

16. Extended Table Types

Motivation

The driver currently returns only "TABLE" and "VIEW" from GetTableTypes(). Snowflake has many specialized table types that tools need to distinguish for correct display, icon rendering, and context-appropriate operations.

Required Types

Table Type	Description	How to Identify
EXTERNAL TABLE	Tables backed by external cloud storage files	SHOW EXTERNAL TABLES IN SCHEMA
MATERIALIZED VIEW	Precomputed views with automatic refresh	SHOW MATERIALIZED VIEWS IN SCHEMA
DYNAMIC TABLE	Declarative tables that auto-refresh from a query	SHOW DYNAMIC TABLES IN SCHEMA
ICEBERG TABLE	Apache Iceberg-format tables (managed or unmanaged)	SHOW ICEBERG TABLES IN SCHEMA
HYBRID TABLE	HTAP tables with row-level locking for transactional workloads	SHOW HYBRID TABLES IN SCHEMA
EVENT TABLE	Tables for logging and telemetry events	SHOW EVENT TABLES IN SCHEMA
TEMPORARY TABLE	Session-scoped temporary tables	is_temporary column in SHOW TABLES
TRANSIENT TABLE	Tables without Fail-safe (reduced storage cost)	retention_time = 0 in SHOW TABLES

Required Metadata by Type

External Tables

Property	Description	Source Column
Location	External storage path	location
File format type	CSV, JSON, Parquet, etc.	file_format_type
Cloud	AWS, Azure, GCP	cloud
Region	Storage region	region
Auto-refresh	Whether auto-refresh is enabled	auto_refresh
Notification channel	Notification endpoint for auto-refresh	notification_channel
Is invalid	Whether the external table is in error state	invalid
Invalid reason	Why the table is invalid	invalid_reason
Last refreshed	Last refresh timestamp	last_refreshed_on
Comment	Description	comment
Owner	Owning role	owner

SHOW EXTERNAL TABLES IN SCHEMA "db"."schema";
DESCRIBE EXTERNAL TABLE "db"."schema"."table" TYPE = COLUMNS;
DESCRIBE EXTERNAL TABLE "db"."schema"."table" TYPE = STAGE;

Materialized Views

Property	Description	Source Column
Name	View identifier	name
Definition	The defining SQL query	text
Is secure	Whether the view is secure	is_secure
Cluster by	Clustering columns	cluster_by
Rows	Approximate row count	rows
Bytes	Storage size	bytes
Refreshed on	Last refresh timestamp	refreshed_on
Behind by	How far behind source data	behind_by
Comment	Description	comment
Owner	Owning role	owner

SHOW MATERIALIZED VIEWS IN SCHEMA "db"."schema";

Dynamic Tables

Property	Description	Source Column
Name	Table identifier	name
Target lag	Desired freshness (e.g., '1 minute', 'DOWNSTREAM')	target_lag
Refresh mode	AUTO, FULL, INCREMENTAL	refresh_mode
Scheduling state	RUNNING, SUSPENDED	scheduling_state
Definition	The defining SQL query	text
Warehouse	Execution warehouse	warehouse
Data timestamp	Freshness of the data	data_timestamp
Last suspended on	Last suspension timestamp	last_suspended_on
Is iceberg	Whether backed by Iceberg format	is_iceberg
Comment	Description	comment
Owner	Owning role	owner

SHOW DYNAMIC TABLES IN SCHEMA "db"."schema";

Iceberg Tables

Property	Description	Source Column
Name	Table identifier	name
Is managed	Snowflake-managed vs externally managed	is_default
External volume	External volume name	external_volume_name
Catalog integration	Catalog integration for external Iceberg	catalog_integration_name
Catalog namespace	Iceberg catalog namespace	catalog_namespace
Catalog table name	Name in external catalog	catalog_table_name
Base location	Storage base location	base_location
Metadata location	Path to current metadata file	metadata_location
Comment	Description	comment
Owner	Owning role	owner

SHOW ICEBERG TABLES IN SCHEMA "db"."schema";

Hybrid Tables

Property	Description	Source Column
Name	Table identifier	name
Comment	Description	comment
Owner	Owning role	owner
Rows	Row count	rows
Bytes	Storage size	bytes

SHOW HYBRID TABLES IN SCHEMA "db"."schema";

Event Tables

Property	Description	Source Column
Name	Table identifier	name
Comment	Description	comment
Owner	Owning role	owner

SHOW EVENT TABLES IN SCHEMA "db"."schema";

---

17. Check Constraints

Motivation

Check constraints enforce data quality rules at the table level. Tools need to display them alongside primary/foreign/unique keys.

Required Metadata

Property	Description
Constraint name	Constraint identifier
Constraint type	CHECK
Expression	The boolean expression (e.g., price > 0)
Column names	Columns referenced by the expression

Current State

The GetObjects schema already includes table_constraints with support for a CHECK type, but it is not populated. Snowflake supports check constraints and lists them in INFORMATION_SCHEMA.TABLE_CONSTRAINTS.

How to Obtain

SELECT CONSTRAINT_NAME, CONSTRAINT_TYPE, TABLE_NAME, CHECK_CLAUSE
FROM "db".INFORMATION_SCHEMA.TABLE_CONSTRAINTS tc
LEFT JOIN "db".INFORMATION_SCHEMA.CHECK_CONSTRAINTS cc
  ON tc.CONSTRAINT_NAME = cc.CONSTRAINT_NAME
WHERE tc.CONSTRAINT_TYPE = 'CHECK'
  AND tc.TABLE_SCHEMA = 'schema_name';

Note: An extension to the ADBC CONSTRAINT_SCHEMA may be needed to carry the constraint_expression text, as the spec only defines constraint_column_names.

---

18. Table Statistics

Motivation

Row counts and storage sizes are fundamental metadata for data tools — displayed in catalog browsers, used for query planning, and needed for capacity monitoring.

Required Metadata

Statistic	ADBC Key	Description
Row count	StatisticRowCountKey (6)	Number of rows
Size in bytes	Custom key	Total storage size
Active bytes	Custom key	Active storage (before Time Travel)
Time travel bytes	Custom key	Time Travel storage
Failsafe bytes	Custom key	Fail-safe storage
Clustering depth	Custom key	Average clustering depth
Retention time (days)	Custom key	Time Travel retention period

How to Obtain

-- From INFORMATION_SCHEMA (fast, cached, may be stale up to ~30 mins)
SELECT TABLE_NAME, ROW_COUNT, BYTES, RETENTION_TIME,
       CLUSTERING_KEY, AUTO_CLUSTERING_ON
FROM "db".INFORMATION_SCHEMA.TABLES
WHERE TABLE_SCHEMA = 'schema_name';

-- From SHOW TABLES (returns live-ish counts)
SHOW TABLES IN SCHEMA "db"."schema";
-- Includes: rows, bytes, retention_time, automatic_clustering, cluster_by

ADBC Spec Alignment

The ADBC standard defines GetStatistics() (since v1.1.0) with a result schema including statistic_key, statistic_value, and statistic_is_approximate. Row count and size would map naturally to this API.

---

Summary Priority Table

#	Category	Object Count	Snowflake API	Impact
1	User Functions (scalar + table)	Per schema	SHOW USER FUNCTIONS + DESCRIBE FUNCTION	High — core SQL development
2	Function/Procedure Parameters	Per function	DESCRIBE FUNCTION/PROCEDURE	High — signature help
3	System Functions	Global	SHOW FUNCTIONS	High — autocomplete
4	Stored Procedures	Per schema	SHOW PROCEDURES + DESCRIBE PROCEDURE	High — core SQL development
5	Sequences	Per schema	SHOW SEQUENCES	Medium
6	Stages	Per schema	SHOW STAGES	Medium — data loading
7	File Formats	Per schema	SHOW FILE FORMATS	Medium — data loading
8	Tasks	Per schema	SHOW TASKS	Medium — pipeline tools
9	Streams	Per schema	SHOW STREAMS	Medium — CDC pipelines
10	Pipes	Per schema	SHOW PIPES	Medium — continuous ingest
11	Warehouses	Account-level	SHOW WAREHOUSES	Medium — admin tools
12	Users & Roles	Account-level	SHOW USERS / SHOW ROLES	Low — admin tools
13	Shares	Account-level	SHOW SHARES	Low — data sharing
14	Storage Integrations	Account-level	SHOW STORAGE INTEGRATIONS	Low — infra tools
15	View Definitions	Per view	INFORMATION_SCHEMA.VIEWS / GET_DDL()	High — SQL development
16	Extended Table Types	Per schema	Various SHOW commands	High — correct object display
17	Check Constraints	Per table	INFORMATION_SCHEMA.TABLE_CONSTRAINTS	Medium
18	Table Statistics	Per table	INFORMATION_SCHEMA.TABLES / SHOW TABLES	Medium

We're happy to provide further details on any of these categories.

[lidavidm]

@mullinsms Snowflake doesn't appear to support CHECK constraints, can you clarify point 17?

[lidavidm]

Stages, File Formats, Tasks, Streams, Pipes, Warehouses, Shares, and Storage Integrations feel rather Snowflake-specific, but as with BigQuery I'm willing to be convinced.

I'm curious why we care about users/roles for Snowflake but not for other systems?

Other requests have been turned into feature requests, see issues referencing this issue above.

[mullinsms]

@mullinsms Snowflake doesn't appear to support CHECK constraints, can you clarify point 17?

@lidavidm Snowflake supports, but doesn't enforce constraints on standard tables. It does however enforce them on hybrid tables. Here is a link to the Snowflake documentation. https://docs.snowflake.com/en/sql-reference/constraints

[mullinsms]

Stages, File Formats, Tasks, Streams, Pipes, Warehouses, Shares, and Storage Integrations feel rather Snowflake-specific, but as with BigQuery I'm willing to be convinced.

I'm curious why we care about users/roles for Snowflake but not for other systems?

Other requests have been turned into feature requests, see issues referencing this issue above.

@lidavidm I think I put this in the class of not expected, but really nice to have. If surfacing these is not implemented in the driver, then we'll have to implement them ourselves with metadata queries to populate the catalog and feed autocomplete. All of the cloud data platforms have these platform specific objects that we still have to display to users.