Release highlights: 1.12.1

This update includes substantial improvements to developer experience and stability, as well as bug fixes. Here's a breakdown of what's new:

---

Placeholder config & secret warning system

PR: #2636

Overview

• dlt now detects placeholder values during config and secret resolution (e.g., "<configure me>", or timestamps like 1768-07-21T02:56:00Z).
• A detailed warning is emitted when placeholders are detected, helping users identify incomplete configuration.
• Prevents confusing errors during pipeline runs and catches misconfigured values early in the development cycle.

Example placeholders

[credentials]
token = "<configure me>"
started_at = "1768-07-21T02:56:73Z"  # random date

Example warning

Placeholder value encountered when resolving config or secret:
resolved_key: host, value: <configure me>, section: sources.sql_database.credentials.host
Most likely, this comes from init-command, which creates basic templates for non-complex configs and secrets. The provider to adjust is secrets.toml at one of these locations:
- /home/me/project/.dlt/secrets.toml
- /home/me/.dlt/secrets.toml

---

Warning for None-only columns

PR: #2633

Overview

• When normalizing data, if a column contains only None, dlt can’t infer its type.
• This PR adds a clear warning message to help users fix this by providing an explicit type hint.
• If no type hint is given, the column will be excluded from the destination table.

Example warning

The column `my_column` in table `my_table` did not receive any data during this load. Therefore, its type couldn't be inferred.

Unless a type hint is provided, the column will not be materialized in the destination.

Recommended fix

@dlt.resource(columns={"my_column": {"data_type": "text"}})
def my_data():
    yield {"my_column": None}

---

Regular and standalone resources are now the same thing

Before

• Inner resources could not receive secrets or configs properly.
• Required the standalone=True argument.

@dlt.resource(standalone=True)
def foo(access_token=dlt.secrets.value):
    yield from generate_json_like_data(access_token)

[sources.foo]
access_token = "your_token"

Now

• All @dlt.resource functions behave identically.
• standalone=True is ignored and no longer required.
• Inner resources adopt the config section from the source they are in: sources.source_module.resource_name this is canonical form.

Cleaner usage

@dlt.resource
def foo(access_token=dlt.secrets.value):
    yield from generate_json_like_data(access_token)

For inner resources:

@dlt.source
def bar():
    @dlt.resource
    def foo(access_token=dlt.secrets.value):
        yield from generate_json_like_data(access_token)

    return foo

[sources.bar.foo]
access_token = "your_token"

---

Resources can return data (instead of yield)

Overview

• You can now return values instead of yielding.
• Useful for static data or short lists.

Example

@dlt.resource
def numbers():
    return [{"key1": 1, "key2": 2}]

return is automatically wrapped into a generator internally.

warning

Not recommended for large-scale use due to readability and stream control.

---

Performance and stability fixes

PR: #2756

1. Snowflake: extremely slow merge query due to OR condition (n² complexity)

• In merge write disposition with delete-insert strategy, the generated SQL used:

DELETE FROM main_table WHERE EXISTS (
  SELECT 1 FROM staging_table
  WHERE main.pk1 = staging.pk1 AND main.pk2 = staging.pk2
     OR main.merge_key = staging.merge_key
);

• This caused very slow performance due to OR logic in EXISTS, especially on large datasets.

New approach

• Rewrites the logic to two separate DELETE queries, executed in the same transaction:

DELETE ... WHERE EXISTS (JOIN ON primary_key);
DELETE ... WHERE EXISTS (JOIN ON merge_key);

info

⚡ Improves performance from 18 hours → 4 seconds on large datasets.

2. Skip deduplication option

• New flag allows users to disable deduplication during delete-insert operations, further reducing merge cost.
• Useful for append-heavy workloads with upstream deduplication.
• Works for any destination that supports delete-insert.

@dlt.resource(
    primary_key="id",
    write_disposition={
        "disposition": "merge",
        "strategy": "delete-insert",
        "deduplicated": False
    }
)
def data_resource():
    yield from get_data()

---

Iceberg: upsert merge strategy added

PR: #2671

Overview

• Iceberg now supports merge with upsert semantics, similar to Delta Lake.
• Enables row-level updates using primary and merge keys.

warning

Known limitations due to current pyiceberg behavior:

• Nested fields and struct joins are not fully supported in Arrow joins (required by upsert).
• Non-unique keys in input data will raise hard errors — Iceberg enforces strict uniqueness.
• Some failing tests stem from current pyiceberg limitations (e.g., recursion limits, Arrow type mismatches).

---

Bonus: Postgres now supports parquet via ADBC

PR: #2685

pipeline = dlt.pipeline(destination="postgres")
pipeline.run(some_source(), loader_file_format="parquet")

---

Full release notes

View the complete list of changes →