ducklake 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 451f6059faf0e7c451599b218bb9dafbbb8b80a2b54416833a5dafb85a0d1a9d
-  data.tar.gz: a1c6c7825da3dd5a077705c1c3f106a0cc1d75bc2e60b17ccb5099f98a3bdf1a
+  metadata.gz: 9366806f0dee7b0277dfee65be6c4718c339b6e9ab53e595e1f88534d5144534
+  data.tar.gz: 3b3c01d575bb8d0c380b49cbbc8b6deb62cc7de5542449dc73d0fd3b5ccffb1f
 SHA512:
-  metadata.gz: 9359cf3aaf06ed4398b9bc0ff35489c0e50008cd8d3426a787339d33f541a6bce8e3a1eca4a2b88b46852545895b591d08613f88ad26902aa4d41cf8b842bc93
-  data.tar.gz: 5bd9818d1b928aa2c0a11301c3455b89f8a4fc99912c6d1b8e52fdd151088adcc331b0ca01169df08412e151841f2f6b50f4c7b85b909ca558d6c21f0ffebd56
+  metadata.gz: 2e1fb28d47b8efeedaedba54727d678bdf946efd6aac1001ae0a2ad9aa9fbdc5b28bf8df06b2a10be13f6684195e2f684eafe71c7bedb27b36eff7af36d65abf
+  data.tar.gz: 4feea1505445b4347e722ea31d87c6dd9ed759489550f7a69997fced7b9a45827b4beaa81c4bc5e12abb801226af82f1c4903aff5ab6fd4b0851b6772b2c83c3

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.1.2 (2025-08-23)
+
+- Added `transaction` method
+
 ## 0.1.1 (2025-08-18)
 
 - Added experimental support for Polars

data/README.md

@@ -163,12 +163,46 @@ Delete data
 ducklake.sql("DELETE * FROM events WHERE id = ?", [1])
 ```
 
+Run multiple statements in a transaction
+
+```ruby
+ducklake.transaction do
+  # ...
+end
+```
+
+Raise `DuckLake::Rollback` to rollback
+
+## Schema Changes
+
 Update the schema
 
 ```ruby
 ducklake.sql("ALTER TABLE events ADD COLUMN active BOOLEAN")
 ```
 
+Set or remove a [partitioning key](https://ducklake.select/docs/stable/duckdb/advanced_features/partitioning)
+
+```ruby
+ducklake.sql("ALTER TABLE events SET PARTITIONED BY (name)")
+# or
+ducklake.sql("ALTER TABLE events RESET PARTITIONED BY")
+```
+
+## Views
+
+Create a view
+
+```ruby
+ducklake.sql("CREATE VIEW events_view AS SELECT * FROM events")
+```
+
+Drop a view
+
+```ruby
+ducklake.sql("DROP VIEW events_view")
+```
+
 ## Snapshots
 
 Get snapshots
@@ -235,9 +269,9 @@ ducklake.set_option("parquet_compression", "zstd", table_name: "events")
 
 ## Read-Only Mode
 
-Note: This feature is experimental
+Note: This feature is experimental and does not prevent the DuckDB engine from writing files via `sql`
 
-Connect to the data lake in read-only mode
+Attach the catalog in read-only mode
 
 ```ruby
 DuckLake::Client.new(read_only: true, ...)
@@ -294,7 +328,7 @@ ducklake.sql("COPY #{quoted_table} FROM #{quoted_file}")
 
 ## Polars
 
-Note: This feature is experimental and does not work on tables with schema changes
+Note: This feature is experimental and does not currently work on tables with schema changes
 
 Query the data with [Ruby Polars](https://github.com/ankane/ruby-polars)
 

data/lib/ducklake/client.rb

@@ -106,6 +106,17 @@ module DuckLake
       execute(sql, params)
     end
 
+    def transaction
+      execute("BEGIN")
+      begin
+        yield
+        execute("COMMIT")
+      rescue => e
+        execute("ROLLBACK")
+        raise e unless e.is_a?(Rollback)
+      end
+    end
+
     def attach(alias_, url)
       type = nil
       extension = nil
@@ -151,6 +162,15 @@ module DuckLake
       symbolize_keys result
     end
 
+    # experimental
+    # TODO use keyword arguments or range?
+    def table_changes(table, start_snapshot, end_snapshot)
+      params = [@catalog, "main", table, start_snapshot, end_snapshot]
+      result = execute("SELECT * FROM ducklake_table_changes(?, ?, ?, ?, ?)", params)
+      # only return changes between snapshots
+      symbolize_keys result.reject { |v| v["snapshot_id"] == start_snapshot }
+    end
+
     # TODO more DDL methods?
     def drop_table(table, if_exists: nil)
       execute("DROP TABLE#{" IF EXISTS" if if_exists} #{quote_identifier(table)}")
@@ -289,15 +309,26 @@ module DuckLake
     end
 
     # experimental
-    # TODO support schema changes
     def polars(table, snapshot_version: nil, snapshot_time: nil)
       files = list_files(table, snapshot_version:, snapshot_time:)
       sources = files.map { |v| v[:data_file] }
+      # TODO support schema changes
+      # column_mapping = [
+      #   "iceberg-column-mapping",
+      #   nil
+      # ]
       deletion_files = [
         "iceberg-position-delete",
         files.map.with_index.select { |v, i| v[:delete_file] }.to_h { |v, i| [i, [v[:delete_file]]] }
       ]
-      Polars.scan_parquet(sources, _deletion_files: deletion_files, storage_options: polars_storage_options)
+      Polars.scan_parquet(
+        sources,
+        storage_options: polars_storage_options,
+        # allow_missing_columns: true,
+        # extra_columns: "ignore",
+        # _column_mapping: column_mapping,
+        _deletion_files: deletion_files
+      )
     end
 
     # libduckdb does not provide function
@@ -370,7 +401,9 @@ module DuckLake
         "Conversion Error: " => ConversionError,
         "Invalid Input Error: " => InvalidInputError,
         "IO Error: " => IOError,
-        "Permission Error: " => PermissionError
+        "Not implemented Error: " => NotImplementedError,
+        "Permission Error: " => PermissionError,
+        "TransactionContext Error: " => TransactionContextError
       }
     end
 

data/lib/ducklake/version.rb

@@ -1,3 +1,3 @@
 module DuckLake
-  VERSION = "0.1.1"
+  VERSION = "0.1.2"
 end

data/lib/ducklake.rb

@@ -15,5 +15,8 @@ module DuckLake
   class ConversionError < Error; end
   class InvalidInputError < Error; end
   class IOError < Error; end
+  class NotImplementedError < Error; end
   class PermissionError < Error; end
+  class Rollback < Error; end
+  class TransactionContextError < Error; end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ducklake
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - Andrew Kane