Parameters Reference¶

Complete reference for all configurable parameters.

read_yaml Parameters¶

Input Parameters¶

Parameter	Type	Default	Description
`path`	VARCHAR or VARCHAR[]	Required	File path, glob pattern, or list of paths

Schema Detection¶

Parameter	Type	Default	Description
`auto_detect`	BOOLEAN	`true`	Enable automatic type detection
`columns`	STRUCT	-	Explicit column types
`sample_size`	INTEGER	`20480`	Rows to sample for schema detection
`maximum_sample_files`	INTEGER	`32`	Files to sample for schema detection

Data Extraction¶

Parameter	Type	Default	Description
`records`	VARCHAR	-	Path to nested array of records (dot notation)

Parsing Behavior¶

Parameter	Type	Default	Description
`multi_document`	BOOLEAN	`true`	Handle multiple YAML documents
`expand_root_sequence`	BOOLEAN	`true`	Expand sequences into rows
`ignore_errors`	BOOLEAN	`false`	Continue on errors
`maximum_object_size`	INTEGER	`16777216`	Max file size (16MB)

auto_detect¶

Controls automatic type detection.

Values:

true (default): Detect types (DATE, TIME, TIMESTAMP, numeric types, BOOLEAN)
false: All columns become VARCHAR

Example:

-- Enable (default)
SELECT * FROM read_yaml('data.yaml', auto_detect = true);

-- Disable
SELECT * FROM read_yaml('data.yaml', auto_detect = false);

columns¶

Specifies explicit column types.

Value: Struct mapping column names to type strings.

Supported Types:

Basic: VARCHAR, INTEGER, BIGINT, DOUBLE, BOOLEAN, DATE, TIMESTAMP
Complex: STRUCT(...), VARCHAR[], INTEGER[], MAP(...), JSON, YAML

Example:

SELECT * FROM read_yaml('data.yaml', columns = {
    'id': 'BIGINT',
    'name': 'VARCHAR',
    'created': 'TIMESTAMP',
    'tags': 'VARCHAR[]',
    'metadata': 'JSON'
});

Interaction with auto_detect:

Specified columns use explicit types
Unspecified columns use auto_detect behavior

records¶

Extract records from a nested array path using dot notation.

Value: String with dot-separated path segments.

Example:

# config.yaml
metadata:
  version: "1.0"
projects:
  - name: alpha
    status: active
  - name: beta
    status: inactive

-- Extract the nested projects array
SELECT * FROM read_yaml('config.yaml', records = 'projects');

Nested paths:

# data.yaml
config:
  database:
    tables:
      - name: users
        rows: 1000
      - name: orders
        rows: 5000

-- Navigate multiple levels deep
SELECT * FROM read_yaml('data.yaml', records = 'config.database.tables');

Error handling:

Condition	Behavior
Path not found	Error (or skip with `ignore_errors = true`)
Path points to non-array	Error (or skip with `ignore_errors = true`)
Array contains non-objects	Non-object elements are filtered out

Interaction with other parameters:

expand_root_sequence is automatically disabled when records is specified
columns can be combined to specify types for the extracted records
ignore_errors allows graceful handling of missing paths

sample_size¶

Number of rows to sample for schema detection.

Values:

Positive integer: Sample this many rows
-1: Unlimited (sample all rows)

Default: 20480

Example:

-- Sample more rows for complex schemas
SELECT * FROM read_yaml('data/*.yaml', sample_size = 50000);

-- Sample all rows
SELECT * FROM read_yaml('data/*.yaml', sample_size = -1);

maximum_sample_files¶

Maximum files to sample for schema detection.

Values:

Positive integer: Sample up to this many files
-1: Unlimited (sample all files)

Default: 32

Example:

-- Sample more files
SELECT * FROM read_yaml('data/*.yaml', maximum_sample_files = 100);

-- Sample all files
SELECT * FROM read_yaml('data/*.yaml', maximum_sample_files = -1);

multi_document¶

Handle multiple YAML documents in a single file.

Values:

true (default): Each document becomes a separate row
false: Only read the first document

Example:

# File with multiple documents
---
name: Doc1
---
name: Doc2

-- Returns 2 rows (default)
SELECT * FROM read_yaml('multi.yaml', multi_document = true);

-- Returns 1 row
SELECT * FROM read_yaml('multi.yaml', multi_document = false);

expand_root_sequence¶

Expand top-level YAML sequences into rows.

Values:

true (default): Each sequence item becomes a row
false: Keep sequence as single value

Example:

# Sequence file
- name: Alice
- name: Bob

-- Returns 2 rows (default)
SELECT * FROM read_yaml('list.yaml', expand_root_sequence = true);

-- Returns 1 row with array
SELECT * FROM read_yaml('list.yaml', expand_root_sequence = false);

ignore_errors¶

Continue processing when errors occur.

Values:

false (default): Stop on first error
true: Skip invalid files/documents

Example:

-- Skip bad files
SELECT * FROM read_yaml('data/*.yaml', ignore_errors = true);

Behavior:

Error Type	ignore_errors=true
File not found	Skip file
Permission denied	Skip file
Invalid YAML	Skip document
Type conversion	Value becomes NULL

maximum_object_size¶

Maximum allowed file size in bytes.

Default: 16777216 (16MB)

Example:

-- Allow larger files (64MB)
SELECT * FROM read_yaml('large.yaml', maximum_object_size = 67108864);

read_yaml_frontmatter Parameters¶

Input Parameters¶

Parameter	Type	Default	Description
`path`	VARCHAR or VARCHAR[]	Required	File path or glob pattern

Output Control¶

Parameter	Type	Default	Description
`as_yaml_objects`	BOOLEAN	`false`	Return raw YAML column
`content`	BOOLEAN	`false`	Include file body
`filename`	BOOLEAN	`false`	Include source filename

as_yaml_objects¶

Return raw YAML instead of expanded columns.

Values:

false (default): Expand frontmatter fields as columns
true: Return single frontmatter column of YAML type

Example:

-- Expanded columns (default)
SELECT title, author, date FROM read_yaml_frontmatter('posts/*.md');

-- Raw YAML
SELECT frontmatter FROM read_yaml_frontmatter('posts/*.md',
    as_yaml_objects = true);

content¶

Include file content after frontmatter.

Values:

false (default): Exclude body content
true: Include as content column

Example:

SELECT title, content FROM read_yaml_frontmatter('posts/*.md',
    content = true);

filename¶

Include source file path.

Values:

false (default): Exclude filename
true: Include as filename column

Example:

SELECT filename, title FROM read_yaml_frontmatter('posts/*.md',
    filename = true);

COPY TO Parameters¶

FORMAT¶

Value	Description
`yaml`	YAML output format

STYLE¶

Value	Description
`block` (default)	Multi-line indented format
`flow`	Compact inline format

LAYOUT¶

Value	Description
`document` (default)	Each row as YAML document
`sequence`	All rows as sequence items

MULTILINE¶

Value	Description
`auto` (default)	Resolves to `literal` for block style, `quoted` for flow
`literal`	Use YAML literal block scalars (`\\|`) for multiline strings
`quoted`	Use quoted strings with escape sequences

INDENT¶

Value	Description
`1`-`10` (default: `2`)	Indentation width for block style output

Example:

COPY (SELECT * FROM data) TO 'output.yaml'
(FORMAT yaml, STYLE block, LAYOUT sequence, MULTILINE literal, INDENT 4);