The functions on this page work on STRUCT values. Functions that access entries by position, such as struct[idx], struct_extract_at, struct_contains, struct_position, and struct_values, also accept TUPLE values, i.e., structs with unnamed entries. struct_concat accepts either only STRUCTs or only TUPLEs. Functions that work with entry names, i.e., struct_insert, struct_keys, struct_update, and remap_struct, require a STRUCT.

Name	Description
struct.entry	Dot notation that serves as an alias for struct_extract from named STRUCTs.
struct[entry]	Bracket notation that serves as an alias for struct_extract from named STRUCTs.
struct[idx]	Bracket notation that serves as an alias for struct_extract from TUPLEs, using an index (1-based).
remap_struct(input, target_type, mapping, defaults)	Remap a STRUCT, LIST, or MAP value into the shape given by the type of target_type, renaming and reordering fields, descending into nested types, and filling missing fields from defaults.
row(any, ...)	Create a TUPLE containing the argument values.
struct_concat(structs...)	Merge the multiple structs into a single STRUCT.
struct_contains(struct, entry)	Check if the STRUCT contains the specified entry.
struct_extract(struct, 'entry')	Extract the named entry from the STRUCT.
struct_extract(struct, idx)	Extract the entry from a TUPLE using an index (1-based).
struct_extract_at(struct, idx)	Extract the entry from a STRUCT or TUPLE using an index (1-based).
struct_insert(struct, name := any, ...)	Add field(s) to an existing STRUCT.
struct_keys(struct)	Return the names of the entries of a STRUCT as a list.
struct_pack(name := any, ...)	Create a STRUCT containing the argument values. The entry name will be the bound variable name.
struct_position(struct, entry)	Return the index of the entry within the STRUCT (1-based), or NULL if not found.
struct_update(struct, name := any, ...)	Add or update field(s) of an existing STRUCT.
struct_values(struct)	Return the values of a STRUCT as a TUPLE.

struct.entry

Description	Dot notation that serves as an alias for struct_extract from named STRUCTs.
Example	({'i': 3, 's': 'string'}).i
Result	3

struct[entry]

Description	Bracket notation that serves as an alias for struct_extract from named STRUCTs.
Example	({'i': 3, 's': 'string'})['i']
Result	3

struct[idx]

Description	Bracket notation that serves as an alias for struct_extract from TUPLEs, using an index (1-based).
Example	(row(42, 84))[1]
Result	42

remap_struct(input, target_type, mapping, defaults)

Description	Remap an input value of type STRUCT, LIST, or MAP into the shape declared by the type of target_type, renaming and reordering fields, descending into nested types, and filling target fields that have no source counterpart from defaults. Only the type of target_type is used, not its value, so any value of the desired type works; a typed NULL literal (e.g. NULL::STRUCT(v1 INT, v2 INT)) is a convenient way to declare the target shape when no suitable value is at hand. mapping is a STRUCT keyed by target field name whose values name the source field to read from; defaults is a STRUCT of constant values for the remaining target fields. Every target field must be covered by either mapping or defaults.
Example	remap_struct({'i': 1, 'j': 2}, NULL::STRUCT(v1 INT, v2 INT, v3 INT), {'v1': 'j', 'v3': 'i'}, {'v2': NULL::INTEGER})
Result	{'v1': 2, 'v2': NULL, 'v3': 1}

To remap a target field that is itself a STRUCT, give its mapping value as ROW(source_field, {inner_mapping}), and nest its defaults the same way. A LIST's single child is addressed by the mapping key 'list', and a MAP's children by the keys 'key' and 'value'.

row(any, ...)

Description	Create a TUPLE containing the argument values.
Example	row(i, i % 4, i / 4)
Result	(10, 2, 2.5)

struct_concat(structs...)

Description	Merge the multiple structs into a single STRUCT.
Example	struct_concat(struct_pack(i := 4), struct_pack(s := 'string'))
Result	{'i': 4, 's': string}

struct_contains(struct, entry)

Description	Check if the STRUCT contains the specified entry.
Example	struct_contains(row(1, 2, 3), 2)
Result	true
Alias	struct_has

struct_extract(struct, 'entry')

Description	Extract the named entry from the STRUCT.
Example	struct_extract({'i': 3, 'v2': 3, 'v3': 0}, 'i')
Result	3

struct_extract(struct, idx)

Description	Extract the entry from a TUPLE using an index (1-based).
Example	struct_extract(row(42, 84), 1)
Result	42

struct_extract_at(struct, idx)

Description	Extract the entry from a STRUCT or TUPLE using an index (1-based).
Example	struct_extract_at({'v1': 10, 'v2': 20, 'v3': 3}, 2)
Result	20

struct_insert(struct, name := any, ...)

Description	Add field(s) to an existing STRUCT.
Example	struct_insert({'a': 1}, b := 2)
Result	{'a': 1, 'b': 2}

struct_keys(struct)

Description	Return the names of the entries of a STRUCT as a list.
Example	struct_keys({'a': 1, 'b': 2})
Result	[a, b]

struct_pack(name := any, ...)

Description	Create a STRUCT containing the argument values. The entry name will be the bound variable name.
Example	struct_pack(i := 4, s := 'string')
Result	{'i': 4, 's': string}

struct_position(struct, entry)

Description	Return the index of the entry within the STRUCT (1-based), or NULL if not found.
Example	struct_position(row(1, 2, 3), 2)
Result	2
Alias	struct_indexof

struct_update(struct, name := any, ...)

Description	Add or update field(s) of an existing STRUCT.
Example	struct_update({'a': 1, 'b': 2}, b := 3, c := 4)
Result	{'a': 1, 'b': 3, 'c': 4}

struct_values(struct)

Description	Return the values of a STRUCT as a TUPLE.
Example	struct_values({'a': 1, 'b': 2, 'c': 3})
Result	(1, 2, 3)