Struct IndexFn 

pub struct IndexFn;

Trait Implementations

impl Debug for IndexFn

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Function for IndexFn

Returns the value or reference at a 1-based row and column within an array or range.

INDEX can operate on both references and array literals. When the first argument is a reference, this implementation resolves a referenced cell and materializes its value in value context.

Remarks

• Indexing is 1-based for both row_num and column_num.
• If column_num is omitted for a 1D array, row_num selects the position in that vector.
• row_num <= 0, column_num <= 0, or out-of-bounds indexes return #REF!.
• Non-numeric index arguments return #VALUE!.

Examples

title: "Pick a value from a 2D table"
grid:
  A1: "Item"
  B1: "Price"
  A2: "Pen"
  B2: 2.5
  A3: "Book"
  B3: 8
formula: '=INDEX(A1:B3,3,2)'
expected: 8

title: "Index into a 1D vector"
grid:
  A1: "Q1"
  A2: "Q2"
  A3: "Q3"
formula: '=INDEX(A1:A3,2)'
expected: "Q2"

related:
  - MATCH
  - XLOOKUP
  - OFFSET
faq:
  - q: "How does INDEX behave when column_num is omitted?"
    a: "For 1D arrays, row_num selects the position along that vector; for 2D arrays, omitted column_num defaults to the first column."
  - q: "Which errors indicate bad indexes?"
    a: "Non-numeric index arguments return #VALUE!, while 0/negative or out-of-bounds indexes return #REF!."

[formualizer-docgen:schema:start] Name: INDEX Type: IndexFn Min args: 2 Max args: 3 Variadic: false Signature: INDEX(arg1: any@range, arg2: number@scalar, arg3?: number@scalar) Arg schema: arg1{kinds=any,required=true,shape=range,by_ref=false,coercion=None,max=None,repeating=None,default=false}; arg2{kinds=number,required=true,shape=scalar,by_ref=false,coercion=NumberStrict,max=None,repeating=None,default=false}; arg3{kinds=number,required=false,shape=scalar,by_ref=false,coercion=NumberStrict,max=None,repeating=None,default=false} Caps: PURE, RETURNS_REFERENCE [formualizer-docgen:schema:end]

fn caps(&self) -> FnCaps

Capability flags for this function

fn name(&self) -> &'static str

fn min_args(&self) -> usize

fn arg_schema(&self) -> &'static [ArgSchema]

fn eval_reference<'a, 'b, 'c>( &self, args: &'c [ArgumentHandle<'a, 'b>], _ctx: &dyn FunctionContext<'b>, ) -> Option<Result<ReferenceType, ExcelError>>

Optional reference result path. Only called by the interpreter/engine when the callsite expects a reference (e.g., range combinators, by-ref argument positions, or spill sources).

fn eval<'a, 'b, 'c>( &self, args: &'c [ArgumentHandle<'a, 'b>], ctx: &dyn FunctionContext<'b>, ) -> Result<CalcValue<'b>, ExcelError>

The unified evaluation path.

fn namespace(&self) -> &'static str

fn variadic(&self) -> bool

fn volatile(&self) -> bool

fn aliases(&self) -> &'static [&'static str]

Optional list of additional alias names (case-insensitive) that should resolve to this function. Default: empty slice. Implementors can override to expose legacy names. Returned slice must have ’static lifetime (typically a static array reference).

fn function_salt(&self) -> u64

fn dispatch<'a, 'b, 'c>( &self, args: &'c [ArgumentHandle<'a, 'b>], ctx: &dyn FunctionContext<'b>, ) -> Result<CalcValue<'b>, ExcelError>

Dispatch to the unified evaluation path with automatic argument validation.