Medohh2120

/*
Name: FILL
Description : Fills blank cells of an array in a given direction.
    Supports 2D arrays.
    Supports multiple directions via array constant {"d", "r", ...} respectively ⚠.
    [directions]: control Filling direction:
    "d" or Omitted (down).
    "u" (up).
    "r" (right).
    "l" (left). 

Medohh2120

/*
Name: WRAPROWS_AT
Description: Wraps a column into rows at every occurrence of a trigger value.
 Supports multiple triggers via array constant {"start", "end", ...}.
 ⚠ Triggers are not case-sensitive "end" and "END" are treated the same.
 Consecutive triggers in data are automatically treated as one -> no empty groups produced.
 Groups are internally joined via CHAR(1) as a safe delimiter, never colliding with real data.
 Optional "pad_with" pads shorter rows to uniform width.
 ⚠ Note: TEXTJOIN cap applies -> max ~32,000 chars per group.
Made By: Medohh2120

Medohh2120

/*
Name: GROUPBY_RETITLE
Description: Replaces the native (GROUPBY+HSTACK) value field headers with provided [custom_headers] array.
 GROUPBY.[Field_headers] must equal 2 or 3.
Made By: Medohh2120
*/

GROUPBY_RETITLE = LAMBDA(groupby_result, [custom_headers],
    LET(
        n_value_fields, SUM(--(TAKE(groupby_result, 1) <> "")),

Medohh2120

/*
Name: BYROW⊟
Description: A high-performance version of BYROW that supports "Array of Arrays" (nested results).
 [orient]   logic to control the assembly of results: 
 (0) or Omitted: (Stacks each row result vertically).
 (1) : (Stacks each row result horizontally). 
Made By: Medohh2120
*/

BYROW⊟ = LAMBDA(array, function, [orient],

Medohh2120

/*
Name: BENCHMARK
Description: Runs a formula N iterations and returns avg & total execution time.
    Func must be wrapped in LAMBDA()  =BENCHMARK(LAMBDA(your_formula), 50)
    [iterations] default: 1
    [time_unit]  default: 0 (ms), 1 = seconds
    Because this function uses NOW(), manual calculation mode is highly recommended.  
Made By: Medohh2120
*/

Medohh2120

/*
Name: REPLACE_VECTOR
Description: Replaces targeted rows and/or columns in an array with given "new_arrays".
    If multiple indices are supplied per axis, the same "new_array" is placed at each target.
    preserves the original order of the remaining rows/columns.
    Optional "pad_with" replaces dimension-mismatch errors in the final result.
Made By: Medohh2120
*/

REPLACE_VECTOR = LAMBDA(array, [row_idxs], [new_row_array], [col_idxs], [new_col_array], [pad_with],

Medohh2120

/* 
Name: REPEAT
Recommended version: v2024 or 365
Description: Repeats a data range based on a frequency column.
    Optimized with Binary Search. 
Made By: Medohh2120
*/

Repeat = LAMBDA(Values,Number_times, 
    LET(

Medohh2120

/*
Name: TEXTBETWEEN
Description: Returns text between delimiters. 
   - Standardized to native TEXTAFTER/BEFORE syntax.
   - Supports Multi-input: Text (Rows) x Delimiters (Columns).
   - Strict Mode: Assumes equal number of start/end delimiters.
Made By: Medohh2120
*/

TEXTBETWEEN = LAMBDA(text, start_delim, end_delim, [instance_start], [instance_end], [case_sensitive_start], [case_sensitive_end], [match_end_start], [match_end_end], [if_not_found],

Medohh2120

/*
Name: UNPIVOT_PLUS	
Description: Given a table or a range with headers, Transforms "Wide" data into a "Long" database format.	
 V4 UPDATE: "pad_blanks_with" feature to replace empty grid cells (prevents Pivot Table type-errors).
 Automatically handles merged cells (Fill-Down or Fill-Right logic). 	
 optionally removing Grid blank entries.	
 optionally removing Grid errors.	
Made By: Medohh2120
*/