Features

Design Goals

Goal	Description
Seamless interchangeability	Works transparently with standard `character` variables via overloaded `//`, `=`, and comparison operators
Rich built-in methods	`split`, `replace`, `join`, `basename`, `camelcase`, `to_number`, and many more as type-bound procedures
Low memory consumption	Only one `character(kind=CK, len=:), allocatable` member — arrays of strings can hold elements of different lengths
Safe	Almost all methods are `elemental` or `pure`
Robust	Test-Driven Developed — a comprehensive doctest suite is embedded directly in the source

Case conversion: upper, lower, swapcase, capitalize, camelcase, snakecase, startcase
Search and replace: replace, search, index, count, scan, verify
Splitting and joining: split, split_chunked, partition, join, strjoin
Padding and trimming: fill, strip, adjustl, adjustr, trim
Transformation: reverse, unique, insert, slice, escape, unescape
Encoding: encode / decode (Base64 via BeFoR64)
Terminal output: colorize (ANSI styles via FACE)

Assign any PENF integer or real kind directly to a string via overloaded assignment(=)
Cast back with to_number(kind=...) — supports all PENF kinds: I1P, I2P, I4P, I8P, R4P, R8P, R16P
Inquiry methods: is_number, is_integer, is_real, is_digit

Read/write files line-by-line or as streams: read_file, write_file, read_line, read_lines, write_line, write_lines
Path components: basedir, basename, extension
File-system glob: glob
Unique temporary names: tempname

Operator	Returns	Notes
`=`	—	From `character`, `string`, or any PENF numeric kind
`//`	`character`	Enables seamless use with Fortran intrinsics
`.cat.`	`string`	Concatenation keeping the `string` type
`==`, `/=`	`logical`	Equality / inequality
`<`, `<=`, `>=`, `>`	`logical`	Lexicographic comparison

Generic interfaces that accept type(string) in place of character:

adjustl · adjustr · count · index · len_trim · repeat · scan · trim · verify