std.fmt

import std.fmt

fmt.format(r"{} of {}", 3, 10)     # "3 of 10"
fmt.fixed(3.14159, 2)               # "3.14"
fmt.inspect([1, { a: 1 }])          # "list[2] [int 1, map{1} {a: int 1}]"

Removed in 0.61

pad_left, pad_right, repeat and truncate duplicated std.str; they were deprecated in 0.59 and removed in 0.61. Calling one is an error naming its replacement, and the replacements give the same answers:

inspect

fmt.inspect(v) renders any value with its types and structure - "list[2] [int 1, map{1} {a: int 1}]" - with map and struct keys sorted, so the output is deterministic, and a secret shown as [secret], so inspecting a config map cannot leak a credential. It was debug.inspect until 0.59.

format needs a raw string

fmt.format(r"{} of {}", 3, 10)     # correct

fmt.format("{} of {}", 3, 10)      # error: Expected expression in string interpolation

A normal string triggers Ecko's own {expr} interpolation at parse time, so {} would be read as an empty expression. The r"..." prefix suppresses interpolation and hands the placeholders to format.

This trips everyone once. If you get a parse error pointing at a format call, that is why.

When to use which

Most of the time you do not need format at all - interpolation is built into the language and reads better:

"{count} of {total}"                # idiomatic
fmt.format(r"{} of {}", count, total)

Reach for format when the template is data: read from a config file, chosen from a table, or reused across call sites with different arguments.

Fixed decimal places

fmt.fixed(value, places) renders a number to a set number of decimals, padding with zeroes rather than trimming the text:

fmt.fixed(1.5, 3)                   # "1.500"
fmt.fixed(2.5, 0)                   # "3"
fmt.fixed(decimal("0.07"), 2)       # "0.07"

Two details worth knowing. Halves go away from zero, so 2.5 at zero places is "3" - not the half-to-even most languages' formatters use. And a decimal is rounded in decimal, never through a float, so an exact 0.07 stays 0.07 instead of picking up a binary artefact. That makes it the one to reach for when formatting money.

Padding and truncating

Use std.str: pad_start, pad_end, center and zfill are character-aware, and a slice s[..n] truncates. For terminal tables where display width differs from character count, the tui package handles width properly.