Many places in Serenity allow you to format strings, similar to printf()
, for example ByteString::formatted()
, StringBuilder::appendff()
, or dbgln()
. These are checked at compile time to ensure the format string matches the
number of parameters. The syntax is largely based on
the C++ std::formatter
syntax
but there are some differences.
For basic usage, any occurrences of {}
in the format string are replaced with the other arguments, converted to string
form, in order:
ByteString::formatted("Well, {} my {} friends!", "hello", 42) == "Well, hello my 42 friends!";
If you want to include a literal {
in the output, use {{
:
ByteString::formatted("{{ {}", "hello") == "{ hello";
You can refer to the arguments by index, if you want to repeat one or change the order:
ByteString::formatted("{2}{0}{1}", "a", "b", "c") == "cab";
To control how the arguments are formatted, add colon after the optional index, and then add format specifier characters:
ByteString::formatted("{:.4}", "cool dude") == "cool";
ByteString::formatted("{0:.4}", "cool dude") == "cool";
In order, the format can contain:
- Fill character and alignment
- Sign
#
Hash0
Zero- Width
- Precision
- Type specifier
Each of these is optional. You can include any combination of them, but they must be in this order.
This is an optional fill character, followed by an alignment. The fill character can be anything apart from {
or }
,
and is used to fill any space left when the input has fewer characters than the format requests. By default, it is a
space. (
)
The alignment characters are:
<
: Align left.>
: Align right.^
: Align centered.
+
: Always display a sign before the number.-
: Display a sign for negative numbers only.- (space): Display a sign for negative numbers, and a leading space for other numbers.
#
causes an "alternate form" to be used.
For integer types, this adds the number-base prefix after the sign:
0b
for binary.0
for octal.0x
for hexadecimal.
0
pads the number with leading zeros.
The width defines the minimum number of characters in the output. The precision is a .
followed by a precision number,
which is used as the precision of floating-point numbers, or a maximum-width for string values.
Both the width and precision can be provided as a replacement field ({}
, optionally including an argument index) which
allows you to use an integer argument instead of a hard-coded number.
Type | Effect | Example output |
---|---|---|
nothing | default format | Anything! :^) |
b | binary | 110 , 0b000110 |
B | binary uppercase | 110 , 0B000110 |
d | decimal | 42 , +0000042 |
o | octal | 043 |
x | hexadecimal | ff0 , 0x00000ff0 |
X | hexadecimal uppercase | FF0 , 0X00000FF0 |
c | character | a |
s | string | well, hello friends! |
p | pointer | 0xdeadc0de |
f | float | 1.234 , -inf |
a | hex float | |
A | hex float uppercase | |
hex-dump | hexadecimal dump | fdfdfdfd , 3030 00 |
Not all type specifiers are compatible with all input types, of course.
You can provide a custom AK::Formatter<Foo>
class to format Foo
values. For the simplest case where you already have
a function that produces a string from your type, that would look like this:
template<>
struct AK::Formatter<Web::CSS::Selector> : Formatter<StringView> {
ErrorOr<void> format(FormatBuilder& builder, Web::CSS::Selector const& selector)
{
return Formatter<StringView>::format(builder, selector.serialize());
}
};
More advanced formatters that make check for format-specifier flags can be written by referring to the fields
in StandardFormatter
(which most Formatter
classes extend).
The AK::HasFormatter<T>
template has a boolean value representing whether T
can be formatted.
The FormatIfSupported<T>
makes use of this to return either the formatted value of T
, or a series of ?
s if the
type cannot be formatted. For example:
// B has a Formatter defined, but A does not.
ByteString::formatted("{}", FormatIfSupported { A {} }) == "?";
ByteString::formatted("{}", FormatIfSupported { B {} }) == "B";