Commit Graph

30 Commits

Author SHA1 Message Date
Hendiadyoin1
35e3e3d483 AK: Make FixedPoint work on platforms without __int128 2023-09-06 07:17:03 -06:00
ronak69
2caf68fd03 AK: Add binary and octal mode formatting for FixedPoint 2023-08-29 11:10:45 +02:00
ronak69
352480338e AK: Refactor FixedPoint's formatter
The main change is the simplification of the expression
`(10^precision * fraction) / 2^precision` to `5^precision * fraction`.

Those expressions overflow or not depends on the value of `precision`
and `fraction`. For the maximum value of `fraction`, the following table
shows for which value of `precision` overflow will occur.

            Old   New
    u32      08    10
    u64      15    20
    u128     30    39

As of now `u64` type is used to calculate the result of the expression.
Meaning that before, only FixedPoints with `precision` less than 15
could be accurately rendered (for every value of fraction) in decimal.
Now, this limit gets increased to 20.

This refactor also fixes, broken decimal render for explicitly specified
precision width in format string, and broken hexadecimal render.
2023-08-29 11:10:45 +02:00
ronak69
d40807681d AK: Cast lhs of left shifts to FixedPoint's underlying integer type
By default, `1` is of the type `int` which is 32-bits wide at max.
Because of that, if `precision` of a `FixedPoint` is greater than 32,
the expression `1 << precision` will get clamped at 32-bits and the
result will always be zero. Casting `1` to the wider underlying type
will make the expression not overflow.
2023-08-29 11:10:45 +02:00
ronak69
4c6ea4a963 AK: Fix off-by-one error in round-to-even logic of FixedPoint
Because of the off-by-one error, the second bit of the fraction was
getting ignored in differentiating between fractions equal to 0.5 or
greater than 0.5. This resulted in numbers like 2.75 being considered
as having fraction equal to 0.5 and getting rounded incorrectly (to 2).
2023-08-29 11:10:45 +02:00
Hendiadyoin1
daacc5c6c2 AK: Rename AK::FixedPoint::round to rint and fix a rounding error
`rint` is a more accurate name for the roudning mode as the fixme above
stated
2023-08-14 14:20:45 -06:00
Hendiadyoin1
05c959e40b AK: Change standard casting method of FixedPoint to truncation
This matches what floats do.

Also fix typo `trunk`->`trunc`
2023-08-14 14:20:45 -06:00
Hendiadyoin1
8526791617 AK: Use wider type for FixedPoint division
This allows us to shift first and then divide, preserving more precision
2023-08-14 14:20:45 -06:00
Hendiadyoin1
e609ac74a3 AK: Fix FixedPoint multiplication rounding and overflow behaviour
We now preform the multiplication in a widened type which makes it
overflow-safe and use the correct bit for rounding direction detection.
2023-08-14 14:20:45 -06:00
Nico Weber
f56b897622 Everywhere: Fix a few typos
Some even user-visible!
2023-04-12 19:37:35 +02:00
Tim Ledbetter
72ea046b68 AK: Add option to the string formatter to use a digit separator
`vformat()` can now accept format specifiers of the form
{:'[numeric-type]}. This will output a number with a comma separator
every 3 digits.

For example:

`dbgln("{:'d}", 9999999);` will output 9,999,999.

Binary, octal and hexadecimal numbers can also use this feature, for
example:

`dbgln("{:'x}", 0xffffffff);` will output ff,fff,fff.
2023-04-11 13:03:30 +02:00
nipos
e35a858bde AK: Remove signbit definition to prevent conflict in FixedPoint 2023-03-04 21:34:54 -07:00
Nico Weber
b898a46d7f AK: Make FixedPoint(FloatingPoint) ctor round instead of truncating
This is needed to have code for creating an in-memory sRGB profile using
the (floating-ppoint) numbers from the sRGB spec and having the
fixed-point values in the profile match what they are in other software
(such as GIMP).

It has the side effect of making the FixedPoint ctor no longer constexpr
(which seems fine; nothing was currently relying on that).

Some of FixedPoint's member functions don't round yet, which requires
tweaking a test.
2023-03-03 09:23:02 +00:00
kleines Filmröllchen
961e263129 AK: Add FixedPoint::clamp 2023-02-25 20:49:41 -07:00
Nico Weber
a30e364f1a AK: Fix printing of negative FixedPoint values
Fixes #17514 by comparing to 0 before truncating the fractional part.
2023-02-18 19:34:10 +01:00
Nico Weber
9c5820326d AK: Make FixedPoint::create_raw public and constexpr 2023-01-23 18:09:23 +00:00
Timothy Flynn
5063e218af AK: Move the AK::FixedPoint formatter to FixedPoint.h
This does not need to be defined in Format.h. This causes FixedPoint.h
to be included everywhere. This is particularly going to be an issue
when trying to include <CoreServices/CoreServices.h> on macOS. The macOS
SDK defines its own FixedPoint structure which will conflict with ours.
2023-01-19 11:29:48 +00:00
Ben Wiederhake
6fd478b6ce Everywhere: Remove unused includes of AK/Format.h
These instances were detected by searching for files that include
AK/Format.h, but don't match the regex:

\\b(CheckedFormatString|critical_dmesgln|dbgln|dbgln_if|dmesgln|FormatBu
ilder|__FormatIfSupported|FormatIfSupported|FormatParser|FormatString|Fo
rmattable|Formatter|__format_value|HasFormatter|max_format_arguments|out
|outln|set_debug_enabled|StandardFormatter|TypeErasedFormatParams|TypeEr
asedParameter|VariadicFormatParams|v_critical_dmesgln|vdbgln|vdmesgln|vf
ormat|vout|warn|warnln|warnln_if)\\b

(Without the linebreaks.)

This regex is pessimistic, so there might be more files that don't
actually use any formatting functions.

Observe that this revealed that Userland/Libraries/LibC/signal.cpp is
missing an include.

In theory, one might use LibCPP to detect things like this
automatically, but let's do this one step after another.
2023-01-02 20:27:20 -05:00
Linus Groh
d26aabff04 Everywhere: Run clang-format 2022-12-03 23:52:23 +00:00
Andreas Kling
ae3ffdd521 AK: Make it possible to not using AK classes into the global namespace
This patch adds the `USING_AK_GLOBALLY` macro which is enabled by
default, but can be overridden by build flags.

This is a step towards integrating Jakt and AK types.
2022-11-26 15:51:34 +01:00
Andrew Kaster
83603d68d2 AK: Make sure we don't include Math.h or math.h from KERNEL 2022-04-23 10:43:32 -07:00
kleines Filmröllchen
98058f7efe AK: Add FixedPoint base 2 logarithm
The log base 2 is implemented using the binary logarithm algorithm
by Clay Turner (see the link in the comment)
2022-02-28 13:59:31 +01:00
kleines Filmröllchen
30002c2ccb AK: Add bit shift to FixedPoint 2022-02-28 13:59:31 +01:00
kleines Filmröllchen
5d6e3441fe AK: Fix FixedPoint to integral comparisons
Add tests to ensure that the fixed point numbers compare correctly to
integrals
2022-02-28 13:59:31 +01:00
Tom
c468a9cc2d AK: Add FixedPoint cast operator for up/downcasting to other sizes
This enables casting between different size FixedPoint variables or
constructing them from other sized FixedPoint values.
2022-01-23 22:45:21 +00:00
Tom
06fc72ca0c AK: Fix warnings when using FixedPoint with a precision >= 32 bits 2022-01-23 22:45:21 +00:00
Tom
f021baf255 AK: Add Formatter<FixedPoint<...>> without floating point
Rather than casting the FixedPoint to double, format the FixedPoint
directly. This avoids using floating point instruction, which in
turn enables this to be used even in the kernel.
2022-01-23 22:45:21 +00:00
Tom
77b3230c80 AK: Loosen FixedPoint template contraints and forward-declare it
Because AK/Concepts.h includes AK/Forward.h and concepts cannot be
forward declared, slightly losen the FixedPoint template arguments
so that we can forward declare it in AK/Forward.h
2022-01-23 22:45:21 +00:00
Andreas Kling
216e21a1fa AK: Convert AK::Format formatting helpers to returning ErrorOr<void>
This isn't a complete conversion to ErrorOr<void>, but a good chunk.
The end goal here is to propagate buffer allocation failures to the
caller, and allow the use of TRY() with formatting functions.
2021-11-17 00:21:13 +01:00
Hediadyoin1
fdef6e5f76 AK: Add FixedPoint arithmetic helper
Co-authored-by: Hendiadyoin1 <leon2002.la@gmail.com>
Co-authored-by: kleines Filmröllchen <malu.bertsch@gmail.com>
2021-08-31 17:03:55 +04:30