* Widen display_width_of() to cover Emoji_Presentation code points
fmt::detail::display_width_of() only treated East Asian Wide/Fullwidth
code points and two hand-picked emoji ranges as two columns wide, so
emoji outside those ranges (e.g. the Dingbats block: cross mark U+274C,
white heavy check mark U+2705) were measured as one column even though
most terminals render them double-width. This produced visibly
inconsistent padding under {:^N} compared to CJK text (fixes#4851).
Replaced the ad hoc boolean expression with a sorted table of ranges
(East Asian Wide/Fullwidth plus the full Emoji_Presentation set from
Unicode's emoji-data.txt) looked up via binary search, and added
regression tests covering the original report plus edge cases:
multiple emoji, mixed emoji/CJK/ASCII content, precision truncation,
alignment/fill variants, newly covered emoji ranges, and regional
indicator (flag) pairs.
* Make wide_cp_ranges C++11-compatible; apply clang-format
* Move wide_cp_range struct into display_width_of()
* Derive wide_cp_ranges from East_Asian_Width data, restore constexpr
* Move wide_cp_ranges to display_width_of function body
* Update format.h
* Update wide_cp_ranges array to include comments
* Add inline to display_width_of and add 1F300-1F5FF and 1F900-1F9FF ranges to match [format.string.std]
* Update display_width_of lo hi varible declaration
---------
Co-authored-by: Eduardo Gomez Saldias <50159560+edugomez102@users.noreply.github.com>
fmt::format("{:6}", nan) gave "nan " where std::format and printf("%6f")
give " nan". write_nonfinite was the only numeric write path taking
write_padded's align::left default, so a width with no explicit align
left-aligned inf and nan while every finite value right-aligned. That
default only reaches align::none, leaving explicit <, > and ^ and the 0
flag unchanged.
fmt::format("{:a}", 0.0) produced 0x0p-1022 where printf's %a gives
0x0p+0, and likewise for -0.0, {:A}, {:#a} and any explicit precision.
basic_fp::assign maps a zero biased exponent to 1 ("subnormals use
biased exponent 1"), which is right for subnormals but not for zero, so
format_hexfloat inherited the minimum subnormal exponent for a value
whose significand carries no information. Reset the exponent there
rather than in assign, which the decimal path shares.
A zero significand can only come from +-0: normals get the implicit bit
added, and a subnormal has a nonzero significand by definition. So the
guard cannot affect anything else, and denorm_min keeps its existing
denormalized form (0x0.0000000000001p-1022, asserted in format_double).
The five new expectations were taken from printf rather than derived,
and all fail without this change.
Formatting an integer with ':c' used the magnitude (abs_value) and never
range-checked, so negatives were mangled and out-of-range values silently
truncated. Copy the value as a character and report an error when it is out
of range, treating all character types as unsigned for portability.
* Handle allocator propagation in basic_memory_buffer::move
Update `basic_memory_buffer::move` to respect `propagate_on_container_move_assignment`allocator trait.
If the allocator should not propagate and differs from the target's allocator,
fallback to copying the buffer instead of transferring ownership.
This avoids potential allocator mismatch issues and ensures exception safety.
* Add test cases for the updated move ctor
- Added two test cases `move_ctor_inline_buffer_non_propagating` and `move_ctor_dynamic_buffer_non_propagating`
- Added `PropageteOnMove` template parameter to `allocator_ref` class to be compatible with the old test cases
- `allocator_ref` now implements `!=` and `==` operators
