Interestingly, Clang's `-Wformat-overflow` seems to only warn
about buffers so small they will *always* underflow,
whereas GCC tries to infer whether they can overflow *at all*.
(I'm guessing they lean towards false-negatives and false-positives resp.)
Anyway, calling `sprintf` here remains safe, since GCC (notably, via CI)
checks our work, and it simplifies the code ever so slightly while
also providing an (admittedly negligible) performance improvement.
Note that there may be more locations where we could use `sprintf`,
but a cursory glance at our other uses of `snprintf` didn't seem fruitful.
(One way to test is to set the target buffer size to 1, and see if a
warning pops up for such a trivially-wrong size. If not, you can be certain
the compiler won't be able to help you with a more realistic size.)
This reverts a change introduced in fd78a9ae8, though it wasn't that commit's main point
so I'm feeling okay with undoing that.
This feels like an overkill change, using a static string is good enough for this
since we never modify this. I have considered using `string_view` instead, to have
the best of both worlds, but that's not NUL-terminated so our print functions
get a little grumpy.
Non-indexed images can still have a PLTE chunk. We should use
`sortRgb` instead of `sortIndexed` for them, since their pixels' colors
may not all be present in the PLTE chunk and would be unsortable.
RGBASM computes these with unsigned arithmetic and casts back, since
signed overflow is UB in C++, but RGBLINK's RPN evaluator used `int32_t`
directly, so `src/link/patch.cpp` tripped UBSan on lines 117, 121, and
124. Share the three operators through `opmath.cpp` so both evaluators
stay in step.
Negative macro arguments count from the end, i.e. `\<-1>`
is equivalent to `\<_NARG>`, even after `shift`ing them.
Negative arguments cannot be used to access shifted values.
- Fix logic for color set comparison (which affects sorting them)
- Prune color sets which are proper subsets of newly-encountered ones
(a comment implied we were already doing this, but we weren't)
- Add more verbose logging to debug this behavior
Instead of reading 64 characters at a time into a rolling buffer
as `shiftChar()` is called, we read 8192 characters at a time into
a complete buffer before any `peek()`/`shiftChar()` operations.
This will let us use C++20 features that GCC 9's experimental
C++20 support did not yet cover, such as "concepts".
This reverts some commits:
- 6bcd79b997
- d5ce5329ea
- 728d14879b
- Add more `assume()` checks in the lexer
- Replace double negative "`disable* = false`" with "`enable* = true`"
- Naming convention of "`std::deque<> *Stack`" when relying on
`.push_front()` and `.pop_back()`
Eldred Habert
Rangi
Rangi
Max Freedom Pollard
vulcandth