It is hard to impossible to work around iconv() implementations which
do not properly set errno according to POSIX. We therefore do no
longer allow to build against such iconv() implementations.
Co-Authored-By: Nikita Popov <nikita.ppv@googlemail.com>
Don't special-case nested arrays/objects in str_replace(), instead
perform a string cast on them as well. For arrays, this will always
result in the usual conversion warning.
This behavior is consistent with preg_replace(). If we didn't want
to cast the array to string here, we should instead perform the
replacement recursively. Silently copying it is just confusing.
Currently, trait methods are aliased will continue to use the
original function name. In a few places in the codebase, we will
try to look up the actual method name instead. However, this does
not work if an aliased method is used indirectly
(https://bugs.php.net/bug.php?id=69180).
I think it would be better to instead actually change the method
name to the alias. This is in principle easy: We have to allow
function_name to be changed even if op array is otherwise shared
(similar to static_variables). This means we need to addref/release
the function_name separately, but I don't think there is a
performance concern here (especially as everything is usually
interned).
There is a bit of complication in opcache, where we need to make
sure that the function name is released the correct number of times
(interning may overwrite the name in the original op_array, but we
need to release it as many times as the op_array is shared).
Fixes bug #69180.
Fixes bug #74939.
Closes GH-5226.
Make ReflectionClassConstant->class the declaring class, not the
class on which the constant was fetched. This matches the behavior
for properties and methods.
While the `$command` passed to `proc_open()` had to be wrapped in
double-quotes manually, that was implicitly done for all other
program execution functions. This could easily introduce bugs and
even security issues when switching from one to another program
execution function.
Furthermore we ensure that the additional quotes are always
unwrapped regardless of what is passed as `$command` by passing
the `/s` flag to cmd.exe. As it was, `shell_exec('path with
spaces/program.exe')` did execute program.exe, but adding an
argument (`shell_exec('path with spaces/program.exe -h)`) failed
to execute program.exe, because cmd.exe stripped the additional
quotes.
While these changes obviously can cause BC breaks, we feel that in
the long run the benefits of having consistent behavior for all
program execution functions outweighs the drawbacks of potentially
breaking some code now.
Any number of arguments can be replaced by a variadic one, so
long as the variadic argument is compatible (in the sense of
contravariance) with the subsumed arguments.
In particular this means that function(...$args) becomes a
near-universal signature: It is compatible with any function
signature that does not accept parameters by-reference.
This also fixes bug #70839, which describes a special case.
Closes GH-5059.
libcurl 7.29.0 has been released almost eight years ago, so this
version is supposed to be available practically everywhere. This bump
also allows us to get rid of quite some conditional code and tests
catering to very old libcurl versions.
While `imagesetinterpolation()` is available as of PHP 5.5.0,
there is no according getter function, so users would have to track the
current interpolation method manually.
To remedy this, we introduce `imagegetinterpolation()` as thin wrapper
for `gdImageGetInterpolationMethod()` (which has been introduced with
libgd 2.1.1), and use `im->interpolation_id` as fallback for older
libgd. Since our bundled libgd does not yet have this function, we add
it.
We also simplify the recently introduced bug79068.phpt, where it is
sufficient to check that the interpolation method has not been changed.
That parameter is mostly useless in practise, and likely has been
directly ported from the underlying `gdImagePolygon()` and friends,
which require that parameter since the number of elements of the point
array would otherwise be unknown. Typical usages of `imagepolygon()`,
`imageopenpolygon()` and `imagefilledpolygon()` pass `count($points)/2`
or hard-code this value as literal. Since explicitly specifying this
parameter is annoying and error-prone, we offer the possibility to omit
it, in which case the `$points` array must have an even number of
elements, and the number of points is calculated as `count($points)/2`.
We actually have to check `$num_points` instead of `2*count($points)`,
because the latter may be greater than the former, but not all elements
of `$points` are guaranteed to be used. This allowed to pass arrays
with excess elements to draw polygons with less than three vertices.
While the current implementation of `gdImagePolygon()` and friends
would allow us to draw monogons and digons, we don't allow that
anymore, because the respective drawing primitives work slightly
different (e.g. drawing lines support anti-aliasing, but drawing
general polygons does not).
To minimize the BC break, we do not fix this longstanding issue for PHP
7, but target PHP 8 only.
`write_dimension` object handlers have to be able to handle `NULL`
`offset`s; for now we simply throw an exception instead of following
the `NULL` pointer.
