Perpetuate a failure to make Windows paths behave reasonably with
path-manipulation functions.
In one case, the new implementation seemed better than the old one, so
I've changed the old implementation (by deleting code) and test cases.
The old code would split "x /y" to "\\?\REL\x " and "y", and the new
one splits to "x /" and "y"; the trailing separator is now enough to
preserve the space character, and it also preserves the directoryness
of the path. Of course, "x /" splits to 'relative and "\\?\REL\x " as
it strips away the trailing "/".
A remaining problem in both implementations: some Windows API
functions implicitly erase a trailing "." in a directory name, making
"x./y" equivalent to "x/y". The Racket path-manipulation functions
don't do that, so splitting and recombining "x./y" does not access the
same path as the original. This apparently hasn't been a problem in
practice, and there are so many terrible hacks already, so I left it
alone.
The new implementation perpetuates also the implementation mistake of
representing paths internally as byte strings. If, in some terrible
universe, I'm forced to do this again, the right choice is probably to
keep the path in a parsed form with enough information to reconstruct
the original, but with the information sorted nicely to make various
normalizations and combinations easy.
This change adjusts the way that trust is threaded through bytecode
and the code inspector. In Racket v6.x, reading bytecode would fail if
the code inspector is non-original and if the bytecode contains a
reference to an unsafe operation. Now, reading bytecode doesn't fail
for that reason, but all bytecode is marked as non-runnable (even
without references to unsafe operations) when loaded under a
non-original code inspector. A `read` operation by itself remains as
safe as ever.
This commit also disables the bytecode validator. For now, the
validate can be re-enabled with `PLT_VALIDATE_LOAD`.
Make `s-exp->fasl` generate an encoding that can be parsed by any
future version of `fasl->s-exp`. The new format does not rely on the
runtime system's bytecode writer and reader.
Some built-in bytes-converter combinations that were not
supposed to require custodian registration were neveretheless
registered, which created a small leak for some programs.
Make the module path index for bindings in a `module->namespace` be a
"self" MPI (with #f for path and base), instead of the MPI associated
with bindings as view from the outside of the module instance. That
makes interactive evalaution in the namespace more closely approximate
expansion within the original module.
Example use: ASL detects a "self" MPI to determine when it should
allow assignment to module-defined variables in the REPL.
This change means that if a macro transformer calls expand (not
local-expand, but top-level expand), the macro debugger won't
receive the inner expand's events. Previously, the macro debugger
tried to parse and then discard the inner expand, but that was
brittle and complicated the grammar.
The old reader used an internal option to short-circuit special-value
reading when a special value acts as a terminating "character". Expose
that shortcut by allowing 'special as an argument to
`peek-{char,byte}-or-special`, and update the reader to use it.
This commit merges changes that were developed in the "racket7" repo.
See that repo (which is no longer modified) for a more fine-grained
change history.
The commit includes experimental support for running Racket on Chez
Scheme, but that "CS" variant is not built by default.
