The new mechanism can deal with more complex scenarios such as
not nil field appearing in a non-default case object branch or
a field within a generic object that may depend on a when branch.
The commit also plugs another hole: the user is no longer able
to create illegal default values through seq.setLen(N).
New issue: since `Table[A, B]` allocates its backing storage with
`newSeq[KeyValuePair[A, B]]`, it's no longer legal to create a table
with `not nil` types used as either keys or values.
* new minor feature: macros for proc types, to be documented
* Finished the implementation and added tests
* [skip ci] Describe the new custom pragmas in the manual and the changelog
Co-authored-by: Zahary Karadjov <zahary@gmail.com>
* Unwind just the "pseudorandom probing" (whole hash-code-keyed variable
stride double hashing) part of recent sets & tables changes (which has
still been causing bugs over a month later (e.g., two days ago
https://github.com/nim-lang/Nim/issues/13794) as well as still having
several "figure this out" implementation question comments in them (see
just diffs of this PR).
This topic has been discussed in many places:
https://github.com/nim-lang/Nim/issues/13393https://github.com/nim-lang/Nim/pull/13418https://github.com/nim-lang/Nim/pull/13440https://github.com/nim-lang/Nim/issues/13794
Alternative/non-mandatory stronger integer hashes (or vice-versa opt-in
identity hashes) are a better solution that is more general (no illusion
of one hard-coded sequence solving all problems) while retaining the
virtues of linear probing such as cache obliviousness and age-less tables
under delete-heavy workloads (still untested after a month of this change).
The only real solution for truly adversarial keys is a hash keyed off of
data unobservable to attackers. That all fits better with a few families
of user-pluggable/define-switchable hashes which can be provided in a
separate PR more about `hashes.nim`.
This PR carefully preserves the better (but still hard coded!) probing
of the `intsets` and other recent fixes like `move` annotations, hash
order invariant tests, `intsets.missingOrExcl` fixing, and the move of
`rightSize` into `hashcommon.nim`.
* Fix `data.len` -> `dataLen` problem.
* stacktraces can now show custom runtime msgs
* improve tests/stdlib/tstackframes.nim
* fix test for --gc:arc
* test --stacktraceMsgs:on and --stacktraceMsgs:off
* --stacktracemsgs:off by default
* Fix typeSym.getImpl for ref types
* Fix a codegen issue affecting the test suite of nim-beacon-chain
* Fix tests/stdlib/tjsonmacro
To understand the fix better it may help to take a look
at the history of the replaced code.
The nil check that is removed in this commit was introduced
in another fix that failed to identify the root cause of the
issue - namely that we allow an object type to exist for which
no ast is present:
https://github.com/nim-lang/Nim/pull/9601/files
The original intention of the code is more obvious here:
https://github.com/nim-lang/Nim/pull/9538/files
* new syntax for lvalue references: `var b {.byaddr.} = expr`
* on type mismatch, `???(0, 0)` not shown anymore
* * compiler now lowers `var a: {.foo.}: MyType = expr` to foo(a, MyType, expr)
* new pragmas.byaddr defined in pure library code exploiting this lowering
* skip `template foo() {.pragma.}`
* Implement SSL/TLS certificate checking #782
* SSL: Add nimDisableCertificateValidation
Remove NIM_SSL_CERT_VALIDATION env var
tests/untestable/thttpclient_ssl.nim ran successfully on Linux with libssl 1.1.1d
* SSL: update integ test to skip flapping tests
* Revert .travis.yml change
* nimDisableCertificateValidation disable imports
Prevent loading symbols that are not defined on older SSL libs
* SSL: disable verification in net.nim
..when nimDisableCertificateValidation is set
* Update changelog
* Fix peername type
* Add define check for windows
* Disable test on windows
* Add exprimental GitHub action CI for SSL
* Test nimDisableCertificateValidation
