Nim 1.4.x compiled the below code without error when using
`--experimental:strictFuncs`
import std/sequtils
type Foo = ref object
let foo1 = Foo()
let foo2 = Foo()
let foos = @[foo1, foo2]
let fooTuples = @[(foo1, 1), (foo2, 2)]
discard repeat(foo1, 3)
discard zip(foos, foos)
discard unzip(fooTuples)
However, since 2020-12-09, devel Nim produced errors like
/tmp/bar.nim(11, 15) template/generic instantiation of `repeat` from here
/foo/nim/pure/collections/sequtils.nim(172, 6) Error: 'repeat' can have side effects
an object reachable from 'x' is potentially mutated
/foo/nim/pure/collections/sequtils.nim(183, 15) the mutation is here
/foo/nim/pure/collections/sequtils.nim(183, 15) is the statement that connected the mutation to the parameter
This commit reverts some `proc` to `func` changes so that code that:
- calls `repeat`, `zip`, or `unzip`
- and instantiates them with types containing `ref`
can once again be compiled with `strictFuncs`. Otherwise, a user might
be forced to drop or alter their `strictFuncs` use when upgrading from
Nim 1.4.x, or when writing new code that uses these procedures (at least
for now, with the current `strictFuncs` implementation).
This commit also adds tests to assert that the remaining funcs in this
module can be compiled with `strictFuncs` when used with types
containing `ref`.
The original batch of `proc` to `func` changes in `sequtils.nim` was in
commit 6f57ebae34, which was partially reverted in 38eb021f81.
See also: https://github.com/nim-lang/Nim/issues/16305
* Fix initrand to avoid random number sequences overlapping
* Minor fix
* Fix compile error on js backend
* Disable new test for js backend
* Minor fix
* tempfiles module uses random.initRand()
* Remove unused module import from lib/std/tempfiles.nim
* Initialize baseState in initRand()
* Run tests/stdlib/trandom.nim from tests/test_nimscript.nims
* baseState is initialized only with sysrand.urandom and quit if failed
* Add comments
* fixes#17369
* megatest is green for --cpu:arm64
* docgen output includes more tags/raises
* implemented 'effectsOf'
* algorithm.nim: uses new effectsOf annotation
* closes#18376
* closes#17475
* closes#13905
* allow effectsOf: [a, b]
* added a test case
* parameters that are not ours cannot be declared as .effectsOf
* documentation
* manual: added the 'sort' example
* bootstrap with the new better options
* Use assert in runnableExamples and improve boundary check
* Add more tests for ceilDiv
* Fix comment in ceilDiv
* Calling ceilDiv with int type T such like sizeof(T) > 8 is error
* Only allow IPv4 literals in strict form
The strict form as defined in RFC 6943, section 3.1.1 only allows the dotted
form ddd.ddd.ddd.ddd of IPv4 literals, where ddd is a one to three digit decimal
number between 0 and 255. Until now octal numbers (with a leading zero) were
interpreted as decimal numbers which has security implications, see
CVE-2021-29922 and CVE-2021-29923.
* Update lib/pure/net.nim
Co-authored-by: Dominik Picheta <dominikpicheta@googlemail.com>
* Remove unnecessary environment tracking
* try to fix windows
* fix delEnv
* make putEnv work on windows even with empty values; improve tests: add tests, add js, vm testing
* [skip ci] fix changelog
Co-authored-by: Caden Haustein <code@brightlysalty.33mail.com>
* Replace calls to C `putenv` with C `setenv` to remove possible memory leaks
* Add test of correct behaviour on invalid input
* Fix style in tests/stdlib/tos.nim
Co-authored-by: Timothee Cour <timothee.cour2@gmail.com>
* Update tests/stdlib/tos.nim
Co-authored-by: Timothee Cour <timothee.cour2@gmail.com>
* Update tests/stdlib/tos.nim
Co-authored-by: Timothee Cour <timothee.cour2@gmail.com>
* Add comment with bug number to tests/stdlib/tos.nim
Co-authored-by: Timothee Cour <timothee.cour2@gmail.com>
* Fix possible msvc arch issues
Co-authored-by: Timothee Cour <timothee.cour2@gmail.com>
* compiler/vmhooks: add getVar to allow vmops with var params
* addFloat vmops with var param
* cgen now renders float32 literals in c backend using roundtrip float to string
* Revert "followup #17398: `getTempDir`, `getConfigDir` now do not have trailing DirSep (#17402)"
This reverts commit 2356d0603f.
* Revert "fix #17393 getHomeDir and expandTilde should not include trailing `/` (#17398)"
This reverts commit bebf2ce24a.
* fix test
Implemented support for Nimble local cache with package directories with
a checksum of the package at the end of their names. Now the compiler
supports package paths in the form:
* /path_to_nimble_cache_dir/pkgs/package_name-1.2.3-
FEBADEAEA2345E777F0F6F8433F7F0A52EDD5D1B
* /path_to_nimble_cache_dir/pkgs/package_name-#head-
042D4BE2B90ED0672E717D71850ABDB0A2D19CD2
* /path_to_nimble_cache_dir/pkgs/package_name-#branch-name-
DBC1F902CB79946E990E38AF51F0BAD36ACFABD9
Related to nim-lang/nimble#127
* deprecate sequtils.delete and add an overload with saner semantics
* AssertionDefect => IndexDefect
* improve tsequtils
* add tests; use splice in js for optimization