| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
* fix #15910
* produce op of op is nil
* Trigger build
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
[backport]
* Fix #8404 JS backend doesn't handle float->int type conversion
* handle conv to uint as cast, discard other cases
* limit to int32, times use int64
* toInt including tyInt64 break times timezones lib, ignore for now
* also affect to vm
* move to tests/misc/t8404.nim
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Update installer.ini
Add support for powerpc64 little endian
* Update buildsh.nimf
Add support for powerpc64 little endian
* Update makefile.nimf
Add support for freebsd/powerpc64 little endian
|
|
|
|
|
|
|
| |
* fix #15916
* add testcase for #15916
* add comments
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* fix #14157
* Update compiler/jsgen.nim
* add changelog
* Update compiler/jsgen.nim
* Update tests/js/tmodify_cstring.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
|
|
|
|
|
|
|
|
|
| |
* fix #14127 from int to int casting
* add test for #14127
* use template for test, also test uint2int
* move to tests/types/t14127_cast_number.nim targets:c cpp js
|
|
|
|
| |
* "eg" is a misspelled "egg", "e.g." is "exempli gratia"
* Also, "ie" is "i.e.".
|
|
|
|
|
|
|
| |
* rst: add support for markdown tables
* change template into proc
* don't create unnecessary `seq[string]`
|
|
|
|
|
| |
* fix #15825
* better fix
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* new: `nim -i cmd`
* rename -i to -e (for eval); consistent with majority of other programing languages
* `nim e -e:cmd` now works; bugfix: `echo cmd | nim e -` now works
* honor --betterRun
* address comments
* --eval alias for -e (replaces undocumented --eval which was a noop)
* --eval now defaults to e (nimscript) instead of r
* address comment: remove -e, only keep --eval
* address comment
* fixup
* Update compiler/nimconf.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
|
|
|
|
|
|
|
|
|
|
|
| |
* fix #15858
* fix space
* fix #15629
* Revert "fix space"
* Revert "fix #15858"
|
|
|
|
|
|
| |
* fix #15858
* fix space
* fix #15629
* Revert "fix #15629"
|
| |
|
|
|
|
| |
* fix #12558
* Update compiler/pragmas.nim
|
| |
|
|
|
|
|
| |
* fix #15706 Underflow not detected when using dec on distinct ranges
* fix #15846 first, jsgen even doesn't care of range in this op
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make `requiresInit` pragma to work for distinct types in addition to
objects. Tagging of distinct types with `requiresInit` pragma was
already supported, but its impact wasn't applied. Now its behavior when
applied on distinct types is as follows.
Given the following distinct type definitions:
```nim
type
DistinctObject {.requiresInit, borrow: `.`.} = distinct MyObject
DistinctString {.requiresInit.} = distinct string
```
The following code blocks will fail to compile:
```nim
var foo: DistinctFoo
foo.x = "test"
doAssert foo.x == "test"
```
```nim
var s: DistinctString
s = "test"
doAssert s == "test"
```
But these ones will compile successfully:
```nim
let foo = DistinctFoo(Foo(x: "test"))
doAssert foo.x == "test"
```
```nim
let s = "test"
doAssert s == "test"
```
|
|
|
|
|
|
|
|
|
| |
* close #9679
* close #7546
* close #9520
* close #6177
|
|
|
|
|
| |
* fix #15609
* fix test
|
|
|
|
|
|
|
|
|
| |
* close #11142
* fix #12636
* undo unwanted changes
* fix illegal recursion case
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
* add parseEnumRange
* fix runnable example
* update changelog
* use parseEnumRange in compiler
* reorganise code
* add changelog, make single normalizer argument
|
| |
|
|
|
|
|
|
|
|
|
| |
* follow #15818 and close #7109
* Update compiler/jsgen.nim
Co-authored-by: Juan Carlos <juancarlospaco@gmail.com>
Co-authored-by: Juan Carlos <juancarlospaco@gmail.com>
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
* fix
* minor
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
`--threads:off` mode ) pragmas supports for JS backend (#15772)
* add global pragma support for js backend
* globalThis
* add support for threadvar
* more tests
* Update compiler/jsgen.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
|
|
|
|
|
| |
* canAlias improvement
* fix comment
|
|
|
|
|
| |
* fix #7658
* more
|
| |
|
|
|
|
|
| |
* fix #15756
* simplify test
|
|
|
|
| |
(#15697)
|
| |
|
|
|
|
|
|
|
| |
show `kind` with --declaredLocs (#15673)
* honor --declaredLocs in more places, including type mismatch errors
* fix tests
* show declaration location also when type mismatch names clash
|
| |
|
| |
|
| |
|
|
|
|
| |
--experimental:views [backport:1.4] (#15737)
|
|
|
| |
When calling procs from Nim in Nimscript you could hit the VM iteration limit even though the code is functioning properly. This resolves that by making the iteration limit reset eachtime you call a proc.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* fix #15704 #15597 wrong VM register was freed
* same treatment for nkCheckedFieldExpr
* note concerning HighRegisterPressure
* bump NimPatch
* Update lib/system.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
|