[stdlib] Implement Int.from_bytes() and Int.as_bytes()
#3768
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
msaelices
changed the title
Int.from_bytes()Int.from_bytes() and Int.as_bytes()
Similar to the Python int.from_bytes() one Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…ular#3770) Some new examples that are being added are being tested as standalone `magic` projects, not with `lit`. Ignore these files from the lit config. We may need to also add tests in CI for these separately, that could be a followup once we have some examples in place. Signed-off-by: Alex Trotta <atrotta@modular.com> Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…3771) Lit config doesn't support wildcards _or_ directories, but it does support entire folders, so here we're just searching for every folder under `examples` and adding it to the exclude. I ran a quick test by copying `hello_interop.mojo` to `examples/project` (and renaming to `hello_interop2.mojo` to ignore the current excludes), then running `magic run examples`. The new file was correctly ignored. Signed-off-by: Alex Trotta <atrotta@modular.com> Signed-off-by: Manuel Saelices <msaelices@gmail.com>
functions Also change to use `stack_allocation()` inside of `check_and_get_or_convert_arg()` -- this simplifies the code needed on the caller side. MODULAR_ORIG_COMMIT_REV_ID: c9ec14f33bac7d1c464058dd78ce11f448afb527 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
* New Mojo manual page documenting operators and dunder methods * An accompanying tutorial showing how to operators for a custom Mojo type * Complete example code for the tutorial MODULAR_ORIG_COMMIT_REV_ID: 40762247b51da9d5b2b958738c846a504d538248 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: 86e3e1b9f8ba59d1012f88d5da528ca6f165f30c Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: c3553f83d8ff65680949a3119a312aeed7c3fa24 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
consistency MODULAR_ORIG_COMMIT_REV_ID: 48e5472978812d786082d4c5906daf844e8a5da0 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…1401 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: e6882d6be8b161bf493fd1dafe0c727f4f4cb428 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: cb3ebf296399f5166893aa5d6e7d5c8560bdfb49 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: 85797bf247f55c80748fc56d468a9d8c2e0e2767 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…1405 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
[External] [stdlib] Fix `math.math.mojo` examples Part of modular#3572. Co-authored-by: soraros <soraros@users.noreply.github.com> Closes modular#3763 MODULAR_ORIG_COMMIT_REV_ID: 9b499c7e23f0fc7cccc3ab3ff75326364d032fe8 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
[External] [stdlib] Clean up `b64encode` (2/N) Co-authored-by: soraros <soraros@users.noreply.github.com> Closes modular#3746 MODULAR_ORIG_COMMIT_REV_ID: e5bf916a6cc953c18bcb23b482de4a86778d5f52 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
[External] [stdlib] Make `StringRef` Representable - [X] Implement `StringRef.__repr__()` - [X] Implement `Writable` for `os.stat_result` Co-authored-by: Manuel Saelices <msaelices@gmail.com> Closes modular#3736 MODULAR_ORIG_COMMIT_REV_ID: f8fe27aa7b0274fa27665b8cd0b19af22f3cac3b Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This change passed tests initially but caused issues specifically on linux + Intel after being merged internally. Reverting fixes the issue for now while we investigate. MODULAR_ORIG_COMMIT_REV_ID: 0aa612d4d89883645c0649593f8f306d31d35268 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: 157fb1e14ce62552ea978397e3bc164925127ce4 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
MODULAR_ORIG_COMMIT_REV_ID: 151283eee2ad3d2b32a9a99c318fce87f2fb33f2 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…1505 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
variables
* Also removed various unused imports of related types.
The Mojo standard library currently contains a pseudo-global-variables
mechanism based on
a global hashmap. This allows one to use functionality that behaves like
global variables, despite
the fact that Mojo does not currently have complete language-level
support for global
variables (i.e. `var` statements at the top level scope).
The current interface in the standard library is very low-level, error
prone, and requires juggling
`OpaquePointer` values.
With this new interface, declaring a global variable containing e.g. a
String looks like:
```mojo
alias MY_GLOBAL = _Global["MY_GLOBAL", String, _init_my_global]
fn _init_my_global() -> String:
return "default value"
```
This new interface adds the following constraints that make it easier to
reason
about and use:
* The data type stored by the global variable must be declared.
* The initializer function cannot take any arguments (no "payload" data)
* The global variable is de-initialized by calling the data type's
`__del__()` method (not an arbitrary function).
MODULAR_ORIG_COMMIT_REV_ID: 40575002cd16e7ecb931a0143102f3956dfb8ef7
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This change passed tests initially but caused issues specifically on linux + Intel after being merged internally. Reverting fixes the issue for now while we investigate. MODULAR_ORIG_COMMIT_REV_ID: 0aa612d4d89883645c0649593f8f306d31d35268 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
variables
* Also removed various unused imports of related types.
The Mojo standard library currently contains a pseudo-global-variables
mechanism based on
a global hashmap. This allows one to use functionality that behaves like
global variables, despite
the fact that Mojo does not currently have complete language-level
support for global
variables (i.e. `var` statements at the top level scope).
The current interface in the standard library is very low-level, error
prone, and requires juggling
`OpaquePointer` values.
With this new interface, declaring a global variable containing e.g. a
String looks like:
```mojo
alias MY_GLOBAL = _Global["MY_GLOBAL", String, _init_my_global]
fn _init_my_global() -> String:
return "default value"
```
This new interface adds the following constraints that make it easier to
reason
about and use:
* The data type stored by the global variable must be declared.
* The initializer function cannot take any arguments (no "payload" data)
* The global variable is de-initialized by calling the data type's
`__del__()` method (not an arbitrary function).
MODULAR_ORIG_COMMIT_REV_ID: 40575002cd16e7ecb931a0143102f3956dfb8ef7
…1606 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This refactors some implicit conversion emission logic in the compiler's OverloadSet processing logic to follow the `CallSyntax` enum instead of using a redundant `allowImplicitConversions` boolean. Beyond unifying the code this exposes a problem where we allowed implicit conversions from implicit conversions (invalid "a->b->c" conversion) in one case. This fixes that, which requires tightening up some logic in the standard library. MODULAR_ORIG_COMMIT_REV_ID: fe26e7bfdb8d36b36f9537d5370a142076a58feb
…1705 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This treats `ref x: Foo` the same as `ref [_] x: Foo`, inferring the origin from the call site. This was discussed in the arg convention resyntaxing thread. MODULAR_ORIG_COMMIT_REV_ID: 6a9d97b8ef01b87dab972a3e783315ac3db8ff12
This moves from `self: Self` to just `self` to clean things up. MODULAR_ORIG_COMMIT_REV_ID: 1a26d34dcb9cc22b9d8a9da0e94402a2730e786c Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This adds `@implicit` to all single-argument constructors to enable explicit by default construction. Implicit conversions are used heavily throughout the codebase, and so this is a required step to get it to compile. This allows us to take off `@implicit` constructors where they don't make sense, and fix the compiler errors one step at a time. MODULAR_ORIG_COMMIT_REV_ID: 931a2f38a0f9fe87007e334158f1fdf399f04bb2 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Require `@implicit` decorator above single argument `__init__` function to allow it to be implicitly constructed. See changelog entry in this commit for an example and more details. This fixes modular#1310 MODULAR_ORIG_COMMIT_REV_ID: ace1a1da978fd7acbd9a8ea60d807448f8889fcf
And slight rewording. MODULAR_ORIG_COMMIT_REV_ID: cc40b7c83b539274076c83197a704c8367fc5fac Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…1816 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…ctions (#47479) [External] [stdlib] Add full unicode support for character casing functions The code I used to generate the lookup tables can be found here https://gist.github.com/mzaks/bbadaeebcf81a5200021af041568b26b Co-authored-by: Maxim Zaks <maxim.zaks@gmail.com> Closes modular#3496 MODULAR_ORIG_COMMIT_REV_ID: 971d447e40e3221b84df526393f984b7fe22fcba Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Since `Span` is just a pointer and length, it's a trivial type, and `StringSlice` is just a wrapper around `Span`. This should provide a small speedup. MODULAR_ORIG_COMMIT_REV_ID: 6ce887ad907e0851f353a9ee9f3cb84eaa53c2cb Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Added conversion from float8 to float32 MODULAR_ORIG_COMMIT_REV_ID: 441ab0aaac7dbf2235f26dc14f0c38bc60f12c64 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Makes mem-ops in memory.mojo more generic. MODULAR_ORIG_COMMIT_REV_ID: 9ac40b0fb895905026ff386e194954bfb130d498 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
using llvm_intrinsic instead of _call_libm on AMD GPUs MODULAR_ORIG_COMMIT_REV_ID: bbd8e506d72f335a62425dbfb62f94b72ec339f9 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
`__ceildiv__`.
The traits currently require `__floordiv__`, `__rfloordiv__`, and
`__neg__`. Modelling `CieldDivable{Raising}` using those traits ignores
unsigned numeric types where `__neg__` doesn't make sense. Having the
types implement `__ceildiv__` directly reduces the change of
implementing it wrong.
MODULAR_ORIG_COMMIT_REV_ID: f043dc093244082d03eb643c8080341c009b8855
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
Adds support for more GPUs to the sleep function. MODULAR_ORIG_COMMIT_REV_ID: c1e1f113c7f51e1386f3bbce5860c4e778e8cdf5 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
…2020 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This refactors some implicit conversion emission logic in the compiler's OverloadSet processing logic to follow the `CallSyntax` enum instead of using a redundant `allowImplicitConversions` boolean. Beyond unifying the code this exposes a problem where we allowed implicit conversions from implicit conversions (invalid "a->b->c" conversion) in one case. This fixes that, which requires tightening up some logic in the standard library. MODULAR_ORIG_COMMIT_REV_ID: fe26e7bfdb8d36b36f9537d5370a142076a58feb Signed-off-by: Manuel Saelices <msaelices@gmail.com>
This adds `@implicit` to all single-argument constructors to enable explicit by default construction. Implicit conversions are used heavily throughout the codebase, and so this is a required step to get it to compile. This allows us to take off `@implicit` constructors where they don't make sense, and fix the compiler errors one step at a time. MODULAR_ORIG_COMMIT_REV_ID: 931a2f38a0f9fe87007e334158f1fdf399f04bb2 Signed-off-by: Manuel Saelices <msaelices@gmail.com>
msaelices
force-pushed
the
int-from-bytes
branch
from
8ba3d1a to
43fa507
Compare
Signed-off-by: Manuel Saelices <msaelices@gmail.com>
|
Closed in favor of #3795 This PR ends up in a mess because of conflicts with the nightly branch |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Similar to the Python's int.from_bytes() and int.to_bytes() one.