# Description
I've made the panic reproducible in test case
`workspace::tests::quoted_command_reference_in_workspace`.
This PR fixes that by parsing + merging 1 more time, IMO it's a small
price to pay for workspace-wide heavy requests.
# User-Facing Changes
bug fix
# Tests + Formatting
made 1 case harder
# After Submitting
# Description
This pull request does a lot of the heavy lifting needed to supported
more complex dtypes like categorical dtypes. It introduces a new
CustomValue, NuDataType and makes NuSchema a full CustomValue. Further
more it introduces a new command `polars into-dtype` that allows a dtype
to be created. This can then be passed into schemas when they are
created.
```nu
> ❯ : let dt = ("str" | polars to-dtype)
> ❯ : [[a b]; ["one" "two"]] | polars into-df -s {a: $dt, b: str} | polars schema
╭───┬─────╮
│ a │ str │
│ b │ str │
╰───┴─────╯
```
# User-Facing Changes
- Introduces new command `polars into-dtype`, allows dtype variables to
be passed in during schema creation.
Bumps [indexmap](https://github.com/indexmap-rs/indexmap) from 2.8.0 to
2.9.0.
<details>
<summary>Changelog</summary>
<p><em>Sourced from <a
href="https://github.com/indexmap-rs/indexmap/blob/main/RELEASES.md">indexmap's
changelog</a>.</em></p>
<blockquote>
<h2>2.9.0 (2025-04-04)</h2>
<ul>
<li>Added a <code>get_disjoint_mut</code> method to
<code>IndexMap</code>, matching Rust 1.86's
<code>HashMap</code> method.</li>
<li>Added a <code>get_disjoint_indices_mut</code> method to
<code>IndexMap</code> and <code>map::Slice</code>,
matching Rust 1.86's <code>get_disjoint_mut</code> method on
slices.</li>
<li>Deprecated the <code>borsh</code> feature in favor of their own
<code>indexmap</code> feature,
solving a cyclic dependency that occured via
<code>borsh-derive</code>.</li>
</ul>
</blockquote>
</details>
<details>
<summary>Commits</summary>
<ul>
<li><a
href="1818d4140d"><code>1818d41</code></a>
Merge pull request <a
href="https://redirect.github.com/indexmap-rs/indexmap/issues/387">#387</a>
from cuviper/release-2.9.0</li>
<li><a
href="9f4998341b"><code>9f49983</code></a>
Release 2.9.0</li>
<li><a
href="582a90fda3"><code>582a90f</code></a>
Merge pull request <a
href="https://redirect.github.com/indexmap-rs/indexmap/issues/386">#386</a>
from cuviper/de-borsh</li>
<li><a
href="90117397b6"><code>9011739</code></a>
Deprecate the "borsh" feature</li>
<li><a
href="0a836e8648"><code>0a836e8</code></a>
Merge pull request <a
href="https://redirect.github.com/indexmap-rs/indexmap/issues/238">#238</a>
from NiklasJonsson/get_many_mut</li>
<li><a
href="434d7ac6d1"><code>434d7ac</code></a>
Avoid let-else for MSRV's sake</li>
<li><a
href="5be552d557"><code>5be552d</code></a>
Implement additional suggestions from review</li>
<li><a
href="4e1d8cef47"><code>4e1d8ce</code></a>
Address review feedback</li>
<li><a
href="5aec9ec674"><code>5aec9ec</code></a>
Implement get_disjoint_mut for arrays of keys</li>
<li><a
href="d10de30e74"><code>d10de30</code></a>
Merge pull request <a
href="https://redirect.github.com/indexmap-rs/indexmap/issues/385">#385</a>
from iajoiner/docs/macros</li>
<li>Additional commits viewable in <a
href="https://github.com/indexmap-rs/indexmap/compare/2.8.0...2.9.0">compare
view</a></li>
</ul>
</details>
<br />
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
I think after that we can close #14790
# Description
So the issue was the tiny time delta between the moment the "date
form-human" command is executed, and the moment the value gets
displayed, using chrono_humanize.
When in inputing "in 30 seconds", we currently get:
```
[crates\nu-protocol\src\value\mod.rs:950:21] HumanTime::from(*val) = HumanTime(
TimeDelta {
secs: 29,
nanos: 992402700,
},
)```
And with "now":
```
crates\nu-protocol\src\value\mod.rs:950:21] HumanTime::from(*val) =
HumanTime(
TimeDelta {
secs: -1,
nanos: 993393200,
},
)
```
My solution is to round this timedelta to seconds and pass this to chrono_humanize.
Example: instead of passing (-1s + 993393200ns), we pass 0s.
Example: instead of passing (29s + 992402700ns), we pass 30s
# User-Facing Changes
Before 🔴
```nushell
~> "in 3 days" | date from-human
Fri, 11 Apr 2025 09:06:36 +0200 (in 2 days)
~> "in 30 seconds" | date from-human
Tue, 8 Apr 2025 09:07:09 +0200 (in 29 seconds)
```
After those changes 🟢
```nushell
~> "in 3 days" | date from-human
Fri, 11 Apr 2025 09:03:47 +0200 (in 3 days)
~> "in 30 seconds" | date from-human
Tue, 8 Apr 2025 09:04:28 +0200 (in 30 seconds)
```
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make sure to [enable developer mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the documentation](https://github.com/nushell/nushell.github.io) after the PR is merged, if necessary. This will help us keep the docs up to date. -->
# Description
Now, with PWD correctly set in #15470 , identifiers in
`use/hide/overlay` commands can be identified using a more robust
method, i.e. module_id from `parser_info`.
# User-Facing Changes
bug fix
# Tests + Formatting
+1 (fails without this PR)
# After Submitting
# Description
The current implementation of `polars into-df` and `polars into-lazy`
will throw an error if `--schema` is provided but not all columns are
defined. This PR seeks to remove this requirement so that when a partial
`--schema` is provided, the types on the defined columns are overridden
while the remaining columns take on their default types.
**Current Implementation**
```
$ [[a b]; [1 "foo"] [2 "bar"]] | polars into-df -s {a: str} | polars schema
Error: × Schema does not contain column: b
╭─[entry #88:1:12]
1 │ [[a b]; [1 "foo"] [2 "bar"]] | polars into-df -s {a: str} | polars schema
· ─────
╰────
```
**New Implementation (no error thrown on partial schema definition)**
Column b is not defined in `--schema`
```
$ [[a b]; [1 "foo"] [2 "bar"]] | polars into-df --schema {a: str} | polars schema
╭───┬─────╮
│ a │ str │
│ b │ str │
╰───┴─────╯
```
# User-Facing Changes
Soft breaking change: The user's previous (erroneous) code that would
have thrown an error would no longer throw an error. The user's previous
working code will still work.
# Tests + Formatting
# After Submitting
No related issue.
Decided in nushell's weekly meeting: see [meeting
notes](https://hackmd.io/rA1YecqjRh6I5m8dTq7BHw)
# Description
Converting a date as a human readable string to a datetime:
- currently: using the ``into datetime`` command
- after this change: using ``date from-human`` command
Also moved the ``--list-human`` flag to the new command.
# User-Facing Changes
- Users have to use a new command for parsing human readable datetimes.
Result:
```nushell
~> date from-human --list
╭────┬───────────────────────────────────┬──────────────╮
│ # │ parseable human datetime examples │ result │
├────┼───────────────────────────────────┼──────────────┤
│ 0 │ Today 18:30 │ in 6 hours │
│ 1 │ 2022-11-07 13:25:30 │ 2 years ago │
│ 2 │ 15:20 Friday │ in 6 days │
│ 3 │ This Friday 17:00 │ in 6 days │
│ 4 │ 13:25, Next Tuesday │ in 3 days │
│ 5 │ Last Friday at 19:45 │ 16 hours ago │
│ 6 │ In 3 days │ in 2 days │
│ 7 │ In 2 hours │ in 2 hours │
│ 8 │ 10 hours and 5 minutes ago │ 10 hours ago │
│ 9 │ 1 years ago │ a year ago │
│ 10 │ A year ago │ a year ago │
│ 11 │ A month ago │ a month ago │
│ 12 │ A week ago │ a week ago │
│ 13 │ A day ago │ a day ago │
│ 14 │ An hour ago │ an hour ago │
│ 15 │ A minute ago │ a minute ago │
│ 16 │ A second ago │ now │
│ 17 │ Now │ now │
╰────┴───────────────────────────────────┴──────────────╯
~> "2 days ago" | date from-human
Thu, 3 Apr 2025 12:03:33 +0200 (2 days ago)
~> "2 days ago" | into datetime
Error: nu:🐚:datetime_parse_error
× Unable to parse datetime: [2 days ago].
╭─[entry #5:1:1]
1 │ "2 days ago" | into datetime
· ──────┬─────
· ╰── datetime parsing failed
╰────
help: Examples of supported inputs:
* "5 pm"
* "2020/12/4"
* "2020.12.04 22:10 +2"
* "2020-04-12 22:10:57 +02:00"
* "2020-04-12T22:10:57.213231+02:00"
* "Tue, 1 Jul 2003 10:52:37 +0200"
```
# Tests + Formatting
Fmt, clippy 🆗
Tests 🆗
> Note: I was able to reactivate one unit test in the ``into datetime``
command
# After Submitting
Here since the user facing changes are significant, I think we should
communicate in the released notes. Otherwise the automatically generated
documentation should be enough IMO.
Closes#15502
# Description
The parsing of Exbibytes used the wrong base unit before converting.
# User-Facing Changes
`1EiB` etc. will now be parsed correctly
# Tests + Formatting
(-)
sub-issue of #10698 according to @sholderbach
(Description largely edited, since the scope of the PR changed)
# Description
Context: `ShellError::OnlySupportsThisInputType` was a duplicate of
`ShellError::PipelineMismatch`
so I
- replaced some occurences of PipelineMismatch by
OnlySupportsThisInputType
For another PR
- replace the remaining occurences
- removed OnlySupportsThisInputType from nu-protocol
# User-Facing Changes
The error message will be different -> but consistent
# Tests + Formatting
OK
# After Submitting
Nothing required
#15499 reminds me of the discrepancies between lsp hover docs and
`--help` outputs.
# Description
# User-Facing Changes
Before:
<img width="610" alt="image"
src="https://github.com/user-attachments/assets/f73f7ace-5c1b-4380-9921-fb4783bdb187"
/>
After:
<img width="610" alt="image"
src="https://github.com/user-attachments/assets/96de3ffe-e37b-41b1-88bb-123eeb72ced2"
/>
Output of `if -h` as a reference:
```
Usage:
> if <cond> <then_block> (else <else_expression>)
Flags:
-h, --help: Display the help message for this command
Parameters:
cond <variable>: Condition to check.
then_block <block>: Block to run if check succeeds.
"else" + <one_of(block, expression)>: Expression or block to run when the condition is false. (optional)
```
# Tests + Formatting
Refined
# After Submitting
# Description
There are some clippy(version 0.1.86) errors on nushell repo. This pr is
trying to fix it.
# User-Facing Changes
Hopefully none.
# Tests + Formatting
NaN
# After Submitting
NaN
# Description
Some editors like neovim will provide "workspace root" as PWD, which can
mess up file completion results.
# User-Facing Changes
bug fix
# Tests + Formatting
adjusted
# After Submitting
# Description
This PR seeks to fix an error in `polars as-datetime` where timezone
information is entirely ignored. This behavior raises a host of silent
errors when dealing with datetime conversions (see example below).
## Current Implementation
Timezones are entirely ignored and datetimes with different timezones
are converted to the same naive datetimes even when the user
specifically indicates that the timezone should be parsed. For example,
"2021-12-30 00:00:00 +0000" and "2021-12-30 00:00:00 -0400" will both be
parsed to "2021-12-30 00:00:00" even when the format string specifically
includes "%z".
```
$ ["2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z"
╭───┬───────────────────────╮
│ # │ datetime │
├───┼───────────────────────┤
│ 0 │ 12/30/2021 12:00:00AM │
│ 1 │ 12/30/2021 12:00:00AM │ <-- Same datetime even though the first is +0000 and second is -0400
╰───┴───────────────────────╯
$ ["2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z" | polars schema
╭──────────┬──────────────╮
│ datetime │ datetime<ns> │
╰──────────┴──────────────╯
```
## New Implementation
Datetimes are converted to UTC and timezone information is retained.
```
$ "2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z"
╭───┬───────────────────────╮
│ # │ datetime │
├───┼───────────────────────┤
│ 0 │ 12/30/2021 12:00:00AM │
│ 1 │ 12/30/2021 04:00:00AM │ <-- Converted to UTC
╰───┴───────────────────────╯
$ ["2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z" | polars schema
╭──────────┬───────────────────╮
│ datetime │ datetime<ns, UTC> │
╰──────────┴───────────────────╯
```
The user may intentionally ignore timezone information by setting the
`--naive` flag.
```
$ ["2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z" --naive
╭───┬───────────────────────╮
│ # │ datetime │
├───┼───────────────────────┤
│ 0 │ 12/30/2021 12:00:00AM │
│ 1 │ 12/30/2021 12:00:00AM │ <-- the -0400 offset is ignored when --naive is set
╰───┴───────────────────────╯
$ ["2021-12-30 00:00:00 +0000" "2021-12-30 00:00:00 -0400"] | polars into-df | polars as-datetime "%Y-%m-%d %H:%M:%S %z" --naive | polars schema
╭──────────┬──────────────╮
│ datetime │ datetime<ns> │
╰──────────┴──────────────╯
```
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
`polars as-datetime` will now account for timezone information and
return type `datetime<ns,UTC>` rather than `datetime<ns>` by default.
The user can replicate the previous behavior by setting `--naive`.
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
Tests that incorporated `polars as-datetime` had to be tweaked to
include `--naive` flag to replicate previous behavior.
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
Fixes#15476
# Description
<!--
Thank you for improving Nushell. Please, check our [contributing
guide](../CONTRIBUTING.md) and talk to the core team before making major
changes.
Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.
-->
Consider PATH when checking for potential_nuscript_in_windows to allow
executing scripts which are in PATH without having to full path address
them. It previously only checked the current working directory so only
relative paths to cwd and full path worked.
The current implementation runs this then through cmd.exe /D /C which
can run it with assoc and ftype set for nushell scripts.
We could instead run it through nu as `std::env::current_exe()` avoiding
the cmd call and the need for assoc and ftype (see:
8b25173f02).
But ive left the current implementation for this intact to not change
implementation details, avoid a bigger change and leave this open for
discussion here since im not sure if this has any major implications.
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
This would now run every external command through PATH an additional
time on windows, so potentially twice. I dont think this has any bigger
effect.
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
# Description
Noticed there is a build failure in #15420, because `ShadowBuilder`
struct is guarded by `build` feature. This pr is going to update it.
# User-Facing Changes
Hopefully none.
# Tests + Formatting
None
# After Submitting
None
---------
Co-authored-by: Stefan Holderbach <sholderbach@users.noreply.github.com>
Issue #9887 which can be closed after this is merged.
# Description
This allows the "into duration" command to accept floats as inputs.
Examples:
<img width="767" alt="image"
src="https://github.com/user-attachments/assets/da181f2a-7ad6-4efb-a6db-f9c6d8929c71"
/>
<img width="710" alt="image"
src="https://github.com/user-attachments/assets/78623a39-33ad-42a0-9324-a147be86f95c"
/>
**How it works:**
Using strings, like `"1.234sec" | into duration`, is already working, so
if a user inputs `1.234 | into duration --sec`, I just convert this back
to a string and use the previous conversion functions.
**Limitations:**
there are some limitation to using floats, but it's a general limitation
that is already present for other use cases:
- only 3 digits are taken into account in the decimal part
- floating durations in nano seconds are always floored and not rounded
<img width="761" alt="image"
src="https://github.com/user-attachments/assets/a9076aab-da03-43f2-927c-c9703fc4f955"
/>
# User-Facing Changes
Users can inject floats with `into duration`
# Tests + Formatting
cargo fmt and clippy OK
Tests OK
# After Submitting
The example I added will automatically become part of the doc, I think
that's enough for documentation.
This should be a more robust method.
# Description
Previously, `export use` with double-space in between will fail to be
recognized as command `export use`.
# User-Facing Changes
minor bug fix
# Tests + Formatting
test cases made harder
# After Submitting
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
# Description
This PR expands the `dtype` parameter of the `polars cast` command to
include `decimal<precision, scale>` type. Setting precision to "*" will
compel inferring the value. Note, however, setting scale to a
non-integer value will throw an explicit error (the underlying polars
crate assigns scale = 0 in such a case, but I opted for throwing an
error instead). .
```
$ [[a b]; [1 2] [3 4]] | polars into-df | polars cast decimal<4,2> a | polars schema
╭───┬──────────────╮
│ a │ decimal<4,2> │
│ b │ i64 │
╰───┴──────────────╯
$ [[a b]; [10.5 2] [3.1 4]] | polars into-df | polars cast decimal<*,2> a | polars schema
╭───┬──────────────╮
│ a │ decimal<*,2> │
│ b │ i64 │
╰───┴──────────────╯
$ [[a b]; [10.05 2] [3.1 4]] | polars into-df | polars cast decimal<5,*> a | polars schema
rror: × Invalid polars data type
╭─[entry #25:1:47]
1 │ [[a b]; [10.05 2] [3.1 4]] | polars into-df | polars cast decimal<5,*> a | polars schema
· ─────┬─────
· ╰── `*` is not a permitted value for scale
╰────
```
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
There are no breaking changes. The user has the additional option to
`polars cast` to a decimal type
# Tests + Formatting
Tests have been added to
`nu_plugin_polars/src/dataframe/values/nu_schema.rs`
# Description
There's been much debate about whether to keep human-date-parser in
`into datetime`. We saw recently that a new version of the crate was
released that addressed some of our concerns. This PR is to make it
easier to test those fixes.
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
# Description
<!--
Thank you for improving Nushell. Please, check our [contributing
guide](../CONTRIBUTING.md) and talk to the core team before making major
changes.
Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.
-->
Follow-up to #15277 and #15392.
Adds examples to `any` and `all` demonstrating using `any {}` or `all
{}` with lists of booleans.
We have a couple options that work for this use-case, but not sure which
we should recommend. The PR currently uses (1).
1. `any {}` / `all {}`
2. `any { $in }` / `all { $in }`
3. `any { $in == true }` / `all { $in == true }`
Would love to hear your thoughts on the above @fennewald @mtimaN
@fdncred @NotTheDr01ds @ysthakur
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
* Added an extra example for `any` and `all`
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
N/A
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
N/A
# Description
As description, I think it's worth to move forward to update rand and
rand_chacha to 0.9.
# User-Facing Changes
Hopefully none
# Tests + Formatting
NaN
# After Submitting
NaN
# Description
```
# table.*
# table_mode (string):
# One of: "default", "basic", "compact", "compact_double", "heavy", "light", "none", "reinforced",
# "rounded", "thin", "with_love", "psql", "markdown", "dots", "restructured", "ascii_rounded",
# or "basic_compact"
# Can be overridden by passing a table to `| table --theme/-t`
$env.config.table.mode = "default"
```
In `doc_config.nu`, it refers to `table_mode` which does not exist under
`$env.config.table`. There is now a short description of this field as
well.
# Description
Closes#14794. This PR enables the strict exact match behavior requested
in #13204 and #14794 for any path containing a slash (#13302 implemented
this for paths ending in slashes).
If any of the components along the way *don't* exactly match a
directory, then the next components will use the old Fish-like
completion behavior rather than the strict behavior.
This change only affects those using prefix matching. Fuzzy matching
remains unaffected.
# User-Facing Changes
Suppose you have the following directory structure:
```
- foo
- bar
- xyzzy
- barbaz
- xyzzy
- foobar
- bar
- xyzzy
- barbaz
- xyzzy
```
- If you type `cd foo<TAB>`, you will be suggested `[foo, foobar]`
- This is because `foo` is the last component of the path, so the strict
behavior isn't activated
- Similarly, `foo/bar` will show you `[foo/bar, foo/barbaz]`
- If you type `foo/bar/x`, you will be suggested `[foo/bar/xyzzy]`
- This is because `foo` and `bar` both exactly matched a directory
- If you type `foo/b/x`, you will be suggested `[foo/bar/xyzzy,
foo/barbaz/xyzzy]`
- This is because `foo` matches a directory exactly, so `foobar/*` won't
be suggested, but `b` doesn't exactly match a directory, so both `bar`
and `barbaz` are suggested
- If you type `f/b/x`, you will be suggested all four of the `xyzzy`
files above
- If you type `f/bar/x`, you will be suggested all four of the `xyzzy`
files above
- Since `f` doesn't exactly match a directory, every component after it
won't use the strict matching behavior (even though `bar` exactly
matches a directory)
# Tests + Formatting
# After Submitting
This is a pretty minor change but should be mentioned somewhere in the
release notes in case it surprises someone.
---------
Co-authored-by: 132ikl <132@ikl.sh>
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
Fixes#14794.
# Description
<!--
Thank you for improving Nushell. Please, check our [contributing
guide](../CONTRIBUTING.md) and talk to the core team before making major
changes.
Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.
-->
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
Makes it so that (even if) the command ends in a slash, exact matches
are still preferred over partial matches.
For example, `foo/bar/as` -> `foo/bar/asdf` but not `foo/bars/asdf`.
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
---------
Co-authored-by: Yash Thakur <45539777+ysthakur@users.noreply.github.com>
# Description
Add parse warnings to LSP diagnostics, not particularly useful but
technically should be done.
# User-Facing Changes
# Tests + Formatting
There's no deprecated command to test for now.
# After Submitting
Fixes#15414 by changing the method used to de-ansi-fy the input. Control characters will now be kept when using `clip copy`, but ANSI escape codes will be removed (when not using `--ansi (-a)`)
Fixes#15441
# Description
Actually I made a small change to the original behavior:
```
^foo<tab>
```
will still show external commands, regardless of whether it's enabled or
not. I think that's the only thing people want to see when they press
tab with a `^` prefix.
# User-Facing Changes
# Tests + Formatting
+1
# After Submitting
Should I document that minor behavior change somewhere in GitHub.io?
---------
Co-authored-by: Yash Thakur <45539777+ysthakur@users.noreply.github.com>
Close#15119 when this is merged
# Description
> Note: my locale is +1
**Before the changes 🔴**
See the issue for more detailed description of the problem.
**After the changes 🟢**
# User-Facing Changes
The ``into datetime`` command will now work with formatting and time
zones or offset together
# Tests + Formatting
Fmt + clippy OK
**Note about the tests I added**: those tests don't really test my
changes, as they were already passing before my changes. Nevertheless I
thought I could push them
# After Submitting
I don't think anything is necessary
No linked issue, it's a follow-up of 2 PRs I recently made to improve
some math commands. (#15319)
# Description
Small refactor to simplify the code. It was suggested in the comments of
my previous PR.
# User-Facing Changes
None
# Tests + Formatting
Tests, fmt and clippy OK
# After Submitting
Nothing more required
fixes#8095
# Description
This approach is a bit straightforward, call access() check with the
flag `X_OK`.
Zsh[^1], Fish perform this check by the same approach.
[^1]:
435cb1b748/Src/exec.c (L6406)
It could also avoid manual xattrs check on other *nix platforms.
BTW, the execution bit for directories in *nix world means permission to
access it's content,
while the read bit means to list it's content. [^0]
[^0]: https://superuser.com/a/169418
# User-Facing Changes
Users could face less permission check bugs in their `cd` usage.
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
---------
Co-authored-by: Stefan Holderbach <sholderbach@users.noreply.github.com>
Closes#15395
# User-Facing Changes
Certain errors no longer leave the argument stack in an unexpected
state:
```diff
let x: any = 1; try { $x | get path } catch { print caught }
-$.path # extra `print` argument from the failed `get` call
caught
```
# Description
If `eval_call` fails in `check_input_types` or `gather_arguments`, the
cleanup code is still executed.
Fixes#14972#15321#14706
# Description
Early returns `NotAConstant` if parsing errors exist in the
subexpression.
I'm not sure when the span of a block will be None, and whether there're
better ways to handle none block spans, like a more suitable ShellError
type.
# User-Facing Changes
# Tests + Formatting
+1, but possibly not the easiest way to do it.
# After Submitting
Closes#15305
# Description
Basically turns off `skip_comments` of the lex function for right hand
side expressions of `let`/`mut`, just as in `parse_const`.
# User-Facing Changes
Should be none.
# Tests + Formatting
+1
# After Submitting
`path add`, when given a record, sets `$env.PATH` according to the value
of the key matching `$nu.os-info.name`. There already existed a check in
place to ensure the correct column existed, but it was never reached
because of an early error on `path expand`ing `null`. This has been
fixed, as well as the out-of-date reference to "darwin" instead of
"macos" in the example.
# User-Facing Changes
`path add` now simply ignores a record that doesn't include a key for the current OS
`path add` also will no longer add duplicate paths.
We only have one valid `datetime` type, but the string representation of
that type was `date`. This PR updates the string representation of the
`datetime` type to be `datetime` and updates other affected
dependencies:
* A `describe` example that used `date`
* The style computer automatically recognized the new change, but also
changed the default `date: purple` to `datetime: purple`.
* Likewise, changed the `default_config.nu` to populate
`$env.config.color_config.datetime`
* Likewise, the dark and light themes in `std/config`
* Updates tests
* Unrelated, but changed the `into value` error messages to use
*"datetime"* if there's an issue.
Fixes#9916 and perhaps others.
## Breaking Changes:
* Code that expected `describe` to return a `date` will now return a
`datetime`
* User configs and themes that override `$env.config.color_config.date`
will need to be updated to use `datetime`
Closes#15373
# Description
Now `ast -f "{||}"` will return
```
╭─content─┬─────shape─────┬─────span──────╮
│ {||} │ shape_closure │ ╭───────┬───╮ │
│ │ │ │ start │ 0 │ │
│ │ │ │ end │ 4 │ │
│ │ │ ╰───────┴───╯ │
╰─────────┴───────────────┴───────────────╯
```
Similar to those of `ast -f "[]"`/`ast -f "{}"`
# User-Facing Changes
# Tests + Formatting
I didn't find the right place to do the test, except for the examples of
`ast` command.
# After Submitting
# Description
Closes#15351
Adds quotes that were missed in #14698 with the proper escaping.
# User-Facing Changes
`to nuon --serialize` will now produce a quoted string instead of
illegal nuon when given a closure
# Tests + Formatting
Reenable the `to nuon` rejection of closures in the base state test.
Added test for quoting.
# Description
This PR solves a circular dependency issue (`nu-test-support` needs
`nu-glob` which needs `nu-protocol` which needs `nu-test-support`). This
was done by making the glob functions that any type that implements
`Interruptible` to remove the dependency on `Signals`.
# After Submitting
Make `Paths.next()` a O(1) operation so that cancellation/interrupt
handling can be moved to the caller (e.g., by wrapping the `Paths`
iterator in a cancellation iterator).
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
# Description
<!--
Thank you for improving Nushell. Please, check our [contributing
guide](../CONTRIBUTING.md) and talk to the core team before making major
changes.
Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.
-->
Adds an `impl From<IoError> for LabeledError`, similar to the existing
`From<ShellError>` implementation. Helpful for plugins.
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
N/A
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
N/A
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
N/A
# Description
Some editors (like zed) will fail to mark the active parameter if not
set in the outmost structure.
# User-Facing Changes
# Tests + Formatting
Adjusted
# After Submitting
# Description
This PR adds a few more columns to the macos version of `ps -l` to bring
it more inline with the Linux and Windows version.
Columns added: user_id, priority, process_threads
I also added some comments that describe the TaskInfo structure. I
couldn't find any good information to add to the BSDInfo structure.
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
# Description
The `job unfreeze` command relies on the `os` feature of the
`nu-protocol` crate, which means that `nu-command` doesn't compile with
`--no-default-features`. This PR gates `job unfreeze` behind
`nu-command`'s `os` feature to avoid this.
No user-facing changes, no tests needed.
# Description
Follow-up to #15272, changing default to disallow DTD as discussed.
Especially applicable for the `http get` case.
# User-Facing Changes
Changes behavior introduced in #15272, so release notes need to be
updated to reflect this
Fixes messed ansi escapes in hover text (manpage):
<img width="392" alt="image"
src="https://github.com/user-attachments/assets/37c16520-d499-4079-93d9-0eccd1cfa8de"
/>
# Description
That bug is introduced in #15115.
Also refactored the hover related code to a separate file, just like
other features.
# User-Facing Changes
# Tests + Formatting
# After Submitting
# Description
`into string` should not modify input strings (even with the
`--group-digits` flag). It's a conversion command, not a formatting
command.
# User-Facing Changes
- For strings, the same behavior from 0.102.0 is preserved.
- Errors are no longer turned into strings, but rather they are returned
as is.
# After Submitting
Create a `format int` and/or `format float` command and so that the
`--group-digits` flag can be transferred to one of those commands.
# Description
Before this PR, `to msgpack`/`to msgpackz` and `to json` serialize
closures as `nil`/`null` respectively, when the `--serialize` option
isn't passed. This PR makes it an error to serialize closures to msgpack
or JSON without the `--serialize` flag, which is the behavior of `to
nuon`.
This PR also adds the `--serialize` flag to `to msgpack`.
This PR also changes `to nuon` and `to json` to return an error if they
cannot find the block contents of a closure, rather than serializing an
empty string or an error string, respectively. This behavior is
replicated for `to msgpack`.
It also changes `to nuon`'s error message for serializing closures
without `--serialize` to be the same as the new errors for `to json` and
`to msgpack`.
# User-Facing Changes
* Add `--serialize` flag to `to msgpack`, similar to the `--serialize`
flag for `to nuon` and `to json`.
* Serializing closures to JSON or msgpack without `--serialize`
Partially fixes#11738
While inspecting the Windows specific code of `ls` for #15311 I stumbled
upon an unrelated issue in the alternate metadata gathering on Windows
(added by #5703).
The handle created by performing `FindFirstFileW` was never closed,
leading to a potential leak. Fixed by running `FindClose` as soon as the
operation succeeds.
https://learn.microsoft.com/en-us/windows/win32/api/fileapi/nf-fileapi-findfirstfilew#remarks
# Description
This PR removes the mimalloc allocator due to run-away memory leaks
recently found.
closes#15311
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
# Description
Found inconsistent behaviors of `directory_completion` and
`file_completion`, https://github.com/nushell/nushell/issues/13951https://github.com/nushell/reedline/pull/886
Also there're failing cases with such file names/dir names `foo(`,
`foo{`, `foo[`.
I think it doesn't harm to be more conservative at adding quotes, even
if it might be unnecessary for paired names like `foo{}`.
# User-Facing Changes
# Tests + Formatting
Adjusted
# After Submitting
Came from [this
discussion](https://discord.com/channels/601130461678272522/1348791953784836147/1349699872059691038)
on discord with @fdncred
# Description
Small refactoring where I rename commands from "SubCommand" to its
proper name. Motivations: better clarity (although subjective), better
searchable, consistency.
The only commands I didn't touch were "split list" and "ansi gradient"
because of name clashes.
# User-Facing Changes
None
# Tests + Formatting
cargo fmt and clippy OK
# After Submitting
nothing required
# Description
Adds a new `--empty/-e` flag to the `default` command.
# User-Facing Changes
Before:
```nushell
$env.FOO = ""
$env.FOO = $env.FOO? | default bar
$env.FOO
# => Empty string
```
After:
```nushell
$env.FOO = ""
$env.FOO = $env.FOO? | default -e bar
$env.FOO
# => bar
```
* Uses `val.is_empty`, which means that empty lists and records are also
replaced
* Empty values in tables (with a column specifier) are also replaced.
# Tests + Formatting
7 tests added and 1 updated + 1 new example
- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`
# After Submitting
N/A
# fixes https://github.com/nushell/nushell/issues/15281
# Description
Provides the ability read dataframes with Categorical and Enum data
The ability to write Categorical and Enum data will provided in a future
PR
# Description
Follow up to https://github.com/nushell/nushell/pull/14634
# User-Facing Changes
`into bits` will be gone for good.
Use it under the new name `format bits`
## Note
Can be removed ahead of the `0.103.0` release as it was deprecated with
`0.102.0`
# Description
Follow up to https://github.com/nushell/nushell/pull/14875
# User-Facing Changes
`fmt` will be gone for good.
Use it under the new name `format number`
## Note
Can be removed ahead of the `0.103.0` release as it was deprecated with
`0.102.0`
Fixes#15077
# Description
Symlinks are currently not shown in directory completions. #14667
modified completions so that symlinks wouldn't be suggested with
trailing slashes, but it did this by treating symlinks as files. This PR
includes symlinks to directories when completing directories, but still
suggests them without trailing slashes.
# User-Facing Changes
Directory completions will once again include symlinks.
# Tests + Formatting
# After Submitting
<!--
if this PR closes one or more issues, you can automatically link the PR
with
them by using one of the [*linking
keywords*](https://docs.github.com/en/issues/tracking-your-work-with-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword),
e.g.
- this PR should close #xxxx
- fixes #xxxx
you can also mention related issues, PRs or discussions!
-->
# Description
<!--
Thank you for improving Nushell. Please, check our [contributing
guide](../CONTRIBUTING.md) and talk to the core team before making major
changes.
Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.
-->
This PR allows `from xml` to parse XML documents with [document type
declarations](https://en.wikipedia.org/wiki/Document_type_declaration)
by default. This is especially notable since many HTML documents start
with `<!DOCTYPE html>`, and `roxmltree` should be able to parse some
simple HTML documents. The security concerns with DTDs are [XXE
attacks](https://en.wikipedia.org/wiki/XML_external_entity_attack), and
[exponential entity expansion
attacks](https://en.wikipedia.org/wiki/Billion_laughs_attack).
`roxmltree` [doesn't
support](d2c7801624/src/tokenizer.rs (L535-L547))
external entities (it parses them, but doesn't do anything with them),
so it is not vulnerable to XXE attacks. Additionally, `roxmltree` has
[some
safeguards](d2c7801624/src/parse.rs (L424-L452))
in place to prevent exponential entity expansion, so enabling DTDs by
default is relatively safe. The worst case is no worse than running
`loop {}`, so I think allowing DTDs by default is best, and DTDs can
still be disabled with `--disallow-dtd` if needed.
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
* Allows `from xml` to parse XML documents with [document type
declarations](https://en.wikipedia.org/wiki/Document_type_declaration)
by default, and adds a `--disallow-dtd` flag to disallow parsing
documents with DTDs.
This PR also improves the errors in `from xml` by pointing at the issue
in the XML source. Example:
```
$ open --raw foo.xml | from xml
Error: × Failed to parse XML
╭─[2:7]
1 │ <html>
2 │ <p<>hi</p>
· ▲
· ╰── Unexpected character <, expected a whitespace
3 │ </html>
╰────
```
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
N/A
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
N/A
Fix failing test by ignoring the local offset when converting times, but still displaying the
resulting date in the local timezone (including applicable DST offset).
# User-Facing Changes
Fix: Unix Epochs now convert consistently regardless of whether DST is
in effect in the local timezone or not.
### Description
Fixes issue #15135
Result
Also this works with other commands: min, max, sum, product, avg...
### User-Facing Changes
Error is returned, instead of console completely blocked and having to
be killed
I chose "Incorrect value", because commands accept inputs of range type,
just cannot work with unbounded ranges.
### Tests + Formatting
- ran cargo fmt, clippy
- added tests
# Description
The upstream crate fixed a bug of position calc, which made some extra
checking in lsp unnecessary.
Also moved some follow-up fixing of #15238 from #15270 here, as it has
something to do with previous position calc bug.
# User-Facing Changes
# Tests + Formatting
Adjusted
# After Submitting
# Description
Continuation of #15271. This PR adds the
`$env.config.filesize.show_unit` option to allow the ability to omit the
filesize unit. Useful if `$env.config.filesize.unit` is set to a fixed
unit, and you don't want the same unit repeated over and over.
# User-Facing Changes
- Adds the `$env.config.filesize.show_unit` option.
# Description
Commands and other pieces of code using `$env.config.format.filesize` to
format filesizes now respect the system locale when formatting the
numeric portion of a file size.
# User-Facing Changes
- System locale is respected when using `$env.config.format.filesize` to
format file sizes.
- Formatting a file size with a binary unit is now exact for large file
sizes and units.
- The output of `to text` is no longer dependent on the config.
# Description
This PR allows the `into string` command to pass the `--group-digits`
flag which already existed in this code but was hard coded to `false`.
Now you can do things like
```nushell
❯ 1234567890 | into string --group-digits
1,234,567,890
❯ ls | into string size --group-digits | last 5
╭─#─┬────────name─────────┬─type─┬──size──┬───modified───╮
│ 0 │ README.md │ file │ 12,606 │ 4 weeks ago │
│ 1 │ rust-toolchain.toml │ file │ 1,125 │ 2 weeks ago │
│ 2 │ SECURITY.md │ file │ 2,712 │ 7 months ago │
│ 3 │ toolkit.nu │ file │ 21,929 │ 2 months ago │
│ 4 │ typos.toml │ file │ 542 │ 7 months ago │
╰─#─┴────────name─────────┴─type─┴──size──┴───modified───╯
❯ "12345" | into string --group-digits
12,345
```
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
# Tests + Formatting
<!--
Don't forget to add tests that cover your changes.
Make sure you've run and fixed any issues with these commands:
- `cargo fmt --all -- --check` to check standard code formatting (`cargo
fmt --all` applies these changes)
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `cargo run -- -c "use toolkit.nu; toolkit test stdlib"` to run the
tests for the standard library
> **Note**
> from `nushell` you can also use the `toolkit` as follows
> ```bash
> use toolkit.nu # or use an `env_change` hook to activate it
automatically
> toolkit check pr
> ```
-->
# After Submitting
<!-- If your PR had any user-facing changes, update [the
documentation](https://github.com/nushell/nushell.github.io) after the
PR is merged, if necessary. This will help us keep the docs up to date.
-->
# Description
As stated in the title, when pressing ctrl-z, I sometimes feel confused
because I return to the REPL without any message. I don't know if the
process has been killed or suspended.
This PR aims to add a message to notify the user that the process has
been frozen.
# User-Facing Changes
After pressing `ctrl-z`. A message will be printed in repl.
# Tests + Formatting
NaN
# After Submitting
NaN
# Description
This fixes#15240, which can be closed after merge.
# User-Facing Changes
- user get now use `to yml` -> exactly the same as `to yaml`
# Tests + Formatting
Cargo fmt and clippy 🆗
I added a test in the only place I could find where `to yaml` was
already tested.
I didn't see the `save.rs::convert_to_extension` function tested
anywhere, but maybe I missed it.
# After Submitting
Not sure this needs an update on the documentation ❓ What do you
suggest?
---------
Co-authored-by: Stefan Holderbach <sholderbach@users.noreply.github.com>
This PR implements the changes proposed in #15112 without any breaking
changes. Should close#15112 post the review.
# Description
Added functionality to generate `uuid` versions 1, 3, 4, 5, 7 instead of
just the version 4.
- Users can now add a `-v n` flag to specify the version of uuid they
want to generate and it maintains backward compatibility by returning a
v4 uuid by default if no flags are passed.
- Versions 3 and 5 have the additional but required namespace (`-s`) and
name (`-n`) arguments too. Version 1 requires a mac address (`-m`).
# User-Facing Changes
- Added support for uuid versions 1, 3, 5 and 7.
- For v3 and v5, the namespace and name arguments are required and hence
there will be an error if those are not passed. Similarly the mac
address for v1.
- Full backward compatibility by setting v4 as default.
# Tests + Formatting
Tests added:
in `nu-command::commands::random`
- generates_valid_uuid4_by_default
- generates_valid_uuid1
- generates_valid_uuid3_with_namespace_and_name
- generates_valid_uuid4
- generates_valid_uuid5_with_namespace_and_name
- generates_valid_uuid7
