<!--
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!
-->
This PR should close#8036, #9028 (in the negative) and #9118.
Fix for #9118 is a bit pedantic. As reported, the issue is:
```
> 2023-05-07T04:08:45+12:00 - 2019-05-10T09:59:12+12:00
3yr 12month 2day 18hr 9min 33sec
```
with this PR, you now get:
```
> 2023-05-07T04:08:45+12:00 - 2019-05-10T09:59:12+12:00
208wk 1day 18hr 9min 33sec
```
Which is strictly correct, but could still fairly be called "weird date
arithmetic".
# Description
* [x] Abide by constraint that Value::Duration remains a number of
nanoseconds with no additional fields.
* [x] `to_string()` only displays weeks .. nanoseconds. Duration doesn't
have base date to compute months or years from.
* [x] `duration | into record` likewise only has fields for weeks ..
nanoseconds.
* [x] `string | into duration` now accepts compound form of duration
to_string() (e.g '2day 3hr`, not just '2day')
* [x] `duration | into string` now works (and produces the same
representation as to_string(), which may be compound).
# User-Facing Changes
## duration -> string -> duration
Now you can "round trip" an arbitrary duration value: convert it to a
string that may include multiple time units (a "compound" value), then
convert that string back into a duration. This required changes to
`string | into duration` and the addition of `duration | into string'.
```
> 2day + 3hr
2day 3hr # the "to_string()" representation (in this case, a compound value)
> 2day + 3hr | into string
2day 3hr # string value
> 2day + 3hr | into string | into duration
2day 3hr # round-trip duration -> string -> duration
```
Note that `to nuon` and `from nuon` already round-tripped durations, but
use a different string representation.
## potentially breaking changes
* string rendering of a duration no longer has 'yr' or 'month' phrases.
* record from `duration | into record` no longer has 'year' or 'month'
fields.
The excess duration is all lumped into the `week` field, which is the
largest time unit you can
convert to without knowing the datetime from which the duration was
calculated.
Scripts that depended on month or year time units on output will need to
be changed.
### Examples
```
> 365day
52wk 1day
## Used to be:
## 1yr
> 365day | into record
╭──────┬────╮
│ week │ 52 │
│ day │ 1 │
│ sign │ + │
╰──────┴────╯
## used to be:
##╭──────┬───╮
##│ year │ 1 │
##│ sign │ + │
##╰──────┴───╯
> (365day + 4wk + 5day + 6hr + 7min + 8sec + 9ms + 10us + 11ns)
56wk 6day 6hr 7min 8sec 9ms 10µs 11ns
## used to be:
## 1yr 1month 3day 6hr 7min 8sec 9ms 10µs 11ns
## which looks reasonable, but was actually only correct in 75% of the years and 25% of the months in the last 4 years.
> (365day + 4wk + 5day + 6hr + 7min + 8sec + 9ms + 10us + 11ns) | into record
╭─────────────┬────╮
│ week │ 56 │
│ day │ 6 │
│ hour │ 6 │
│ minute │ 7 │
│ second │ 8 │
│ millisecond │ 9 │
│ microsecond │ 10 │
│ nanosecond │ 11 │
│ sign │ + │
╰─────────────┴────╯
```
Strictly speaking, these changes could break an existing user script.
Losing years and months as time units is arguably a regression in
behavior.
Also, the corrected duration calculation could break an existing script
that was calibrated using the old algorithm.
# Tests + Formatting
```
> toolkit check pr
```
- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`
# 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: Bob Hyman <bobhy@localhost.localdomain>
# Description
This PR changes `Value::columns` to return a slice of columns instead of
cloning said columns. If the caller needs an owned copy, they can use
`slice::to_vec` or the like. This eliminates unnecessary Vec clones
(e.g., in `update cells`).
# User-Facing Changes
Breaking change for `nu_protocol` API.
# Description
`Span` is `Copy`, so we probably should not be passing references of
`Span` around. This PR replaces all instances of `&Span` with `Span`,
copying spans where necessary.
# User-Facing Changes
This alters some public functions to take `Span` instead of `&Span` as
input. Namely, `EngineState::get_span_contents`,
`nu_protocol::extract_value`, a bunch of the math commands, and
`Gstat::gstat`.
# Description
If we reach the conclusion that the fields of a list are of `Type::Any`
there is no need to continue as the type will remain `Type::Any`
This should improve runtimes of `Value.get_type()` for lists with mixed
types.
# User-Facing Changes
None, a speedup in some cases.
# Tests + Formatting
Relies on existing tests
# Description
This PR ensures functions exist to extract and create each and every
`Value` case. It also renames `Value::boolean` to `Value::bool` to match
`Value::test_bool`, `Value::as_bool`, and `Value::Bool`. Similarly,
`Value::as_integer` was renamed to `Value::as_int` to be consistent with
`Value::int`, `Value::test_int`, and `Value::Int`. These two renames can
be undone if necessary.
# User-Facing Changes
No user facing changes, but two public functions were renamed which may
affect downstream dependents.
# Description
- A new one is the removal of unnecessary `#` in raw strings without `"`
inside.
-
https://rust-lang.github.io/rust-clippy/master/index.html#/needless_raw_string_hashes
- The automatically applied removal of `.into_iter()` touched several
places where #9648 will change to the use of the record API. If
necessary I can remove them @IanManske to avoid churn with this PR.
- Manually applied `.try_fold` in two places
- Removed a dead `if`
- Manual: Combat rightward-drift with early return
# Description
A couple clippy 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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.
-->
- fixes#9501
# Description
`chrono::Datetime::to_rfc2822()` panics when it would format a negative
year. 3339 does not. This makes negative-year datetimes inconsistent
compared to positive ones, but it's better than a panic.
- improves usability of datetime's in displayed text
-
# Description
Creates a config point for specifying long / short date time formats.
Defaults to humanized as we have today.
Provides for adding strftime formats into config.nu such as:
```nu
datetime_format: {
normal: "%Y-%m-%d %H:%M:%S"
table: "%Y-%m-%d"
}
```
Example:
```bash
> $env.config.datetime_format
┏━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ normal ┃ %a, %d %b %Y %H:%M:%S %z ┃
┃ table ┃ %m/%d/%y %I:%M:%S%p ┃
┗━━━━━━━━┻━━━━━━━━━━━━━━━━━━━━━━━━━━┛
> let a = (date now)
> echo $a
Thu, 22 Jun 2023 10:21:23 -0700
> echo [$a]
┏━━━┳━━━━━━━━━━━━━━━━━━━━━┓
┃ 0 ┃ 06/22/23 10:21:23AM ┃
┗━━━┻━━━━━━━━━━━━━━━━━━━━━┛
```
# User-Facing Changes
Any place converting a datetime to a user displayed value should be
impacted.
# Tests + Formatting
- `cargo fmt --all -- --check` Done
- `cargo clippy --workspace -- -D warnings -D clippy::unwrap_used -A
clippy::needless_collect -A clippy::result_large_err` Done
- `cargo test --workspace` Done
- `cargo run -- crates/nu-std/tests/run.nu` Not done - doesn't seem to
work
```bash
> use toolkit.nu # or use an `env_change` hook to activate it automatically
> toolkit check pr
``` - Done
---------
Co-authored-by: Darren Schroeder <343840+fdncred@users.noreply.github.com>
Co-authored-by: Antoine Stevan <44101798+amtoine@users.noreply.github.com>
This PR reverts https://github.com/nushell/nushell/pull/9391
We try not to revert PRs like this, though after discussion with the
Nushell team, we decided to revert this one.
The main reason is that Nushell, as a codebase, isn't ready for these
kinds of optimisations. It's in the part of the development cycle where
our main focus should be on improving the algorithms inside of Nushell
itself. Once we have matured our algorithms, then we can look for
opportunities to switch out technologies we're using for alternate
forms.
Much of Nushell still has lots of opportunities for tuning the codebase,
paying down technical debt, and making the codebase generally cleaner
and more robust. This should be the focus. Performance improvements
should flow out of that work.
Said another, optimisation that isn't part of tuning the codebase is
premature at this stage. We need to focus on doing the hard work of
making the engine, parser, etc better.
# User-Facing Changes
Reverts the HashMap -> ahash change.
cc @FilipAndersson245
<!--
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 removes ZB and ZiB from file size type, as they
were showing incorrect values due to an integer overflow.
Fixes: #9337
# 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
see https://github.com/nushell/nushell/issues/9390
using `ahash` instead of the default hasher. this will not affect
compile time as we where already building `ahash`.
# 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
Fixes: #9165
It's because `sys` returns a lazy record, and `insert`, `update`,
`upsert` can't operate on lazy record yet.
# User-Facing 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
Fixes the clippy warnings we're about to get hit with next time we
upgrade Rust.
The big one was shrinking ShellError and related under 128 bytes.
# User-Facing Changes
Shouldn't notice much difference. In theory, we could see a tiny perf
improvement, but I didn't notice one.
# 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
Fixes#9254.
# User-Facing Changes
upserting data of a cellpath that doesn't exist into a record now
creates the cellpath.
# Tests + Formatting
```
~/CodingProjects/nushell> mut a = {}
~/CodingProjects/nushell> $a.b.c = 99
~/CodingProjects/nushell> $a
╭───┬────────────╮
│ │ ╭───┬────╮ │
│ b │ │ c │ 99 │ │
│ │ ╰───┴────╯ │
╰───┴────────────╯
```
<!--
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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
Despite the innocent-looking title, this PR involves quite a few backend
changes as the existing LazyRecord trait was not at all friendly towards
the idea of these values being generated on the fly from Nu code.
In particular, here are a few changes involved:
- The LazyRecord trait now involves a lifetime `'a`, and this lifetime
is used in the return value of `get_column_names`. This means it no
longer returns `'static str`s (but implementations still can return
these). This is more stringent on the consumption side.
- The LazyRecord trait now must be able to clone itself via a new
`clone_value` method (as requiring `Clone` is not object safe). This
pattern is borrowed from `Value::CustomValue`.
- LazyRecord no longer requires being serde serializable and
deserializable.
These, in hand, allow for the following:
- LazyRecord can now clone itself, which means that they don't have to
be collected into a Record when being cloned.
- This is especially useful in Stack, which is cloned on each repl line
and in a few other cases. This would mean that _every_ LazyRecord
instance stored in a variable would be collected in its entirety and
cloned, which can be catastrophic for performance. See: `let nulol =
$nu`.
- LazyRecord's columns don't have to be static, they can have the same
lifetime of the struct itself, so different instances of the same
LazyRecord type can have different columns and values (like the new
`NuLazyRecord`)
- Serialization and deserialization are no longer meaningless, they are
simply less.
I would consider this PR very "drafty", but everything works. It
probably requires some cleanup and testing, though, but I'd like some
eyes and pointers first.
# User-Facing Changes
New command. New restrictions are largely internal. Maybe there are some
plugins affected?
Example of new command's usage:
```
lazy make --columns [a b c] --get-value { |name| print $"getting ($name)"; $name | str upcase }
```
You can also trivially implement something like `lazy make record` to
take a record of closures and turn it into a getter-like lazy struct:
```
def "lazy make record" [
record: record
] {
let columns = ($record | columns)
lazy make --columns $columns --get-value { |col| do ($record | get $col) }
}
```
Open to bikeshedding. `lazy make` is similar to `error make` which is
also in the core commands. I didn't like `make lazy` since it sounded
like some transformation was going on.
# Tour for reviewers
Take a look at LazyMake's examples. They have `None` as the results, as
such they aren't _really_ correct and aren't being tested at all. I
didn't do this because creating the Value::LazyRecord is a little tricky
and didn't want to risk messing it up, especially as the necessary
variables aren't available when creating the examples (like stack and
engine state).
Also take a look at NuLazyRecord's get_value implementation, or in
general. It uses an Arc<Mutex<_>> for the stack, which must be accessed
mutably for eval_block but get_value only provides us with a `&self`.
This is a sad state of affairs, but I don't know if there's a better
way.
On the same code path, we also have pipeline handling, and any pipeline
that isn't a Pipeline::Value will return Value::nothing. I believe
returning a Value::Error is probably better, or maybe some other
handling. Couldn't decide on which ShellError to settle with for that
branch.
The "unfortunate casualty" in the columns.rs file. I'm not sure just how
bad that is, though, I simply had to fight a little with the borrow
checker.
A few leftover comments like derives, comments about the now
non-existing serde requirements, and impls. I'll definitely get around
to those eventually but they're in atm
Should NuLazyRecord implement caching? I'm leaning heavily towards
**yes**, this was one of the main reasons not to use a record of
closures (besides convenience), but maybe it could be opt-out. I'd
wonder about its implementation too, but a simple way would be to move a
HashMap into the mutex state and keep cached values there.
(*third* try at posting this PR, #9104, like #9084, got polluted with
unrelated commits. I'm never going to pull from the github feature
branch again!)
# 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.
-->
Show parameter defaults in scope command signature, where they're
available for display by help.
per https://github.com/nushell/nushell/issues/8928.
I found unexpected ramifications in one completer (NuHelpCompleter) and
plugins, which both use the flag-formatting routine from builtin help.
For the moment I made the minimum necessary changes to get the mainline
scenario to pass tests and run. But we should circle back on what to do
with plugins and help completer..
# User-Facing Changes
<!-- List of all changes that impact the user experience here. This
helps us keep track of breaking changes. -->
1. New `parameter_default` column to `signatures` table in
`$nu.scope.commands`
It is populated with whatever parameters can be defaulted: currently
positional args and named flags.
2. Built in help (both `help <command>` and `<command> --help` will
display the defaults
3. Help completer will display defaults for flags, but not for
positionals.
Example:
A custom command with some default parameters:
```
〉cat ~/work/dflts.nu
# sample function to show defaults in help
export def main [
arg1: string # mandatory positional
arg2:string=abc # optional positional
--switch # no default here
--named:int # named flag, no default
--other:string=def # flag
--hard:record<foo:int bar:string, bas:bool> # default can be compound type
= {foo:22, bar:"other worlds", bas:false}
] { {arg1: $arg1,
arg2: $arg2,
switch: $switch,
named: $named,
other: $other,
hard: $hard, }
}
〉use ~/work/dflts.nu
〉$nu.scope.commands | where name == 'dflts' | get signatures.0.any | reject short_flag description custom_completion
╭───┬────────────────┬────────────────┬──────────────────────────────────────────┬─────────────┬───────────────────────────╮
│ # │ parameter_name │ parameter_type │ syntax_shape │ is_optional │ parameter_default │
├───┼────────────────┼────────────────┼──────────────────────────────────────────┼─────────────┼───────────────────────────┤
│ 0 │ │ input │ any │ false │ │
│ 1 │ arg1 │ positional │ string │ false │ │
│ 2 │ arg2 │ positional │ string │ true │ abc │
│ 3 │ switch │ switch │ │ true │ │
│ 4 │ named │ named │ int │ true │ │
│ 5 │ other │ named │ string │ true │ def │
│ 6 │ hard │ named │ record<foo: int, bar: string, bas: bool> │ true │ ╭───────┬───────────────╮ │
│ │ │ │ │ │ │ foo │ 22 │ │
│ │ │ │ │ │ │ bar │ other worlds │ │
│ │ │ │ │ │ │ bas │ false │ │
│ │ │ │ │ │ ╰───────┴───────────────╯ │
│ 7 │ │ output │ any │ false │ │
╰───┴────────────────┴────────────────┴──────────────────────────────────────────┴─────────────┴───────────────────────────╯
〉help dflts
sample function to show defaults in help
Usage:
> dflts {flags} <arg1> (arg2)
Flags:
--switch - switch -- no default here
--named <Int> - named flag, typed, but no default
--other <String> - flag with default (default: 'def')
--hard <Record([("foo", Int), ("bar", String), ("bas", Boolean)])> - default can be compound type (default: {foo: 22, bar: 'other worlds', bas: false})
-h, --help - Display the help message for this command
Parameters:
arg1 <string>: mandatory positional
arg2 <string>: optional positional (optional, default: 'abc')
```
Compared to (relevant bits of) help output previously:
```
Flags:
-h, --help - Display the help message for this command
-, --switch - no default here
-, --named <int> - named flag, no default
-, --other <string> - flag
-, --hard <record<foo: int, bar: string, bas: bool>> - default can be compound type
Signatures:
<any> | dflts <string> <string> -> <any>
Parameters:
arg1 <string>: mandatory positional
(optional) arg2 <string>: optional positional
```
# 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 -A
clippy::needless_collect -A clippy::result_large_err` to check that
you're using the standard code style
- `cargo test --workspace` to check that all tests pass
- `cargo run -- crates/nu-std/tests/run.nu` 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
> [x] 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.
-->
This PR makes `?` work with `reject`. For example:
```bash
> {} | reject foo
Error: nu:🐚:column_not_found
× Cannot find column
╭─[entry #2:1:1]
1 │ {} | reject foo
· ───┬── ─┬─
· │ ╰── cannot find column 'foo'
· ╰── value originates here
╰────
> {} | reject foo?
╭──────────────╮
│ empty record │
╰──────────────╯
```
This was prompted by [a user
question](https://discord.com/channels/601130461678272522/614593951969574961/1091466428546306078).
I would like to get this in for 0.78, I think it's low-risk and I want
the `?` feature to be as polished as possible for its debut.
Prior to this PR, the less/greater than operators (`<`, `>`, `<=`, `>=`)
would throw an error if either side was null. After this PR, these
operators return null if either side (or both) is null.
### Examples
```bash
1 < 3 # true
1 < null # null
null < 3 # null
null < null # null
```
### Motivation
JT [asked the C#
folks](https://discord.com/channels/601130461678272522/615329862395101194/1086137515053957140)
and this is apparently the approach they would choose for comparison
operators if they could start from scratch.
This PR makes `where` more convenient to use on jagged/missing data. For
example, we can now filter on columns that may not be present in every
row:
```
> [{foo: 123} {}] | where foo? > 10
╭───┬─────╮
│ # │ foo │
├───┼─────┤
│ 0 │ 123 │
╰───┴─────╯
```
Some minor code cleanup.
We've accumulated a few macros over the years that arguably don't need
to be macros. This PR removes 4 macros by either:
1. Inlining the macro
2. Replacing the macro with a local function
3. Replacing the macro with a closure
# Description
This adds `match` and basic pattern matching.
An example:
```
match $x {
1..10 => { print "Value is between 1 and 10" }
{ foo: $bar } => { print $"Value has a 'foo' field with value ($bar)" }
[$a, $b] => { print $"Value is a list with two items: ($a) and ($b)" }
_ => { print "Value is none of the above" }
}
```
Like the recent changes to `if` to allow it to be used as an expression,
`match` can also be used as an expression. This allows you to assign the
result to a variable, eg) `let xyz = match ...`
I've also included a short-hand pattern for matching records, as I think
it might help when doing a lot of record patterns: `{$foo}` which is
equivalent to `{foo: $foo}`.
There are still missing components, so consider this the first step in
full pattern matching support. Currently missing:
* Patterns for strings
* Or-patterns (like the `|` in Rust)
* Patterns for tables (unclear how we want to match a table, so it'll
need some design)
* Patterns for binary values
* And much more
# User-Facing Changes
[see above]
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
> **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.
This is a follow-up to https://github.com/nushell/nushell/pull/8379 and
https://github.com/nushell/nushell/discussions/8502.
This PR makes it so that the new `?` syntax for marking a path member as
optional short-circuits, as voted on in the
[8502](https://github.com/nushell/nushell/discussions/8502) poll.
Previously, `{ foo: 123 }.bar?.baz` would raise an error:
```
> { foo: 123 }.bar?.baz
× Data cannot be accessed with a cell path
╭─[entry #15:1:1]
1 │ { foo: 123 }.bar?.baz
· ─┬─
· ╰── nothing doesn't support cell paths
╰────
```
Here's what was happening:
1. The `bar?` path member access returns `nothing` because there is no field named `bar` on the record
2. The `baz` path member access fails when trying to access a `baz` field on that `nothing` value
After this change, `{ foo: 123 }.bar?.baz` returns `nothing`; the failed `bar?` access immediately returns `nothing` and the `baz` access never runs.
# Description
Resolves issue #8370
Adds the following flags to commands `from csv` and `from tsv`:
- `--flexible`: allow the number of fields in records to be variable
- `-c --comment`: a comment character to ignore lines starting with it
- `-q --quote`: a quote character to ignore separators in strings,
defaults to '\"'
- `-e --escape`: an escape character for strings containing the quote
character
Internally, the `Value` struct has an additional helper function
`as_char` which converts it to a single `char`
# User-Facing Changes
The single quoted string `'\t'` can no longer be used as a parameter for
the flag `--separator '\t'` as it is interpreted as a two-character
string. One needs to use from now on the flag with a double quoted
string like so: `-s "\t"` which correctly interprets the string as a
single `char`.
This is a follow up from https://github.com/nushell/nushell/pull/7540.
Please provide feedback if you have the time!
## Summary
This PR lets you use `?` to indicate that a member in a cell path is
optional and Nushell should return `null` if that member cannot be
accessed.
Unlike the previous PR, `?` is now a _postfix_ modifier for cell path
members. A cell path of `.foo?.bar` means that `foo` is optional and
`bar` is not.
`?` does _not_ suppress all errors; it is intended to help in situations
where data has "holes", i.e. the data types are correct but something is
missing. Type mismatches (like trying to do a string path access on a
date) will still fail.
### Record Examples
```bash
{ foo: 123 }.foo # returns 123
{ foo: 123 }.bar # errors
{ foo: 123 }.bar? # returns null
{ foo: 123 } | get bar # errors
{ foo: 123 } | get bar? # returns null
{ foo: 123 }.bar.baz # errors
{ foo: 123 }.bar?.baz # errors because `baz` is not present on the result from `bar?`
{ foo: 123 }.bar.baz? # errors
{ foo: 123 }.bar?.baz? # returns null
```
### List Examples
```
〉[{foo: 1} {foo: 2} {}].foo
Error: nu:🐚:column_not_found
× Cannot find column
╭─[entry #30:1:1]
1 │ [{foo: 1} {foo: 2} {}].foo
· ─┬ ─┬─
· │ ╰── cannot find column 'foo'
· ╰── value originates here
╰────
〉[{foo: 1} {foo: 2} {}].foo?
╭───┬───╮
│ 0 │ 1 │
│ 1 │ 2 │
│ 2 │ │
╰───┴───╯
〉[{foo: 1} {foo: 2} {}].foo?.2 | describe
nothing
〉[a b c].4? | describe
nothing
〉[{foo: 1} {foo: 2} {}] | where foo? == 1
╭───┬─────╮
│ # │ foo │
├───┼─────┤
│ 0 │ 1 │
╰───┴─────╯
```
# Breaking changes
1. Column names with `?` in them now need to be quoted.
2. The `-i`/`--ignore-errors` flag has been removed from `get` and
`select`
1. After this PR, most `get` error handling can be done with `?` and/or
`try`/`catch`.
4. Cell path accesses like this no longer work without a `?`:
```bash
〉[{a:1 b:2} {a:3}].b.0
2
```
We had some clever code that was able to recognize that since we only
want row `0`, it's OK if other rows are missing column `b`. I removed
that because it's tricky to maintain, and now that query needs to be
written like:
```bash
〉[{a:1 b:2} {a:3}].b?.0
2
```
I think the regression is acceptable for now. I plan to do more work in
the future to enable streaming of cell path accesses, and when that
happens I'll be able to make `.b.0` work again.
# Description
The correction made here concerns the issue #8431. Indeed, the algorithm
initially proposed to remove elements of a `vector` performed a loop
with `remove` and an incident therefore appeared when several values
were equal because the deletion was done outside the length of the
vector:
```rust
let mut found = false;
for (i, col) in cols.clone().iter().enumerate() {
if col == col_name {
cols.remove(i);
vals.remove(i);
found = true;
}
}
```
Then, `[[a, a]; [1, 2]] | reject a: ` gave `thread 'main' panicked at
'removal index (is 1) should be < len (is 1)',
crates/nu-protocol/src/value/mod.rs:1213:54`.
The proposed correction is therefore the implementation of the
`retain_mut` utility dedicated to this functionality.
```rust
let mut found = false;
let mut index = 0;
cols.retain_mut(|col| {
if col == col_name {
found = true;
vals.remove(index);
false
} else {
index += 1;
true
}
});
```
# Description
Our `ShellError` at the moment has a `std::mem::size_of<ShellError>` of
136 bytes (on AMD64). As a result `Value` directly storing the struct
also required 136 bytes (thanks to alignment requirements).
This change stores the `Value::Error` `ShellError` on the heap.
Pro:
- Value now needs just 80 bytes
- Should be 1 cacheline less (still at least 2 cachelines)
Con:
- More small heap allocations when dealing with `Value::Error`
- More heap fragmentation
- Potential for additional required memcopies
# Further code changes
Includes a small refactor of `try` due to a type mismatch in its large
match.
# User-Facing Changes
None for regular users.
Plugin authors may have to update their matches on `Value` if they use
`nu-protocol`
Needs benchmarking to see if there is a benefit in real world workloads.
**Update** small improvements in runtime for workloads with high volume
of values. Significant reduction in maximum resident set size, when many
values are held in memory.
# Tests + Formatting
# Description
Fix for data ambiguity noted in #8244.
Basic change is to use nanosecond resolution for unix timestamps (stored
in type Int). Previously, a timestamp might have seconds, milliseconds
or nanoseconds, but it turned out there were overlaps in data ranges
between different resolutions, so there wasn't always a unique mapping
back to date/time.
Due to higher precision, the *range* of dates that timestamps can map to
is restricted. Unix timestamps with seconds resolution and 64 bit
storage can cover all dates from the Big Bang to eternity. Timestamps
with seconds resolution and 32 bit storage can only represent dates from
1901-12-13 through 2038-01-19. The nanoseconds resolution and 64 bit
storage used with this fix can represent dates from 1677-09-21T00:12:44
to 2262-04-11T23:47:16, something of a compromise.
# User-Facing Changes
_(List of all changes that impact the user experience here. This helps
us keep track of breaking changes.)_
## `<datetime> | into int`
Converts to nanosecond resolution
```rust
〉date now | into int
1678084730502126846
```
This is the number of non-leap nanoseconds after the unix epoch date:
1970-01-01T00:00:00+00:00.
Conversion fails for dates outside the supported range:
```rust
〉1492-10-12 | into int
Error: nu:🐚:incorrect_value
× Incorrect value.
╭─[entry #51:1:1]
1 │ 1492-10-12 | into int
· ────┬───
· ╰── DateTime out of timestamp range 1677-09-21T00:12:43 and 2262-04-11T23:47:16
╰────
```
## `<int> | into datetime`
Can no longer fail or produce incorrect results for any 64-bit input:
```rust
〉0 | into datetime
Thu, 01 Jan 1970 00:00:00 +0000 (53 years ago)
〉"7fffffffffffffff" | into int -r 16 | into datetime
Fri, 11 Apr 2262 23:47:16 +0000 (in 239 years)
〉("7fffffffffffffff" | into int -r 16) * -1 | into datetime
Tue, 21 Sep 1677 00:12:43 +0000 (345 years ago)
```
## `<date> | date to-record` and `<date> | date to-table`
Now both have a `nanosecond` field.
```rust
〉"7fffffffffffffff" | into int -r 16 | into datetime | date to-record
╭────────────┬───────────╮
│ year │ 2262 │
│ month │ 4 │
│ day │ 11 │
│ hour │ 23 │
│ minute │ 47 │
│ second │ 16 │
│ nanosecond │ 854775807 │
│ timezone │ +00:00 │
╰────────────┴───────────╯
〉"7fffffffffffffff" | into int -r 16 | into datetime | date to-table
╭───┬──────┬───────┬─────┬──────┬────────┬────────┬────────────┬──────────╮
│ # │ year │ month │ day │ hour │ minute │ second │ nanosecond │ timezone │
├───┼──────┼───────┼─────┼──────┼────────┼────────┼────────────┼──────────┤
│ 0 │ 2262 │ 4 │ 11 │ 23 │ 47 │ 16 │ 854775807 │ +00:00 │
╰───┴──────┴───────┴─────┴──────┴────────┴────────┴────────────┴──────────╯
```
This change was not mandated by the OP problem, but it is nice to be
able to see the nanosecond bits that were present in Nushell `date` type
all along.
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
# 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.
Fixes#8341.
The `CustomValue::to_json()` function is an odd duck; it defaults to
returning `null`, and no `CustomValue` implementations override it to do
anything useful. I forgot to implement `to_json()` for `SQLiteDatabase`,
so `open foo.db | to json` was returning `null`.
To fix this, I've removed `CustomValue::to_json()` and now `to json`
will collect a `CustomValue` into a regular `Value` before doing a JSON
conversion.
Continuation of #8229 and #8326
# Description
The `ShellError` enum at the moment is kind of messy.
Many variants are basic tuple structs where you always have to reference
the implementation with its macro invocation to know which field serves
which purpose.
Furthermore we have both variants that are kind of redundant or either
overly broad to be useful for the user to match on or overly specific
with few uses.
So I set out to start fixing the lacking documentation and naming to
make it feasible to critically review the individual usages and fix
those.
Furthermore we can decide to join or split up variants that don't seem
to be fit for purpose.
# Call to action
**Everyone:** Feel free to add review comments if you spot inconsistent
use of `ShellError` variants.
# User-Facing Changes
(None now, end goal more explicit and consistent error messages)
# Tests + Formatting
(No additional tests needed so far)
# Commits (so far)
- Remove `ShellError::FeatureNotEnabled`
- Name fields on `SE::ExternalNotSupported`
- Name field on `SE::InvalidProbability`
- Name fields on `SE::NushellFailed` variants
- Remove unused `SE::NushellFailedSpannedHelp`
- Name field on `SE::VariableNotFoundAtRuntime`
- Name fields on `SE::EnvVarNotFoundAtRuntime`
- Name fields on `SE::ModuleNotFoundAtRuntime`
- Remove usused `ModuleOrOverlayNotFoundAtRuntime`
- Name fields on `SE::OverlayNotFoundAtRuntime`
- Name field on `SE::NotFound`
Continuation of #8229
# Description
The `ShellError` enum at the moment is kind of messy.
Many variants are basic tuple structs where you always have to reference
the implementation with its macro invocation to know which field serves
which purpose.
Furthermore we have both variants that are kind of redundant or either
overly broad to be useful for the user to match on or overly specific
with few uses.
So I set out to start fixing the lacking documentation and naming to
make it feasible to critically review the individual usages and fix
those.
Furthermore we can decide to join or split up variants that don't seem
to be fit for purpose.
**Everyone:** Feel free to add review comments if you spot inconsistent
use of `ShellError` variants.
- Name fields of `SE::IncorrectValue`
- Merge and name fields on `SE::TypeMismatch`
- Name fields on `SE::UnsupportedOperator`
- Name fields on `AssignmentRequires*` and fix doc
- Name fields on `SE::UnknownOperator`
- Name fields on `SE::MissingParameter`
- Name fields on `SE::DelimiterError`
- Name fields on `SE::IncompatibleParametersSingle`
# User-Facing Changes
(None now, end goal more explicit and consistent error messages)
# Tests + Formatting
(No additional tests needed so far)
# Description
The `ShellError` enum at the moment is kind of messy.
Many variants are basic tuple structs where you always have to reference
the implementation with its macro invocation to know which field serves
which purpose.
Furthermore we have both variants that are kind of redundant or either
overly broad to be useful for the user to match on or overly specific
with few uses.
So I set out to start fixing the lacking documentation and naming to
make it feasible to critically review the individual usages and fix
those.
Furthermore we can decide to join or split up variants that don't seem
to be fit for purpose.
Feel free to add review comments if you spot inconsistent use of
`ShellError` variants.
- Name fields on `ShellError::OperatorOverflow`
- Name fields on `ShellError::PipelineMismatch`
- Add doc to `ShellError::OnlySupportsThisInputType`
- Name `ShellError::OnlySupportsThisInputType`
- Name field on `ShellError::PipelineEmpty`
- Comment about issues with `TypeMismatch*`
- Fix a few `exp_input_type`s
- Name fields on `ShellError::InvalidRange`
# User-Facing Changes
(None now, end goal more explicit and consistent error messages)
# Tests + Formatting
(No additional tests needed so far)
# Description
As title, I found this feature is useful to me too :)
Closes: #8039
# User-Facing Changes
```
❯ 3 * "ab"
ababab
❯ 3 * [1 2 3]
╭───┬───╮
│ 0 │ 1 │
│ 1 │ 2 │
│ 2 │ 3 │
│ 3 │ 1 │
│ 4 │ 2 │
│ 5 │ 3 │
│ 6 │ 1 │
│ 7 │ 2 │
│ 8 │ 3 │
╰───┴───╯
```
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
# 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.
I noticed that `open some_big_file | into binary` cannot be cancelled
with `ctrl+c`.
This small PR fixes that by checking `ctrl+c` in
`RawStream::into_bytes()`, and does the same in
`RawStream::into_string()` for good measure.
This PR makes `++` (the append operator) work with strings and binary
values. Can now do things like:
```bash
〉"a" ++ "b"
ab
〉0x[01 02] ++ 0x[03]
Length: 3 (0x3) bytes | printable whitespace ascii_other non_ascii
00000000: 01 02 03 •••
```
Closes#8015.
# Description
Closes#6768.
BEFORE:
```
〉{ foo: [{a:1, b:2},2,3,4,5] }
╭─────┬────────────────╮
│ foo │ [table 5 rows] │
╰─────┴────────────────╯
```
AFTER:
```
〉{ foo: [{a:1, b:2},2,3,4,5] }
╭─────┬────────────────╮
│ foo │ [list 5 items] │
╰─────┴────────────────╯
```
# User-Facing Changes
See above.
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
# 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
Lint: `clippy::uninlined_format_args`
More readable in most situations.
(May be slightly confusing for modifier format strings
https://doc.rust-lang.org/std/fmt/index.html#formatting-parameters)
Alternative to #7865
# User-Facing Changes
None intended
# Tests + Formatting
(Ran `cargo +stable clippy --fix --workspace -- -A clippy::all -D
clippy::uninlined_format_args` to achieve this. Depends on Rust `1.67`)
# Description
This PR bumps the required rust version to 1.66.1.
# User-Facing 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
# 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.
This is an attempt to implement a new `Value::LazyRecord` variant for
performance reasons.
`LazyRecord` is like a regular `Record`, but it's possible to access
individual columns without evaluating other columns. I've implemented
`LazyRecord` for the special `$nu` variable; accessing `$nu` is
relatively slow because of all the information in `scope`, and [`$nu`
accounts for about 2/3 of Nu's startup time on
Linux](https://github.com/nushell/nushell/issues/6677#issuecomment-1364618122).
### Benchmarks
I ran some benchmarks on my desktop (Linux, 12900K) and the results are
very pleasing.
Nu's time to start up and run a command (`cargo build --release;
hyperfine 'target/release/nu -c "echo \"Hello, world!\""' --shell=none
--warmup 10`) goes from **8.8ms to 3.2ms, about 2.8x faster**.
Tests are also much faster! Running `cargo nextest` (with our very slow
`proptest` tests disabled) goes from **7.2s to 4.4s (1.6x faster)**,
because most tests involve launching a new instance of Nu.
### Design (updated)
I've added a new `LazyRecord` trait and added a `Value` variant wrapping
those trait objects, much like `CustomValue`. `LazyRecord`
implementations must implement these 2 functions:
```rust
// All column names
fn column_names(&self) -> Vec<&'static str>;
// Get 1 specific column value
fn get_column_value(&self, column: &str) -> Result<Value, ShellError>;
```
### Serializability
`Value` variants must implement `Serializable` and `Deserializable`, which poses some problems because I want to use unserializable things like `EngineState` in `LazyRecord`s. To work around this, I basically lie to the type system:
1. Add `#[typetag::serde(tag = "type")]` to `LazyRecord` to make it serializable
2. Any unserializable fields in `LazyRecord` implementations get marked with `#[serde(skip)]`
3. At the point where a `LazyRecord` normally would get serialized and sent to a plugin, I instead collect it into a regular `Value::Record` (which can be serialized)
I have changed `assert!(a == b)` calls to `assert_eq!(a, b)`, which give
better error messages. Similarly for `assert!(a != b)` and
`assert_ne!(a, b)`. Basically all instances were comparing primitives
(string slices or integers), so there is no loss of generality from
special-case macros,
I have also fixed a number of typos in comments, variable names, and a
few user-facing messages.
# Description
_(Description of your pull request goes here. **Provide examples and/or
screenshots** if your changes affect the user experience.)_
I implemented the status bar we talk about yesterday. The idea was
inspired by the progress bar of `wget`.
I decided to go for the second suggestion by `@Reilly`
> 2. add an Option<usize> or whatever to RawStream (and ListStream?) for
situations where you do know the length ahead of time
For now only works with the command `save` but after the approve of this
PR we can see how we can implement it on commands like `cp` and `mv`
When using `fetch` nushell will check if there is any `content-length`
attribute in the request header. If so, then `fetch` will send it
through the new `Option` variable in the `RawStream` to the `save`.
If we know the total size we show the progress bar
but if we don't then we just show the stats like: data already saved,
bytes per second, and time lapse.
Please let me know If I need to make any changes and I will be happy to
do it.
# User-Facing Changes
A new flag (`--progress` `-p`) was added to the `save` command
Examples:
```nu
fetch https://github.com/torvalds/linux/archive/refs/heads/master.zip | save --progress -f main.zip
fetch https://releases.ubuntu.com/22.04.1/ubuntu-22.04.1-desktop-amd64.iso | save --progress -f main.zip
open main.zip --raw | save --progress main.copy
```
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
-
I am getting some errors and its weird because the errors are showing up
in files i haven't touch. Is this normal?
# 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: Reilly Wood <reilly.wood@icloud.com>
# Description
This closes#7498, as well as fixes an issue reported in
https://github.com/nushell/nushell/pull/7002#issuecomment-1368340773
BEFORE:
```
〉[{foo: 'bar'} {}] | get foo
Error: nu:🐚:column_not_found (link)
× Cannot find column
╭─[entry #5:1:1]
1 │ [{foo: 'bar'} {}] | get foo
· ────────┬──────── ─┬─
· │ ╰── value originates here
· ╰── cannot find column 'Empty cell'
╰────
〉[{foo: 'bar'} {}].foo
╭───┬─────╮
│ 0 │ bar │
│ 1 │ │
╰───┴─────╯
```
AFTER:
```
〉[{foo: 'bar'} {}] | get foo
Error: nu:🐚:column_not_found (link)
× Cannot find column
╭─[entry #1:1:1]
1 │ [{foo: 'bar'} {}] | get foo
· ─┬ ─┬─
· │ ╰── cannot find column 'foo'
· ╰── value originates here
╰────
〉[{foo: 'bar'} {}].foo
Error: nu:🐚:column_not_found (link)
× Cannot find column
╭─[entry #3:1:1]
1 │ [{foo: 'bar'} {}].foo
· ─┬ ─┬─
· │ ╰── cannot find column 'foo'
· ╰── value originates here
╰────
```
EDIT: This also changes the semantics of `get`/`select` `-i` somewhat.
I've decided to leave it like this because it works more intuitively
with `default` and `compact`.
BEFORE:
```
〉[{a:1} {b:2} {a:3}] | select -i foo | to nuon
null
```
AFTER:
```
〉[{a:1} {b:2} {a:3}] | select -i foo | to nuon
[[foo]; [null], [null], [null]]
```
# User-Facing Changes
See above. EDIT: the issue with holes in cases like ` [{foo: 'bar'}
{}].foo.0` versus ` [{foo: 'bar'} {}].0.foo` has been resolved.
# 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 -A
clippy::needless_collect` to check that you're using the standard code
style
- `cargo test --workspace` to check that all tests pass
# 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.
