Commit Graph

1357 Commits

Author SHA1 Message Date
02b2c55146 Rolling and cumulative commands (#3960)
* rolling and cumulative operations

* update polars to 0.15.1

* change reference in function
2021-08-24 09:10:29 -05:00
JT
487fafbca3 Add a 'tutor' command (#3949)
* Add a 'tutor' command

* clippy
2021-08-21 19:41:54 +12:00
188a352c6f added help --find to search usage and extra_usage text for strings (#3948)
* added help --find to search usage and extra_usage text for strings

* changed my mind
2021-08-20 17:55:56 -05:00
e11b400a75 Allow source to accept paths with emojis (#3939)
* Allow sourcing paths with emojis

* Add source command tests for emoji paths

* Fmt

* Disable source tests on Windows with illegal paths

* Test sourcing also ASCII and single-quoted paths
2021-08-19 19:06:18 +12:00
6db5692be4 Only allow unaliasing in current scope, add tests (#3936)
* unalias only removes aliases in the current scope

* Add a test and fix previous ones which did not function as expected
2021-08-19 19:05:36 +12:00
JT
ead4029d49 Bump rustyline and add unalias test (#3935) 2021-08-18 05:55:34 +12:00
9bd408449e Add the ability to remove and list aliases (#3879)
* Add the ability to remove and list aliases

* Fix failing unit tests

* Add a test to check unalias shadowing blocks
2021-08-17 08:56:35 -05:00
ab961a78cb Add trailing slash in completion of symlinked dir (#3921) 2021-08-17 07:13:59 +12:00
65c639cf13 allow fetch to follow redirects (#3923) 2021-08-16 12:31:02 -05:00
0cf5dc11e3 Fix 'Inimplemented' typo. (#3922) 2021-08-16 09:17:31 -05:00
b873fa7a5f The zip command. (#3919)
We introduce it here and allow it to work with regular lists (tables with no columns) as well as symmetric tables. Say we have two lists and wish to zip them, like so:

```
[0 2 4 6 8] | zip {
  [1 3 5 7 9]
} | flatten

───┬───
 0 │ 0
 1 │ 1
 2 │ 2
 3 │ 3
 4 │ 4
 5 │ 5
 6 │ 6
 7 │ 7
 8 │ 8
 9 │ 9
───┴───
```

In the case for two tables instead:

```
[[symbol]; ['('] ['['] ['{']] | zip {
  [[symbol]; [')'] [']'] ['}']]
} | each {
  get symbol | $'($in.0)nushell($in.1)'
}

───┬───────────
 0 │ (nushell)
 1 │ [nushell]
 2 │ {nushell}
───┴───────────
```
2021-08-14 23:36:08 -05:00
ee563ecf4e PROMPT_STRING env variable (#3918)
* prompt string env variable

* cargo clippy
2021-08-15 06:14:14 +12:00
183b35d683 Rustyline bug fixes (#3916)
* Mitigate history file bug in Rustyline

Rustyline's duplicate ignoring code has a bug that can cause data loss and
history file corruption. Testing seems to indicate that disabling this behavior
and allowing duplicates will prevent the bug from showing up. Many people have
complained about this issue, I think it is worthwhile to fix the bug at the cost
of permitting duplicate history entries.

Upstream bug: https://github.com/kkawakam/rustyline/issues/559

* Increase Rustyline historyfile limit

Rustyline will only store 100 history items by default. This is quite a small
limit for a shell that people use as a daily driver. Especially when the
deduplication code is removed, we will hit that limit quickly and start to lose
history. This commit bumps the limit up to 10k. We can discuss if this is an
inappropriate limit or if we should allow users to specify this setting in their
nushell config file instead.
2021-08-14 06:56:28 +12:00
463dd48180 Flexible dropping of rows (by desired row number) (#3917)
We very well support `nth 0 2 3 --skip 1 4` to select particular rows and skip some using a flag. However, in practice we deal with tables (whether they come from parsing or loading files and whatnot) where we don't know the size of the table up front (and everytime we have these, they may have different sizes). There are also other use cases when we use intermediate tables during processing and wish to always drop certain rows and **keep the rest**.

Usage:

```
... | drop nth 0
... | drop nth 3 8
```
2021-08-13 12:48:05 -05:00
c9b87c4c03 Count the size of the directories when calculating the size in DirInfo (#3902)
* take dir entry size into consideration when calculting the size of a directory in DirInfo

* fmt check
2021-08-08 05:50:02 +12:00
38848082ae describe command (#3907) 2021-08-08 05:48:54 +12:00
JT
b6728efcd4 in/not-in for strings (#3906) 2021-08-07 09:49:37 +12:00
cd814851da Use bigdecimal-rs patch (#3905)
* Use bigdecimal-rs patch

* fix nu-serde's bigdecimal dependency
2021-08-07 09:27:19 +12:00
6646daab45 source config from $NU_CONFIG_DIR if it exists (#3883) 2021-08-06 11:47:11 -05:00
ba483155d7 Reimplement parsers with nu-serde (#3880)
The nu-serde crate allows us to become much more generic with respect to how we
convert output to `nu-protocol::Value`s. This allows us to remove a lot of the
special-case code that we wrote for deserializing JSON values.

Co-authored-by: Darren Schroeder <343840+fdncred@users.noreply.github.com>
2021-08-06 11:46:19 -05:00
63abe1cb3e Datetime commands (#3894)
* date and duration from nu

* date commands

* Import to feature flag

* corrected to-csv example

* corrected sample example
2021-08-05 17:18:53 -05:00
7dcc08985c Implement PartialEq for ReturnSuccess (#3888)
This makes ReturnValue and ReturnSuccess usable in assert_eq!.
2021-08-05 14:55:41 +12:00
325f45fa66 Fix typo: patter → pattern (#3890) 2021-08-03 18:48:23 -05:00
JT
bc682066d8 Bump to 0.35 (#3884) 2021-08-03 20:01:09 +12:00
0a1cdc5107 Add the nu-serde crate (#3878)
* Add the nu-serde crate

nu-serde is a crate that can be used to turn a value implementing
`serde::Serialize` into a `nu-protocol::Value`. This has the potential to
significantly simplify plugin authorship.

This crate was the previously independent
[serde-nu](https://github.com/lily-mara/serde-nu) but the nushell maintainers
expressed an interest in having it added to the mainline nushell repository.

* fixup! Add the nu-serde crate
2021-07-31 22:03:13 -05:00
6984185e61 Better representation in nested dataframes (#3875)
* better dataframe representation in nested df

* Error message correction
2021-07-31 09:02:32 -05:00
5826126284 simple contains arguments (#3874) 2021-07-31 09:01:05 -05:00
762e528ec5 Support equals sign in shorthand environment variable values (#3869)
Some environment variables, such as `RUST_LOG` include equals signs. Nushell
should support this in the shorthand environment variable syntax so that
developers using these variables can control them easily. We accomplish this by
swapping `std::str::split` for `std::str::splitn`, which ensures that we only
consider the first equals sign in the string instead of all of them, which we
did previously.

Closes #3867
2021-07-31 09:10:28 +12:00
JT
c3de9848b4 Revert "Unify use of the surf crate (#3855)" (#3871)
This reverts commit 7f7af2bbaa.
2021-07-31 09:04:01 +12:00
370ae8c20c Add support for mult-doc YAML files (#3870)
Single doc YAML files are returned as before. Multi-doc YAML files are
wrapped in an outer Table. Empty YAML files return Nothing.
2021-07-30 15:45:19 -05:00
69083bfca0 Adding parse fix for power operator error on negative integers and de… (#3821)
* Adding parse fix for power operator error on negative integers and decimal

* Adding correct formatting

* Changed is negative check to follow conventions

* Adding tests

* Added fix for the negatives and added tests

* Removed comments
2021-07-30 07:08:57 -05:00
1e15f26e98 fix interpolated strings when using unicode (#3866)
* fix interpolated strings when using unicode

* added test case
2021-07-29 19:07:34 -05:00
23ba01d89c add performance metrics for measuring startup time (#3854)
* add performance metrics for measureing startup time

* removed some comments

* update so tests pass

* update default.toml for tests, merged main

* fix clippy lints

* wording changes
2021-07-29 18:52:40 -05:00
653cbe651f Going deeper (#3864)
* nuframe in its own type in UntaggedValue

* Removed eager dataframe from enum

* Dataframe created from list of values

* Corrected order in dataframe columns

* Returned tag from stream collection

* Removed series from dataframe commands

* Arithmetic operators

* forced push

* forced push

* Replace all command

* String commands

* appending operations with dfs

* Testing suite for dataframes

* Unit test for dataframe commands

* improved equality for dataframes

* moving all dataframe operations to protocol

* objects in dataframes

* Removed cloning when converting to row
2021-07-30 09:16:30 +12:00
JT
f3e487e829 Add named positionals to all (#3863) 2021-07-30 09:12:24 +12:00
JT
e602647d4d Fix clippy lint and disable broken lint (#3865) 2021-07-30 08:11:47 +12:00
9696e4d315 Improve md5 and sha256 code (#3841)
* Refactor Hash code to simplify md5 and sha256 implementations

Md5 and Sha256 (and other future digests) require less boilerplate code
now. Error reporting includues the name of the hash again.

* Add missing hash sha256 test
2021-07-29 10:22:16 -05:00
7f7af2bbaa Unify use of the surf crate (#3855)
This brings the features used by the `nu_plugin_fetch` and
`nu_plugin_post` in line and drops the default-features, reducing
the number of pulled-in dependencies and avoiding a second round of
compilations.

Retry of #3777 but with different features, post and fetch plugin tested

Signed-off-by: Daniel Egger <daniel@eggers-club.de>
2021-07-29 19:26:38 +12:00
b190051e15 Fix select to insert nulls in sparse tables instead of ignoring absent values (#3857)
Select used to ignore absent values resulting in "squashing" where
columns for multiple rows could be squashed into a single row.

This fixes the problem by inserting null/nothing into absent value, thus
preserving the structure of rows. This makes sure that all selected
columns are present in each row.
2021-07-28 16:39:42 -05:00
83b28cad8d Remove dependencies (#3853)
* nuframe in its own type in UntaggedValue

* Removed eager dataframe from enum

* Dataframe created from list of values

* Corrected order in dataframe columns

* Returned tag from stream collection

* Removed series from dataframe commands

* Arithmetic operators

* forced push

* forced push

* Replace all command

* String commands

* appending operations with dfs

* Testing suite for dataframes

* Unit test for dataframe commands

* improved equality for dataframes

* moving all dataframe operations to protocol
2021-07-27 14:20:06 -05:00
e4c282f0a6 added the ability to compare time units like 1hr < 2hr (#3845) 2021-07-26 15:19:32 -05:00
d54d7cc431 append dataframes (#3839) 2021-07-26 08:36:09 +12:00
111477aa74 Add sha256 to the hash command (#3836)
Hashers now uses on Rust Crypto Digest trait which makes it trivial to
implement additional hash functions.

The original `md5` crate does not implement the Digest trait and was
replaced by `md-5` crate which does. Sha256 uses already included `sha2`
crate.
2021-07-25 14:08:08 -05:00
JT
226739d13f Bump to 0.34.1 (#3835) 2021-07-25 22:58:33 +12:00
f1ee9113ac All is a DataFrame (#3812)
* nuframe in its own type in UntaggedValue

* Removed eager dataframe from enum

* Dataframe created from list of values

* Corrected order in dataframe columns

* Returned tag from stream collection

* Removed series from dataframe commands

* Arithmetic operators

* forced push

* forced push

* Replace all command

* String commands

* appending operations with dfs

* Testing suite for dataframes

* Unit test for dataframe commands

* improved equality for dataframes
2021-07-25 22:01:54 +12:00
9120a64cfb use chrono_humanize for datetime formatting (#3834)
* use chrono_humanize for datetime formatting

* fix tests
2021-07-25 20:38:45 +12:00
fcd624a722 add date humanize command (#3833)
* add `date humanize` command

* add docs
2021-07-25 17:33:31 +12:00
e6af7f75a1 Read from standard input in rm (#3763)
* Read from standard input in `rm`

With this change, rm falls back to reading from the standard input if no
arguments are supplied. This leads to more intuitive pipes as seen in
https://github.com/nushell/nushell/issues/2824

 ls | get name | sort-by | first 20 | each {rm $it} becomes
 ls | get name | sort-by | first 20 | rm

* [Fix] Run cargo fmt, make files a rest parameter, and fix cargo build warnings

* [Fix] Fix clippy suggestions
2021-07-25 15:01:53 +12:00
27f1e7b60c change describe so it doesn't output colored strings (#3832) 2021-07-25 06:34:11 +12:00
2e24de7f47 Support other variables than PATH in pathvar (2nd attempt) (#3828)
* Fix swapped PATH env var separators

* Support pathvar to manipulate other vars than PATH

* Add tests for pathvar and its subcommands

* Adjust pathvar tests to comply with env quirks

* Make pathvar tests work on non-Windows as well

* Compact the comments in pathvar tests

* Fix last failing test

Co-authored-by: Jakub Žádník <jakub.zadnik@tuni.fi>
2021-07-24 11:44:36 -05:00