* Begin allowing comments and multiline scripts.
* clippy
* Finish moving to groups. Test pass
* Keep going
* WIP
* WIP
* BROKEN WIP
* WIP
* WIP
* Fix more tests
* WIP: alias starts working
* Broken WIP
* Broken WIP
* Variables begin to work
* captures start working
* A little better but needs fixed scope
* Shorthand env setting
* Update main merge
* Broken WIP
* WIP
* custom command parsing
* Custom commands start working
* Fix coloring and parsing of block
* Almost there
* Add some tests
* Add more param types
* Bump version
* Fix benchmark
* Fix stuff
v0.29.0 and earlier versions of `uom` fail to compile on nightly because
of now-ambiguous trait bounds. The issue was corrected in v0.30.0 of
`uom`. `uom` and `heim` dependencies have been updated to the
latest version to include this fix and allow nushell to compile on
nightly.
Co-authored-by: Boutin, Michael <mjboutin@ecolab.com>
* updated & added date related commands based on the new design
* added proper error handling when date format string is invalid
* fixed format issue
* fixed an issue caused due to the change in primitive Date type
* added `date list-timezone` command to list all supported time zones and updated `date to-timezone` accordingly
This makes the errors slightly better. It took me a while to realize I was missing the `--raw` flag.
```
open "data.csv" | from csv --separator ';'
```
error: Could not parse as CSV split by ',' (Line 1: expected 1 fields, found 14)
┌─ shell:1:1
│
1 │ open "data.csv" | from csv --separator ';'
│ ^^^^ ------------------------------------------------- value originates from here
│ │
│ input cannot be parsed as CSV split by ','. Sample input:
Name;Data
Ugly;row
AnotherUgly;row
I think this still needs some refinement. Maybe we don't want to show
the separator all the time, omitting the defaults or the separator
on other formats.
* Bump Rustyline to 7.0.0
* Append history instead of always save
* Add associated type to Hinter
* Convert to using Rustyline KeyEvent
* Use AcceptOrInsertLine as struct
* Cargo fmt
* Make convert_keyevent pub
* Better naming for RL conversion
* WIP try testing hash command
Ensure test worked
fmt
WIP get it working for other types of base64
Use optional named arg
WIP
* rebased and refactored a little with encoding and decoding
Fix some typos
Add some more charactersets
refactor several args into the encoding config struct and fix character_set arg. It needs to match the field
Add main hash command so it can be found via help
Added tests for running the whole pipeline
* add test case to cover invalid character sets
* clippy and fmt
* Add string argument support for path subcommands
* Add --replace option to 'path extension' command
* Add examples of replacing for path extension
* Refactor path extension and its example
* Add replacement functionality to path basename
* Refactor path subcommands to support more args
This adds a lot of redundancy to non-relevant subcommands such as type,
exists or expand.
* Add replace and num_levels options to path dirname
* Rename num_levels option to num-levels
* Remove commented code
* Clean up path basename
* Fix path dirname description
* Add path filestem opts; Rename extension -> suffix
* Add prefix option and examples to path filestem
* Fix broken num-levels of path dirname
* Fix failing example test of path filestem
* Fix failing test of path extension
* Formatting
* Add Windows-specific path subcommand examples
`path expand` is still broken but otherwise seems to fix all examples
on Windows
* Fix weird path expand on Windows
Also disable example tests for path expand. Failed caconicalization
(e.g., due to path not existing) returns the original path so the
examples always fail.
* Formatting
* Return path datatype when appropriate
* Do not append empty remainder to path dirname
* Add tests for path subcommands
* Formatting
* Revisit path subcommand description strings
* Apply clippy suggestions; Formatting
* Remove problematic test checking '~' expansion
Wouldn't run on minimal due to useing optional dependency.
The test success was also deending on the presence of home dir on the
testing machine which might not be completely robust.
* Add missing newline to file
* first step of making selector
* wip
* wip tests working
* probably good enough for a first pass
* oops, missed something.
* and something else...
* grrrr version errors
* seq command - WIP
* why, oh why
* works with parameters
* widths should've been optional
* dbg messages
* working. rest had to be first.
* updated so that it outputs a table instead of just strings
* made to work with floats, allowed separator be more than 1 char
* clippy
* fixed tests
* changed terminator help desc
* commit to get ci moving again
* Change alias shape inference to proposal of RFC#4
* Remove commented code
* Fix typo
* Change comment to be more informative
* Make match statement to lookup in table
* Remove resolved question
https://github.com/nushell/nushell/pull/2685#discussion_r509832054
* Pick ...or_insert_dependency functions into pieces
Previously there was get_shape_of_expr_or_insert dependency, now there is
get_shape_of_expr and get_shape_of_expr_or_insert_dependency
2 new functions have been added: get_result_shape_of_math_expr and
get_result_shape_of_math_expr_or_insert_dependency
* Remove flattening of deep binary expressions
Previously deep binary expressions have been flattened through the insertion of
fake vars. This logic was quite complicated. Now if a variable depends on the
result shape of a binary expression and the result shape can't be computed,
the variable simply depends on the whole binary.
* Change Expression::Variable(Variable::It(...)) to Expression::Variable(...)
* Simplify get_result_shapes_in_math_expr
* Simplify infer_shapes_in_binary_expr
* Clarify comment
* Clarify comment
* Fix clippy lint
* Move check for real var into checked_insert
* Remove comment
* Rename var
* added math round
* added math floor
* added math ceil
* added math.md examples
* moved the detection of nonnumerical values in ceil/floor/round
* math round now works on streams
* math floor now works on streams
* math ceil now works on streams
Continuing on anchoring and improvements on Nu's overall internal commands (#2635).
`move column` sub command has been turned into the command `move` since
we use it to move exclusively columns. Examples added as well.
Fixed it to carry along any anchor locations that might be in place if
table to be moved originates from other sources.
* Add parser improvements
Previously everything starting with "$" was parsed as a column path.
With this commit applied, the lite_arg starting with $ is parsed as
the most appropriate thing
- $true/$false ==> Expression::Boolean
- $(...) ==> Invocation
- $it ==> ColumnPath
- Anything with at least one '.' ==> ColumnPath
- Anything else ==> Variable
* Ignore failing tests
* refactor and cleanup to md
* Add padding around values in each row
* Add padding to test
* Update code to satisfy Clippy and pass other failing tests
* make sort-by fail gracefully if mismatched types are compared
* Added a test to check if sorted-by with invalid types exists gracefully
* Linter changes
* removed redundant pattern matching
* Changed the error message
* Added a comma after every argument
* Changed the test to accomodate the new err messages
* Err message for sort-by invalid types now shows the mismatched types
* Lints problems
* Changed unwrap to expect
* Added the -f flag to rm command
Now when you a use rm -f there will be no error message, even if the
file doesnt actually exist
* Lint problems
* Fixed the wrong line
* Removed println
* Spelling mistake
* Fix problems when you mv a file into itself
* Lint mistakes
* Remove unecessary filtering in most cases
* Allow the removal of sockets
* Conditional compilations to systems without socket
* Add a size-format option to ls command
* Added kib and mib formating
* Make patterns lowercase
* New subcommand to format, filesize
* Forgot the linter once more
* Remove the ls changes since its no longer needed
* CI mistakes
* Lint stuff
* Fix lint
* Added formatting for bytes
* fix lint
* Changed the usage comment
* First draft for adding a `pretty` flag to `to md`
* rustfmt
* Fix Clippy warnings
* rustfmt
* Using Clippy suggestion broken code, reverting and putting in a statement to ignore clippy warning
* Add test for `to md -p`
Multiline pastes wait for the user to hit enter before running,
because they enter a special paste mode in rustyline called
'bracketed paste' by default. This commit disables that mode
by default for nushell, causing multiline pastes to be executed
immediately, treating each new line as a separate command.
* xpath prototype
* new xpath engine is finally working
* nearly there
* closer
* working with list, started to add test, code cleanup
* broken again
* working again - time for some cleanup
* cleaned up code, added error handling and test
* update example, fix clippy
* removed commented char
Nu has many commands that allow the nuño to customize behavior such
as UI and behavior. Today, coloring can be customized, the line editor,
and other things. The more options there are, the higher the complexity
in managing them.
To mitigate this Nu can store configuration options as nested properties.
But to add and edit them can be taxing. With column path support we can
work with them easier.
* [wasi] Update time & instant crates
In https://github.com/nushell/nushell/pull/2643 instant was updated by adding it as a hard dependency in Cargo.toml, but it's better to avoid it and only update in Cargo.lock via `cargo update -p ...`.
Additionally, updated `time` crate so that now some basic commands like `ls` work too, although formatting is pretty bad.
* Update default terminal width to 80
If termsize can't return anything, use 80 chars (e.g. on WASI).
* make sort-by fail gracefully if mismatched types are compared
* Added a test to check if sorted-by with invalid types exists gracefully
* Linter changes
* removed redundant pattern matching
* Changed the error message
* Added a comma after every argument
* Changed the test to accomodate the new err messages
* Err message for sort-by invalid types now shows the mismatched types
* Lints problems
* Changed unwrap to expect
* Added the -f flag to rm command
Now when you a use rm -f there will be no error message, even if the
file doesnt actually exist
* Lint problems
* Fixed the wrong line
* Removed println
* Spelling mistake
* Fix problems when you mv a file into itself
* Lint mistakes
* Remove unecessary filtering in most cases
* Allow the removal of sockets
* Conditional compilations to systems without socket
* make sort-by fail gracefully if mismatched types are compared
* Added a test to check if sorted-by with invalid types exists gracefully
* Linter changes
* removed redundant pattern matching
* Changed the error message
* Added a comma after every argument
* Changed the test to accomodate the new err messages
* Err message for sort-by invalid types now shows the mismatched types
* Lints problems
* Changed unwrap to expect
* Added the -f flag to rm command
Now when you a use rm -f there will be no error message, even if the
file doesnt actually exist
* Lint problems
* Fixed the wrong line
* Removed println
* Spelling mistake
* Fix problems when you mv a file into itself
* Lint mistakes
* Remove unecessary filtering in most cases
* Refactor scope to have parents
* Refactor scope to have parents
* Refactor scope to have parents
* Clippy
Co-authored-by: Jonathan Turner <jonathan@pop-os.localdomain>
* Implement passthrough for benchmark
Add a new option -p,--passthrough to benchmark.
With this option, the benchmark command prints its results to stdout and passes the block's output to the next command in the pipeline.
* Add execution block for benchmark -p
`benchmark --passthrough` now takes a block where the benchmark output is sent.
Example:
`benchmark -p {save bench.toml} {ls}`
* Add global mode to str trim
The global mode allows skipping non-string values,
and processes rows and tables as well
* Add tests to action with ActionMode::Global
* Add system, user and idle times to benchmark command
* Feature-gate dependency on heim in benchmark
* Reorder let bindings in benchmark
* Fully feature-gate rich-benchmark and print 0sec on zero duration
* Remove completion code from help/value shells
* Tweak the `Completer` trait for nushell.
Previously, this trait was built around rustyline's completion traits, and for
`Shell` instances. Now it is built for individual completers inside of nushell
that will complete a specific location based on a partial string. For example,
for completing a partially typed command in the command position.
We introduce the `plugin` nu sub command (`nu plugin`) with basic plugin
loading support. We can choose to load plugins from a directory. Originally
introduced to make integration tests faster (by not loading any plugins on startup at all)
but `nu plugin --load some_path ; test_pipeline_that_uses_plugins_just_loaded` does not see it.
Therefore, a `nu_with_plugins!` macro for tests was introduced on top of nu`s `--skip-plugins`
switch executable which is set to true when running the integration tests that use the `nu!` macro now..
* Implement exclusive and inclusive ranges with .. and ..=
This commit adds right-exclusive ranges.
The original a..b inclusive syntax was changed to reflect the Rust notation.
New a..=b syntax was introduced to have the old behavior.
Currently, both a.. and b..= is valid, and it is unclear whether it's valid
to impose restrictions.
The original issue suggests .. for inclusive and ..< for exclusive ranges,
this can be implemented by making simple changes to this commit.
* Fix collect tests by changing ranges to ..=
* Fix clippy lints in exclusive range matching
* Implement exclusive ranges using `..<`
