nushell/crates/nu-command/src/filesystem/save.rs
JT 1e3e034021
Spanned Value step 1: span all value cases (#10042)
# Description

This doesn't really do much that the user could see, but it helps get us
ready to do the steps of the refactor to split the span off of Value, so
that values can be spanless. This allows us to have top-level values
that can hold both a Value and a Span, without requiring that all values
have them.

We expect to see significant memory reduction by removing so many
unnecessary spans from values. For example, a table of 100,000 rows and
5 columns would have a savings of ~8megs in just spans that are almost
always duplicated.

# User-Facing Changes

Nothing yet

# 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 -- -c "use std testing; testing run-tests --path
crates/nu-std"` 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.
-->
2023-08-25 08:48:05 +12:00

434 lines
14 KiB
Rust

use nu_engine::current_dir;
use nu_engine::CallExt;
use nu_path::expand_path_with;
use nu_protocol::ast::Call;
use nu_protocol::engine::{Command, EngineState, Stack};
use nu_protocol::{
Category, Example, PipelineData, RawStream, ShellError, Signature, Span, Spanned, SyntaxShape,
Type, Value,
};
use std::fs::File;
use std::io::Write;
use std::path::{Path, PathBuf};
use std::thread;
use crate::progress_bar;
#[derive(Clone)]
pub struct Save;
impl Command for Save {
fn name(&self) -> &str {
"save"
}
fn usage(&self) -> &str {
"Save a file."
}
fn search_terms(&self) -> Vec<&str> {
vec![
"write",
"write_file",
"append",
"redirection",
"file",
"io",
">",
">>",
]
}
fn signature(&self) -> nu_protocol::Signature {
Signature::build("save")
.input_output_types(vec![(Type::Any, Type::Nothing)])
.required("filename", SyntaxShape::Filepath, "the filename to use")
.named(
"stderr",
SyntaxShape::Filepath,
"the filename used to save stderr, only works with `-r` flag",
Some('e'),
)
.switch("raw", "save file as raw binary", Some('r'))
.switch("append", "append input to the end of the file", Some('a'))
.switch("force", "overwrite the destination", Some('f'))
.switch("progress", "enable progress bar", Some('p'))
.category(Category::FileSystem)
}
fn run(
&self,
engine_state: &EngineState,
stack: &mut Stack,
call: &Call,
input: PipelineData,
) -> Result<PipelineData, ShellError> {
let raw = call.has_flag("raw");
let append = call.has_flag("append");
let force = call.has_flag("force");
let progress = call.has_flag("progress");
let span = call.head;
let cwd = current_dir(engine_state, stack)?;
let path_arg = call.req::<Spanned<PathBuf>>(engine_state, stack, 0)?;
let path = Spanned {
item: expand_path_with(path_arg.item, &cwd),
span: path_arg.span,
};
let stderr_path = call
.get_flag::<Spanned<PathBuf>>(engine_state, stack, "stderr")?
.map(|arg| Spanned {
item: expand_path_with(arg.item, cwd),
span: arg.span,
});
match input {
PipelineData::ExternalStream { stdout: None, .. } => {
// Open files to possibly truncate them
let _ = get_files(&path, &stderr_path, append, force)?;
Ok(PipelineData::empty())
}
PipelineData::ExternalStream {
stdout: Some(stream),
stderr,
..
} => {
let (file, stderr_file) = get_files(&path, &stderr_path, append, force)?;
// delegate a thread to redirect stderr to result.
let handler = stderr.map(|stderr_stream| match stderr_file {
Some(stderr_file) => thread::Builder::new()
.name("stderr redirector".to_string())
.spawn(move || stream_to_file(stderr_stream, stderr_file, span, progress))
.expect("Failed to create thread"),
None => thread::Builder::new()
.name("stderr redirector".to_string())
.spawn(move || {
let _ = stderr_stream.into_bytes();
Ok(PipelineData::empty())
})
.expect("Failed to create thread"),
});
let res = stream_to_file(stream, file, span, progress);
if let Some(h) = handler {
h.join().map_err(|err| ShellError::ExternalCommand {
label: "Fail to receive external commands stderr message".to_string(),
help: format!("{err:?}"),
span,
})??;
res
} else {
res
}
}
PipelineData::ListStream(ls, _)
if raw || prepare_path(&path, append, force)?.0.extension().is_none() =>
{
let (mut file, _) = get_files(&path, &stderr_path, append, force)?;
for val in ls {
file.write_all(&value_to_bytes(val)?)
.map_err(|err| ShellError::IOError(err.to_string()))?;
file.write_all("\n".as_bytes())
.map_err(|err| ShellError::IOError(err.to_string()))?;
}
file.flush()?;
Ok(PipelineData::empty())
}
input => {
let bytes =
input_to_bytes(input, Path::new(&path.item), raw, engine_state, stack, span)?;
// Only open file after successful conversion
let (mut file, _) = get_files(&path, &stderr_path, append, force)?;
file.write_all(&bytes)
.map_err(|err| ShellError::IOError(err.to_string()))?;
file.flush()?;
Ok(PipelineData::empty())
}
}
}
fn examples(&self) -> Vec<Example> {
vec![
Example {
description: "Save a string to foo.txt in the current directory",
example: r#"'save me' | save foo.txt"#,
result: None,
},
Example {
description: "Append a string to the end of foo.txt",
example: r#"'append me' | save --append foo.txt"#,
result: None,
},
Example {
description: "Save a record to foo.json in the current directory",
example: r#"{ a: 1, b: 2 } | save foo.json"#,
result: None,
},
Example {
description: "Save a running program's stderr to foo.txt",
example: r#"do -i {} | save foo.txt --stderr foo.txt"#,
result: None,
},
Example {
description: "Save a running program's stderr to separate file",
example: r#"do -i {} | save foo.txt --stderr bar.txt"#,
result: None,
},
]
}
}
/// Convert [`PipelineData`] bytes to write in file, possibly converting
/// to format of output file
fn input_to_bytes(
input: PipelineData,
path: &Path,
raw: bool,
engine_state: &EngineState,
stack: &mut Stack,
span: Span,
) -> Result<Vec<u8>, ShellError> {
let ext = if raw {
None
// if is extern stream , in other words , not value
} else if let PipelineData::ExternalStream { .. } = input {
None
} else if let PipelineData::Value(Value::String { .. }, ..) = input {
None
} else {
path.extension()
.map(|name| name.to_string_lossy().to_string())
};
if let Some(ext) = ext {
convert_to_extension(engine_state, &ext, stack, input, span)
} else {
let value = input.into_value(span);
value_to_bytes(value)
}
}
/// Convert given data into content of file of specified extension if
/// corresponding `to` command exists. Otherwise attempt to convert
/// data to bytes as is
fn convert_to_extension(
engine_state: &EngineState,
extension: &str,
stack: &mut Stack,
input: PipelineData,
span: Span,
) -> Result<Vec<u8>, ShellError> {
let converter = engine_state.find_decl(format!("to {extension}").as_bytes(), &[]);
let output = match converter {
Some(converter_id) => {
let output = engine_state.get_decl(converter_id).run(
engine_state,
stack,
&Call::new(span),
input,
)?;
output.into_value(span)
}
None => input.into_value(span),
};
value_to_bytes(output)
}
/// Convert [`Value::String`] [`Value::Binary`] or [`Value::List`] into [`Vec`] of bytes
///
/// Propagates [`Value::Error`] and creates error otherwise
fn value_to_bytes(value: Value) -> Result<Vec<u8>, ShellError> {
match value {
Value::String { val, .. } => Ok(val.into_bytes()),
Value::Binary { val, .. } => Ok(val),
Value::List { vals, .. } => {
let val = vals
.into_iter()
.map(|it| it.as_string())
.collect::<Result<Vec<String>, ShellError>>()?
.join("\n")
+ "\n";
Ok(val.into_bytes())
}
// Propagate errors by explicitly matching them before the final case.
Value::Error { error, .. } => Err(*error),
other => Ok(other.as_string()?.into_bytes()),
}
}
/// Convert string path to [`Path`] and [`Span`] and check if this path
/// can be used with given flags
fn prepare_path(
path: &Spanned<PathBuf>,
append: bool,
force: bool,
) -> Result<(&Path, Span), ShellError> {
let span = path.span;
let path = &path.item;
if !(force || append) && path.exists() {
Err(ShellError::GenericError(
"Destination file already exists".into(),
format!(
"Destination file '{}' already exists",
path.to_string_lossy()
),
Some(span),
Some("you can use -f, --force to force overwriting the destination".into()),
Vec::new(),
))
} else {
Ok((path, span))
}
}
fn open_file(path: &Path, span: Span, append: bool) -> Result<File, ShellError> {
let file = match (append, path.exists()) {
(true, true) => std::fs::OpenOptions::new()
.write(true)
.append(true)
.open(path),
_ => std::fs::File::create(path),
};
file.map_err(|err| {
ShellError::GenericError(
"Permission denied".into(),
err.to_string(),
Some(span),
None,
Vec::new(),
)
})
}
/// Get output file and optional stderr file
fn get_files(
path: &Spanned<PathBuf>,
stderr_path: &Option<Spanned<PathBuf>>,
append: bool,
force: bool,
) -> Result<(File, Option<File>), ShellError> {
// First check both paths
let (path, path_span) = prepare_path(path, append, force)?;
let stderr_path_and_span = stderr_path
.as_ref()
.map(|stderr_path| prepare_path(stderr_path, append, force))
.transpose()?;
// Only if both files can be used open and possibly truncate them
let file = open_file(path, path_span, append)?;
let stderr_file = stderr_path_and_span
.map(|(stderr_path, stderr_path_span)| {
if path == stderr_path {
Err(ShellError::GenericError(
"input and stderr input to same file".to_string(),
"can't save both input and stderr input to the same file".to_string(),
Some(stderr_path_span),
Some("you should use `o+e> file` instead".to_string()),
vec![],
))
} else {
open_file(stderr_path, stderr_path_span, append)
}
})
.transpose()?;
Ok((file, stderr_file))
}
fn stream_to_file(
mut stream: RawStream,
file: File,
span: Span,
progress: bool,
) -> Result<PipelineData, ShellError> {
// https://github.com/nushell/nushell/pull/9377 contains the reason
// for not using BufWriter<File>
let mut writer = file;
let mut bytes_processed: u64 = 0;
let bytes_processed_p = &mut bytes_processed;
let file_total_size = stream.known_size;
let mut process_failed = false;
let process_failed_p = &mut process_failed;
// Create the progress bar
// It looks a bit messy but I am doing it this way to avoid
// creating the bar when is not needed
let (mut bar_opt, bar_opt_clone) = if progress {
let tmp_bar = progress_bar::NuProgressBar::new(file_total_size);
let tmp_bar_clone = tmp_bar.clone();
(Some(tmp_bar), Some(tmp_bar_clone))
} else {
(None, None)
};
let result = stream
.try_for_each(move |result| {
let buf = match result {
Ok(v) => match v {
Value::String { val, .. } => val.into_bytes(),
Value::Binary { val, .. } => val,
// Propagate errors by explicitly matching them before the final case.
Value::Error { error, .. } => return Err(*error),
other => {
return Err(ShellError::OnlySupportsThisInputType {
exp_input_type: "string or binary".into(),
wrong_type: other.get_type().to_string(),
dst_span: span,
src_span: other.span(),
});
}
},
Err(err) => {
*process_failed_p = true;
return Err(err);
}
};
// If the `progress` flag is set then
if progress {
// Update the total amount of bytes that has been saved and then print the progress bar
*bytes_processed_p += buf.len() as u64;
if let Some(bar) = &mut bar_opt {
bar.update_bar(*bytes_processed_p);
}
}
if let Err(err) = writer.write(&buf) {
*process_failed_p = true;
return Err(ShellError::IOError(err.to_string()));
}
Ok(())
})
.map(|_| PipelineData::empty());
// If the `progress` flag is set then
if progress {
// If the process failed, stop the progress bar with an error message.
if process_failed {
if let Some(bar) = bar_opt_clone {
bar.abandoned_msg("# Error while saving #".to_owned());
}
}
}
// And finally return the stream result.
result
}