nushell/crates/nu-cli/src/util.rs
Devyn Cairns d7392f1f3b
Internal representation (IR) compiler and evaluator (#13330)
# Description

This PR adds an internal representation language to Nushell, offering an
alternative evaluator based on simple instructions, stream-containing
registers, and indexed control flow. The number of registers required is
determined statically at compile-time, and the fixed size required is
allocated upon entering the block.

Each instruction is associated with a span, which makes going backwards
from IR instructions to source code very easy.

Motivations for IR:

1. **Performance.** By simplifying the evaluation path and making it
more cache-friendly and branch predictor-friendly, code that does a lot
of computation in Nushell itself can be sped up a decent bit. Because
the IR is fairly easy to reason about, we can also implement
optimization passes in the future to eliminate and simplify code.
2. **Correctness.** The instructions mostly have very simple and
easily-specified behavior, so hopefully engine changes are a little bit
easier to reason about, and they can be specified in a more formal way
at some point. I have made an effort to document each of the
instructions in the docs for the enum itself in a reasonably specific
way. Some of the errors that would have happened during evaluation
before are now moved to the compilation step instead, because they don't
make sense to check during evaluation.
3. **As an intermediate target.** This is a good step for us to bring
the [`new-nu-parser`](https://github.com/nushell/new-nu-parser) in at
some point, as code generated from new AST can be directly compared to
code generated from old AST. If the IR code is functionally equivalent,
it will behave the exact same way.
4. **Debugging.** With a little bit more work, we can probably give
control over advancing the virtual machine that `IrBlock`s run on to
some sort of external driver, making things like breakpoints and single
stepping possible. Tools like `view ir` and [`explore
ir`](https://github.com/devyn/nu_plugin_explore_ir) make it easier than
before to see what exactly is going on with your Nushell code.

The goal is to eventually replace the AST evaluator entirely, once we're
sure it's working just as well. You can help dogfood this by running
Nushell with `$env.NU_USE_IR` set to some value. The environment
variable is checked when Nushell starts, so config runs with IR, or it
can also be set on a line at the REPL to change it dynamically. It is
also checked when running `do` in case within a script you want to just
run a specific piece of code with or without IR.

# Example

```nushell
view ir { |data|
  mut sum = 0
  for n in $data {
    $sum += $n
  }
  $sum
}
```
  
```gas
# 3 registers, 19 instructions, 0 bytes of data
   0: load-literal           %0, int(0)
   1: store-variable         var 904, %0 # let
   2: drain                  %0
   3: drop                   %0
   4: load-variable          %1, var 903
   5: iterate                %0, %1, end 15 # for, label(1), from(14:)
   6: store-variable         var 905, %0
   7: load-variable          %0, var 904
   8: load-variable          %2, var 905
   9: binary-op              %0, Math(Plus), %2
  10: span                   %0
  11: store-variable         var 904, %0
  12: load-literal           %0, nothing
  13: drain                  %0
  14: jump                   5
  15: drop                   %0          # label(0), from(5:)
  16: drain                  %0
  17: load-variable          %0, var 904
  18: return                 %0
```

# Benchmarks

All benchmarks run on a base model Mac Mini M1.

## Iterative Fibonacci sequence

This is about as best case as possible, making use of the much faster
control flow. Most code will not experience a speed improvement nearly
this large.

```nushell
def fib [n: int] {
  mut a = 0
  mut b = 1
  for _ in 2..=$n {
    let c = $a + $b
    $a = $b
    $b = $c
  }
  $b
}
use std bench
bench { 0..50 | each { |n| fib $n } }
```

IR disabled:

```
╭───────┬─────────────────╮
│ mean  │ 1ms 924µs 665ns │
│ min   │ 1ms 700µs 83ns  │
│ max   │ 3ms 450µs 125ns │
│ std   │ 395µs 759ns     │
│ times │ [list 50 items] │
╰───────┴─────────────────╯
```

IR enabled:

```
╭───────┬─────────────────╮
│ mean  │ 452µs 820ns     │
│ min   │ 427µs 417ns     │
│ max   │ 540µs 167ns     │
│ std   │ 17µs 158ns      │
│ times │ [list 50 items] │
╰───────┴─────────────────╯
```

![explore ir
view](https://github.com/nushell/nushell/assets/10729/d7bccc03-5222-461c-9200-0dce71b83b83)

##
[gradient_benchmark_no_check.nu](https://github.com/nushell/nu_scripts/blob/main/benchmarks/gradient_benchmark_no_check.nu)

IR disabled:

```
╭───┬──────────────────╮
│ 0 │ 27ms 929µs 958ns │
│ 1 │ 21ms 153µs 459ns │
│ 2 │ 18ms 639µs 666ns │
│ 3 │ 19ms 554µs 583ns │
│ 4 │ 13ms 383µs 375ns │
│ 5 │ 11ms 328µs 208ns │
│ 6 │  5ms 659µs 542ns │
╰───┴──────────────────╯
```

IR enabled:

```
╭───┬──────────────────╮
│ 0 │       22ms 662µs │
│ 1 │ 17ms 221µs 792ns │
│ 2 │ 14ms 786µs 708ns │
│ 3 │ 13ms 876µs 834ns │
│ 4 │  13ms 52µs 875ns │
│ 5 │ 11ms 269µs 666ns │
│ 6 │  6ms 942µs 500ns │
╰───┴──────────────────╯
```

##
[random-bytes.nu](https://github.com/nushell/nu_scripts/blob/main/benchmarks/random-bytes.nu)

I got pretty random results out of this benchmark so I decided not to
include it. Not clear why.

# User-Facing Changes
- IR compilation errors may appear even if the user isn't evaluating
with IR.
- IR evaluation can be enabled by setting the `NU_USE_IR` environment
variable to any value.
- New command `view ir` pretty-prints the IR for a block, and `view ir
--json` can be piped into an external tool like [`explore
ir`](https://github.com/devyn/nu_plugin_explore_ir).

# Tests + Formatting
All tests are passing with `NU_USE_IR=1`, and I've added some more eval
tests to compare the results for some very core operations. I will
probably want to add some more so we don't have to always check
`NU_USE_IR=1 toolkit test --workspace` on a regular basis.

# After Submitting
- [ ] release notes
- [ ] further documentation of instructions?
- [ ] post-release: publish `nu_plugin_explore_ir`
2024-07-10 17:33:59 -07:00

337 lines
10 KiB
Rust

use nu_cmd_base::hook::eval_hook;
use nu_engine::{eval_block, eval_block_with_early_return};
use nu_parser::{escape_quote_string, lex, parse, unescape_unquote_string, Token, TokenContents};
use nu_protocol::{
debugger::WithoutDebug,
engine::{EngineState, Stack, StateWorkingSet},
report_error, report_error_new, PipelineData, ShellError, Span, Value,
};
#[cfg(windows)]
use nu_utils::enable_vt_processing;
use nu_utils::perf;
use std::path::Path;
// This will collect environment variables from std::env and adds them to a stack.
//
// In order to ensure the values have spans, it first creates a dummy file, writes the collected
// env vars into it (in a "NAME"="value" format, quite similar to the output of the Unix 'env'
// tool), then uses the file to get the spans. The file stays in memory, no filesystem IO is done.
//
// The "PWD" env value will be forced to `init_cwd`.
// The reason to use `init_cwd`:
//
// While gathering parent env vars, the parent `PWD` may not be the same as `current working directory`.
// Consider to the following command as the case (assume we execute command inside `/tmp`):
//
// tmux split-window -v -c "#{pane_current_path}"
//
// Here nu execute external command `tmux`, and tmux starts a new `nushell`, with `init_cwd` value "#{pane_current_path}".
// But at the same time `PWD` still remains to be `/tmp`.
//
// In this scenario, the new `nushell`'s PWD should be "#{pane_current_path}" rather init_cwd.
pub fn gather_parent_env_vars(engine_state: &mut EngineState, init_cwd: &Path) {
gather_env_vars(std::env::vars(), engine_state, init_cwd);
}
fn gather_env_vars(
vars: impl Iterator<Item = (String, String)>,
engine_state: &mut EngineState,
init_cwd: &Path,
) {
fn report_capture_error(engine_state: &EngineState, env_str: &str, msg: &str) {
report_error_new(
engine_state,
&ShellError::GenericError {
error: format!("Environment variable was not captured: {env_str}"),
msg: "".into(),
span: None,
help: Some(msg.into()),
inner: vec![],
},
);
}
fn put_env_to_fake_file(name: &str, val: &str, fake_env_file: &mut String) {
fake_env_file.push_str(&escape_quote_string(name));
fake_env_file.push('=');
fake_env_file.push_str(&escape_quote_string(val));
fake_env_file.push('\n');
}
let mut fake_env_file = String::new();
// Write all the env vars into a fake file
for (name, val) in vars {
put_env_to_fake_file(&name, &val, &mut fake_env_file);
}
match init_cwd.to_str() {
Some(cwd) => {
put_env_to_fake_file("PWD", cwd, &mut fake_env_file);
}
None => {
// Could not capture current working directory
report_error_new(
engine_state,
&ShellError::GenericError {
error: "Current directory is not a valid utf-8 path".into(),
msg: "".into(),
span: None,
help: Some(format!(
"Retrieving current directory failed: {init_cwd:?} not a valid utf-8 path"
)),
inner: vec![],
},
);
}
}
// Lex the fake file, assign spans to all environment variables and add them
// to stack
let span_offset = engine_state.next_span_start();
engine_state.add_file(
"Host Environment Variables".into(),
fake_env_file.as_bytes().into(),
);
let (tokens, _) = lex(fake_env_file.as_bytes(), span_offset, &[], &[], true);
for token in tokens {
if let Token {
contents: TokenContents::Item,
span: full_span,
} = token
{
let contents = engine_state.get_span_contents(full_span);
let (parts, _) = lex(contents, full_span.start, &[], &[b'='], true);
let name = if let Some(Token {
contents: TokenContents::Item,
span,
}) = parts.first()
{
let mut working_set = StateWorkingSet::new(engine_state);
let bytes = working_set.get_span_contents(*span);
if bytes.len() < 2 {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got empty name.",
);
continue;
}
let (bytes, err) = unescape_unquote_string(bytes, *span);
if let Some(err) = err {
working_set.error(err);
}
if working_set.parse_errors.first().is_some() {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got unparsable name.",
);
continue;
}
bytes
} else {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got empty name.",
);
continue;
};
let value = if let Some(Token {
contents: TokenContents::Item,
span,
}) = parts.get(2)
{
let mut working_set = StateWorkingSet::new(engine_state);
let bytes = working_set.get_span_contents(*span);
if bytes.len() < 2 {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got empty value.",
);
continue;
}
let (bytes, err) = unescape_unquote_string(bytes, *span);
if let Some(err) = err {
working_set.error(err);
}
if working_set.parse_errors.first().is_some() {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got unparsable value.",
);
continue;
}
Value::string(bytes, *span)
} else {
report_capture_error(
engine_state,
&String::from_utf8_lossy(contents),
"Got empty value.",
);
continue;
};
// stack.add_env_var(name, value);
engine_state.add_env_var(name, value);
}
}
}
pub fn eval_source(
engine_state: &mut EngineState,
stack: &mut Stack,
source: &[u8],
fname: &str,
input: PipelineData,
allow_return: bool,
) -> i32 {
let start_time = std::time::Instant::now();
let exit_code = match evaluate_source(engine_state, stack, source, fname, input, allow_return) {
Ok(code) => code.unwrap_or(0),
Err(err) => {
report_error_new(engine_state, &err);
1
}
};
stack.add_env_var(
"LAST_EXIT_CODE".to_string(),
Value::int(exit_code.into(), Span::unknown()),
);
// reset vt processing, aka ansi because illbehaved externals can break it
#[cfg(windows)]
{
let _ = enable_vt_processing();
}
perf!(
&format!("eval_source {}", &fname),
start_time,
engine_state.get_config().use_ansi_coloring
);
exit_code
}
fn evaluate_source(
engine_state: &mut EngineState,
stack: &mut Stack,
source: &[u8],
fname: &str,
input: PipelineData,
allow_return: bool,
) -> Result<Option<i32>, ShellError> {
let (block, delta) = {
let mut working_set = StateWorkingSet::new(engine_state);
let output = parse(
&mut working_set,
Some(fname), // format!("entry #{}", entry_num)
source,
false,
);
if let Some(warning) = working_set.parse_warnings.first() {
report_error(&working_set, warning);
}
if let Some(err) = working_set.parse_errors.first() {
report_error(&working_set, err);
return Ok(Some(1));
}
if let Some(err) = working_set.compile_errors.first() {
report_error(&working_set, err);
// Not a fatal error, for now
}
(output, working_set.render())
};
engine_state.merge_delta(delta)?;
let pipeline = if allow_return {
eval_block_with_early_return::<WithoutDebug>(engine_state, stack, &block, input)
} else {
eval_block::<WithoutDebug>(engine_state, stack, &block, input)
}?;
let status = if let PipelineData::ByteStream(..) = pipeline {
pipeline.print(engine_state, stack, false, false)?
} else {
if let Some(hook) = engine_state.get_config().hooks.display_output.clone() {
let pipeline = eval_hook(
engine_state,
stack,
Some(pipeline),
vec![],
&hook,
"display_output",
)?;
pipeline.print(engine_state, stack, false, false)
} else {
pipeline.print(engine_state, stack, true, false)
}?
};
Ok(status.map(|status| status.code()))
}
#[cfg(test)]
mod test {
use super::*;
#[test]
fn test_gather_env_vars() {
let mut engine_state = EngineState::new();
let symbols = r##" !"#$%&'()*+,-./:;<=>?@[\]^_`{|}~"##;
gather_env_vars(
[
("FOO".into(), "foo".into()),
("SYMBOLS".into(), symbols.into()),
(symbols.into(), "symbols".into()),
]
.into_iter(),
&mut engine_state,
Path::new("t"),
);
let env = engine_state.render_env_vars();
assert!(
matches!(env.get(&"FOO".to_string()), Some(&Value::String { val, .. }) if val == "foo")
);
assert!(
matches!(env.get(&"SYMBOLS".to_string()), Some(&Value::String { val, .. }) if val == symbols)
);
assert!(
matches!(env.get(&symbols.to_string()), Some(&Value::String { val, .. }) if val == "symbols")
);
assert!(env.get(&"PWD".to_string()).is_some());
assert_eq!(env.len(), 4);
}
}