Allow filesystem commands to access files with glob metachars in name (#10694)

(squashed version of #10557, clean commit history and review thread)

Fixes #10571, also potentially: #10364, #10211, #9558, #9310,


# Description
Changes processing of arguments to filesystem commands that are source
paths or globs.
Applies to `cp, cp-old, mv, rm, du` but not `ls` (because it uses a
different globbing interface) or `glob` (because it uses a different
globbing library).

The core of the change is to lookup the argument first as a file and
only glob if it is not. That way,
a path containing glob metacharacters can be referenced without glob
quoting, though it will have to be single quoted to avoid nushell
parsing.

Before: A file path that looks like a glob is not matched by the glob
specified as a (source) argument and takes some thinking about to
access. You might say the glob pattern shadows a file with the same
spelling.
```
> ls a*
╭───┬────────┬──────┬──────┬────────────────╮
│ # │  name  │ type │ size │    modified    │
├───┼────────┼──────┼──────┼────────────────┤
│ 0 │ a[bc]d │ file │  0 B │ 34 seconds ago │
│ 1 │ abd    │ file │  0 B │ now            │
│ 2 │ acd    │ file │  0 B │ now            │
╰───┴────────┴──────┴──────┴────────────────╯

> cp --verbose 'a[bc]d' dest
copied /home/bobhy/src/rust/work/r4/abd to /home/bobhy/src/rust/work/r4/dest/abd
copied /home/bobhy/src/rust/work/r4/acd to /home/bobhy/src/rust/work/r4/dest/acd

> ## Note -- a[bc]d *not* copied, and seemingly hard to access.
> cp --verbose 'a\[bc\]d' dest
Error:   × No matches found
   ╭─[entry #33:1:1]
 1 │ cp --verbose 'a\[bc\]d' dest
   ·              ─────┬────
   ·                   ╰── no matches found
   ╰────

> #.. but is accessible with enough glob quoting.
> cp --verbose 'a[[]bc[]]d' dest
copied /home/bobhy/src/rust/work/r4/a[bc]d to /home/bobhy/src/rust/work/r4/dest/a[bc]d
```
Before_2: if file has glob metachars but isn't a valid pattern, user
gets a confusing error:

```
> touch 'a[b'
> cp 'a[b' dest
Error:   × Pattern syntax error near position 30: invalid range pattern
   ╭─[entry #13:1:1]
 1 │ cp 'a[b' dest
   ·    ──┬──
   ·      ╰── invalid pattern
   ╰────
```

After: Args to cp, mv, etc. are tried first as literal files, and only
as globs if not found to be files.

```
> cp --verbose 'a[bc]d' dest
copied /home/bobhy/src/rust/work/r4/a[bc]d to /home/bobhy/src/rust/work/r4/dest/a[bc]d
> cp --verbose '[a][bc]d' dest
copied /home/bobhy/src/rust/work/r4/abd to /home/bobhy/src/rust/work/r4/dest/abd
copied /home/bobhy/src/rust/work/r4/acd to /home/bobhy/src/rust/work/r4/dest/acd
```
After_2: file with glob metachars but invalid pattern just works.
(though Windows does not allow file name to contain `*`.).

```
> cp --verbose 'a[b' dest
copied /home/bobhy/src/rust/work/r4/a[b to /home/bobhy/src/rust/work/r4/dest/a[b
```

So, with this fix, a file shadows a glob pattern with the same spelling.
If you have such a file and really want to use the glob pattern, you
will have to glob quote some of the characters in the pattern. I think
that's less confusing to the user: if ls shows a file with a weird name,
s/he'll still be able to copy, rename or delete it.

# User-Facing Changes
Could break some existing scripts. If user happened to have a file with
a globbish name but was using a glob pattern with the same spelling, the
new version will process the file and not expand the glob.

# 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` to
check that you're using the standard code style
- `cargo test --workspace` to check that all tests pass (on Windows make
sure to [enable developer
mode](https://learn.microsoft.com/en-us/windows/apps/get-started/developer-mode-features-and-debugging))
- `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.
-->

---------

Co-authored-by: Darren Schroeder <343840+fdncred@users.noreply.github.com>
This commit is contained in:
Bob Hyman 2023-10-18 11:31:15 -07:00 committed by GitHub
parent 88a87158c2
commit 09b3dab35d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
13 changed files with 400 additions and 145 deletions

4
Cargo.lock generated
View File

@ -2713,9 +2713,13 @@ dependencies = [
"indexmap 2.0.2", "indexmap 2.0.2",
"miette", "miette",
"nu-engine", "nu-engine",
"nu-glob",
"nu-parser", "nu-parser",
"nu-path", "nu-path",
"nu-protocol", "nu-protocol",
"nu-test-support",
"nu-utils",
"rstest",
] ]
[[package]] [[package]]

View File

@ -11,8 +11,15 @@ version = "0.86.1"
[dependencies] [dependencies]
nu-engine = { path = "../nu-engine", version = "0.86.1" } nu-engine = { path = "../nu-engine", version = "0.86.1" }
nu-glob = { path = "../nu-glob", version = "0.86.1" }
nu-parser = { path = "../nu-parser", version = "0.86.1" } nu-parser = { path = "../nu-parser", version = "0.86.1" }
nu-path = { path = "../nu-path", version = "0.86.1" } nu-path = { path = "../nu-path", version = "0.86.1" }
nu-protocol = { version = "0.86.1", path = "../nu-protocol" } nu-protocol = { path = "../nu-protocol", version = "0.86.1" }
indexmap = { version = "2.0" } nu-utils = { path = "../nu-utils", version = "0.86.1" }
miette = { version = "5.10", features = ["fancy-no-backtrace"] }
indexmap = "2.0"
miette = "5.10.0"
[dev-dependencies]
nu-test-support = { path = "../nu-test-support", version = "0.86.1" }
rstest = "0.18.2"

View File

@ -0,0 +1,207 @@
// utilities for expanding globs in command arguments
use nu_glob::{glob_with_parent, MatchOptions, Paths};
use nu_protocol::{ShellError, Spanned};
use std::fs;
use std::path::{Path, PathBuf};
// standard glob options to use for filesystem command arguments
const GLOB_PARAMS: MatchOptions = MatchOptions {
case_sensitive: true,
require_literal_separator: false,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
// handle an argument that could be a literal path or a glob.
// if literal path, return just that (whether user can access it or not).
// if glob, expand into matching paths, using GLOB_PARAMS options.
pub fn arg_glob(
pattern: &Spanned<String>, // alleged path or glob
cwd: &Path, // current working directory
) -> Result<Paths, ShellError> {
arg_glob_opt(pattern, cwd, GLOB_PARAMS)
}
// variant of [arg_glob] that requires literal dot prefix in pattern to match dot-prefixed path.
pub fn arg_glob_leading_dot(pattern: &Spanned<String>, cwd: &Path) -> Result<Paths, ShellError> {
arg_glob_opt(
pattern,
cwd,
MatchOptions {
require_literal_leading_dot: true,
..GLOB_PARAMS
},
)
}
fn arg_glob_opt(
pattern: &Spanned<String>,
cwd: &Path,
options: MatchOptions,
) -> Result<Paths, ShellError> {
// remove ansi coloring (?)
let pattern = {
Spanned {
item: nu_utils::strip_ansi_string_unlikely(pattern.item.clone()),
span: pattern.span,
}
};
// if there's a file with same path as the pattern, just return that.
let pp = cwd.join(&pattern.item);
let md = fs::metadata(pp);
#[allow(clippy::single_match)]
match md {
Ok(_metadata) => {
return Ok(Paths::single(&PathBuf::from(pattern.item), cwd));
}
// file not found, but also "invalid chars in file" (e.g * on Windows). Fall through and glob
Err(_) => {}
}
// user wasn't referring to a specific thing in filesystem, try to glob it.
match glob_with_parent(&pattern.item, options, cwd) {
Ok(p) => Ok(p),
Err(pat_err) => {
Err(ShellError::InvalidGlobPattern(
pat_err.msg.into(),
pattern.span, // improve specificity
))
}
}
}
#[cfg(test)]
mod test {
use super::*;
use nu_glob::GlobResult;
use nu_protocol::{Span, Spanned};
use nu_test_support::fs::Stub::EmptyFile;
use nu_test_support::playground::Playground;
use rstest::rstest;
fn spanned_string(str: &str) -> Spanned<String> {
Spanned {
item: str.to_string(),
span: Span::test_data(),
}
}
#[test]
fn does_something() {
let act = arg_glob(&spanned_string("*"), &PathBuf::from("."));
assert!(act.is_ok());
for f in act.expect("checked ok") {
match f {
Ok(p) => {
assert!(!p.to_str().unwrap().is_empty());
}
Err(e) => panic!("unexpected error {:?}", e),
};
}
}
#[test]
fn glob_format_error() {
let act = arg_glob(&spanned_string(r#"ab]c[def"#), &PathBuf::from("."));
assert!(act.is_err());
}
#[rstest]
#[case("*", 4, "no dirs")]
#[case("**/*", 7, "incl dirs")]
fn glob_subdirs(#[case] pat: &str, #[case] exp_count: usize, #[case] case: &str) {
Playground::setup("glob_subdirs", |dirs, sandbox| {
sandbox.with_files(vec![
EmptyFile("yehuda.txt"),
EmptyFile("jttxt"),
EmptyFile("andres.txt"),
]);
sandbox.mkdir(".children");
sandbox.within(".children").with_files(vec![
EmptyFile("timothy.txt"),
EmptyFile("tiffany.txt"),
EmptyFile("trish.txt"),
]);
let p: Vec<GlobResult> = arg_glob(&spanned_string(pat), &dirs.test)
.expect("no error")
.collect();
assert_eq!(
exp_count,
p.iter().filter(|i| i.is_ok()).count(),
" case: {case} ",
);
// expected behavior -- that directories are included in results (if name matches pattern)
let t = p
.iter()
.any(|i| i.as_ref().unwrap().to_string_lossy().contains(".children"));
assert!(t, "check for dir, case {case}");
})
}
#[rstest]
#[case("yehuda.txt", true, 1, "matches literal path")]
#[case("*", false, 3, "matches glob")]
#[case(r#"bad[glob.foo"#, true, 1, "matches literal, would be bad glob pat")]
fn exact_vs_glob(
#[case] pat: &str,
#[case] exp_matches_input: bool,
#[case] exp_count: usize,
#[case] case: &str,
) {
Playground::setup("exact_vs_glob", |dirs, sandbox| {
sandbox.with_files(vec![
EmptyFile("yehuda.txt"),
EmptyFile("jttxt"),
EmptyFile("bad[glob.foo"),
]);
let res = arg_glob(&spanned_string(pat), &dirs.test)
.expect("no error")
.collect::<Vec<GlobResult>>();
eprintln!("res: {:?}", res);
if exp_matches_input {
assert_eq!(
exp_count,
res.len(),
" case {case}: matches input, but count not 1? "
);
assert_eq!(
&res[0].as_ref().unwrap().to_string_lossy(),
pat, // todo: is it OK for glob to return relative paths (not to current cwd, but to arg cwd of arg_glob)?
);
} else {
assert_eq!(exp_count, res.len(), " case: {}: matched glob", case);
}
})
}
#[rstest]
#[case(r#"realbad[glob.foo"#, true, 1, "error, bad glob")]
fn exact_vs_bad_glob(
// if path doesn't exist but pattern is not valid glob, should get error.
#[case] pat: &str,
#[case] _exp_matches_input: bool,
#[case] _exp_count: usize,
#[case] _tag: &str,
) {
Playground::setup("exact_vs_bad_glob", |dirs, sandbox| {
sandbox.with_files(vec![
EmptyFile("yehuda.txt"),
EmptyFile("jttxt"),
EmptyFile("bad[glob.foo"),
]);
let res = arg_glob(&spanned_string(pat), &dirs.test);
assert!(res
.expect_err("expected error")
.to_string()
.contains("Invalid glob pattern"));
})
}
}

View File

@ -1,4 +1,7 @@
mod arg_glob;
pub mod formats; pub mod formats;
pub mod hook; pub mod hook;
pub mod input_handler; pub mod input_handler;
pub mod util; pub mod util;
pub use arg_glob::arg_glob;
pub use arg_glob::arg_glob_leading_dot;

View File

@ -4,6 +4,7 @@ use std::path::PathBuf;
use std::sync::atomic::AtomicBool; use std::sync::atomic::AtomicBool;
use std::sync::Arc; use std::sync::Arc;
use nu_cmd_base::arg_glob;
use nu_engine::env::current_dir; use nu_engine::env::current_dir;
use nu_engine::CallExt; use nu_engine::CallExt;
use nu_path::{canonicalize_with, expand_path_with}; use nu_path::{canonicalize_with, expand_path_with};
@ -19,13 +20,6 @@ use super::util::try_interaction;
use crate::filesystem::util::FileStructure; use crate::filesystem::util::FileStructure;
use crate::progress_bar; use crate::progress_bar;
const GLOB_PARAMS: nu_glob::MatchOptions = nu_glob::MatchOptions {
case_sensitive: true,
require_literal_separator: false,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
#[derive(Clone)] #[derive(Clone)]
pub struct Cp; pub struct Cp;
@ -81,12 +75,6 @@ impl Command for Cp {
_input: PipelineData, _input: PipelineData,
) -> Result<PipelineData, ShellError> { ) -> Result<PipelineData, ShellError> {
let src: Spanned<String> = call.req(engine_state, stack, 0)?; let src: Spanned<String> = call.req(engine_state, stack, 0)?;
let src = {
Spanned {
item: nu_utils::strip_ansi_string_unlikely(src.item),
span: src.span,
}
};
let dst: Spanned<String> = call.req(engine_state, stack, 1)?; let dst: Spanned<String> = call.req(engine_state, stack, 1)?;
let recursive = call.has_flag("recursive"); let recursive = call.has_flag("recursive");
let verbose = call.has_flag("verbose"); let verbose = call.has_flag("verbose");
@ -95,7 +83,6 @@ impl Command for Cp {
let update_mode = call.has_flag("update"); let update_mode = call.has_flag("update");
let current_dir_path = current_dir(engine_state, stack)?; let current_dir_path = current_dir(engine_state, stack)?;
let source = current_dir_path.join(src.item.as_str());
let destination = current_dir_path.join(dst.item.as_str()); let destination = current_dir_path.join(dst.item.as_str());
let path_last_char = destination.as_os_str().to_string_lossy().chars().last(); let path_last_char = destination.as_os_str().to_string_lossy().chars().last();
@ -110,7 +97,7 @@ impl Command for Cp {
let span = call.head; let span = call.head;
// Get an iterator with all the source files. // Get an iterator with all the source files.
let sources: Vec<_> = match nu_glob::glob_with(&source.to_string_lossy(), GLOB_PARAMS) { let sources: Vec<_> = match arg_glob(&src, &current_dir_path) {
Ok(files) => files.collect(), Ok(files) => files.collect(),
Err(e) => { Err(e) => {
return Err(ShellError::GenericError( return Err(ShellError::GenericError(
@ -124,13 +111,7 @@ impl Command for Cp {
}; };
if sources.is_empty() { if sources.is_empty() {
return Err(ShellError::GenericError( return Err(ShellError::FileNotFound(src.span));
"No matches found".into(),
"no matches found".into(),
Some(src.span),
None,
Vec::new(),
));
} }
if sources.len() > 1 && !destination.is_dir() { if sources.len() > 1 && !destination.is_dir() {
@ -189,9 +170,7 @@ impl Command for Cp {
} }
let res = if src == dst { let res = if src == dst {
let message = format!( let message = format!("src and dst identical: {:?} (not copied)", src);
"src {source:?} and dst {destination:?} are identical(not copied)"
);
return Err(ShellError::GenericError( return Err(ShellError::GenericError(
"Copy aborted".into(), "Copy aborted".into(),

View File

@ -1,6 +1,7 @@
use std::path::{Path, PathBuf}; use std::path::{Path, PathBuf};
use super::util::try_interaction; use super::util::try_interaction;
use nu_cmd_base::arg_glob;
use nu_engine::env::current_dir; use nu_engine::env::current_dir;
use nu_engine::CallExt; use nu_engine::CallExt;
use nu_protocol::ast::Call; use nu_protocol::ast::Call;
@ -10,13 +11,6 @@ use nu_protocol::{
Spanned, SyntaxShape, Type, Value, Spanned, SyntaxShape, Type, Value,
}; };
const GLOB_PARAMS: nu_glob::MatchOptions = nu_glob::MatchOptions {
case_sensitive: true,
require_literal_separator: false,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
#[derive(Clone)] #[derive(Clone)]
pub struct Mv; pub struct Mv;
@ -70,12 +64,6 @@ impl Command for Mv {
) -> Result<PipelineData, ShellError> { ) -> Result<PipelineData, ShellError> {
// TODO: handle invalid directory or insufficient permissions when moving // TODO: handle invalid directory or insufficient permissions when moving
let spanned_source: Spanned<String> = call.req(engine_state, stack, 0)?; let spanned_source: Spanned<String> = call.req(engine_state, stack, 0)?;
let spanned_source = {
Spanned {
item: nu_utils::strip_ansi_string_unlikely(spanned_source.item),
span: spanned_source.span,
}
};
let spanned_destination: Spanned<String> = call.req(engine_state, stack, 1)?; let spanned_destination: Spanned<String> = call.req(engine_state, stack, 1)?;
let verbose = call.has_flag("verbose"); let verbose = call.has_flag("verbose");
let interactive = call.has_flag("interactive"); let interactive = call.has_flag("interactive");
@ -88,17 +76,11 @@ impl Command for Mv {
let source = path.join(spanned_source.item.as_str()); let source = path.join(spanned_source.item.as_str());
let destination = path.join(spanned_destination.item.as_str()); let destination = path.join(spanned_destination.item.as_str());
let mut sources = nu_glob::glob_with(&source.to_string_lossy(), GLOB_PARAMS) let mut sources =
.map_or_else(|_| Vec::new(), Iterator::collect); arg_glob(&spanned_source, &path).map_or_else(|_| Vec::new(), Iterator::collect);
if sources.is_empty() { if sources.is_empty() {
return Err(ShellError::GenericError( return Err(ShellError::FileNotFound(spanned_source.span));
"File(s) not found".into(),
"could not find any files matching this glob pattern".into(),
Some(spanned_source.span),
None,
Vec::new(),
));
} }
// We have two possibilities. // We have two possibilities.

View File

@ -7,6 +7,7 @@ use std::path::PathBuf;
use super::util::try_interaction; use super::util::try_interaction;
use nu_cmd_base::arg_glob_leading_dot;
use nu_engine::env::current_dir; use nu_engine::env::current_dir;
use nu_engine::CallExt; use nu_engine::CallExt;
use nu_protocol::ast::Call; use nu_protocol::ast::Call;
@ -16,13 +17,6 @@ use nu_protocol::{
Spanned, SyntaxShape, Type, Value, Spanned, SyntaxShape, Type, Value,
}; };
const GLOB_PARAMS: nu_glob::MatchOptions = nu_glob::MatchOptions {
case_sensitive: true,
require_literal_separator: false,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
const TRASH_SUPPORTED: bool = cfg!(all( const TRASH_SUPPORTED: bool = cfg!(all(
feature = "trash-support", feature = "trash-support",
not(any(target_os = "android", target_os = "ios")) not(any(target_os = "android", target_os = "ios"))
@ -49,8 +43,8 @@ impl Command for Rm {
.input_output_types(vec![(Type::Nothing, Type::Nothing)]) .input_output_types(vec![(Type::Nothing, Type::Nothing)])
.required( .required(
"filename", "filename",
SyntaxShape::Filepath, SyntaxShape::GlobPattern,
"the path of the file you want to remove", "the file or files you want to remove",
) )
.switch( .switch(
"trash", "trash",
@ -252,13 +246,7 @@ fn rm(
} }
let path = currentdir_path.join(&target.item); let path = currentdir_path.join(&target.item);
match nu_glob::glob_with( match arg_glob_leading_dot(&target, &currentdir_path) {
&path.to_string_lossy(),
nu_glob::MatchOptions {
require_literal_leading_dot: true,
..GLOB_PARAMS
},
) {
Ok(files) => { Ok(files) => {
for file in files { for file in files {
match file { match file {

View File

@ -1,5 +1,6 @@
use nu_engine::CallExt; use nu_cmd_base::arg_glob;
use nu_path::expand_to_real_path; use nu_engine::{current_dir, CallExt};
use nu_glob::GlobResult;
use nu_protocol::{ use nu_protocol::{
ast::Call, ast::Call,
engine::{Command, EngineState, Stack}, engine::{Command, EngineState, Stack},
@ -10,12 +11,6 @@ use uu_cp::{BackupMode, UpdateMode};
// TODO: related to uucore::error::set_exit_code(EXIT_ERR) // TODO: related to uucore::error::set_exit_code(EXIT_ERR)
// const EXIT_ERR: i32 = 1; // const EXIT_ERR: i32 = 1;
const GLOB_PARAMS: nu_glob::MatchOptions = nu_glob::MatchOptions {
case_sensitive: true,
require_literal_separator: false,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
#[cfg(not(target_os = "windows"))] #[cfg(not(target_os = "windows"))]
const PATH_SEPARATOR: &str = "/"; const PATH_SEPARATOR: &str = "/";
@ -154,20 +149,22 @@ impl Command for UCp {
Vec::new(), Vec::new(),
)); ));
}; };
// paths now contains the sources // paths now contains the sources
let sources: Vec<Vec<PathBuf>> = paths
.iter() let cwd = current_dir(engine_state, stack)?;
.map(|p| { let mut sources: Vec<PathBuf> = Vec::new();
// Need to expand too make it work with globbing
let expanded_src = expand_to_real_path(&p.item); for p in paths {
match nu_glob::glob_with(&expanded_src.to_string_lossy(), GLOB_PARAMS) { let exp_files = arg_glob(&p, &cwd)?.collect::<Vec<GlobResult>>();
Ok(files) => { if exp_files.is_empty() {
let f = files.filter_map(Result::ok).collect::<Vec<PathBuf>>();
if f.is_empty() {
return Err(ShellError::FileNotFound(p.span)); return Err(ShellError::FileNotFound(p.span));
} };
let any_source_is_dir = f.iter().any(|f| matches!(f, f if f.is_dir())); let mut app_vals: Vec<PathBuf> = Vec::new();
if any_source_is_dir && !recursive { for v in exp_files {
match v {
Ok(path) => {
if !recursive && path.is_dir() {
return Err(ShellError::GenericError( return Err(ShellError::GenericError(
"could_not_copy_directory".into(), "could_not_copy_directory".into(),
"resolves to a directory (not copied)".into(), "resolves to a directory (not copied)".into(),
@ -175,22 +172,20 @@ impl Command for UCp {
Some("Directories must be copied using \"--recursive\"".into()), Some("Directories must be copied using \"--recursive\"".into()),
Vec::new(), Vec::new(),
)); ));
};
app_vals.push(path)
}
Err(e) => {
return Err(ShellError::ErrorExpandingGlob(
format!("error {} in path {}", e.error(), e.path().display()),
p.span,
));
}
}
}
sources.append(&mut app_vals);
} }
Ok(f)
}
Err(e) => Err(ShellError::GenericError(
e.to_string(),
"invalid pattern".to_string(),
Some(p.span),
None,
Vec::new(),
)),
}
})
.collect::<Result<Vec<Vec<PathBuf>>, ShellError>>()?;
let sources = sources.into_iter().flatten().collect::<Vec<PathBuf>>();
let options = uu_cp::Options { let options = uu_cp::Options {
overwrite, overwrite,
reflink_mode, reflink_mode,

View File

@ -1,28 +1,21 @@
use crate::{DirBuilder, DirInfo, FileInfo}; use crate::{DirBuilder, DirInfo, FileInfo};
use nu_engine::CallExt; use nu_cmd_base::arg_glob;
use nu_glob::{GlobError, MatchOptions, Pattern}; use nu_engine::{current_dir, CallExt};
use nu_glob::{GlobError, Pattern};
use nu_protocol::{ use nu_protocol::{
ast::Call, ast::Call,
engine::{Command, EngineState, Stack}, engine::{Command, EngineState, Stack},
Category, Example, IntoInterruptiblePipelineData, PipelineData, ShellError, Signature, Spanned, Category, Example, IntoInterruptiblePipelineData, PipelineData, ShellError, Signature, Span,
SyntaxShape, Type, Value, Spanned, SyntaxShape, Type, Value,
}; };
use serde::Deserialize; use serde::Deserialize;
use std::path::PathBuf;
const GLOB_PARAMS: MatchOptions = MatchOptions {
case_sensitive: true,
require_literal_separator: true,
require_literal_leading_dot: false,
recursive_match_hidden_dir: true,
};
#[derive(Clone)] #[derive(Clone)]
pub struct Du; pub struct Du;
#[derive(Deserialize, Clone, Debug)] #[derive(Deserialize, Clone, Debug)]
pub struct DuArgs { pub struct DuArgs {
path: Option<Spanned<PathBuf>>, path: Option<Spanned<String>>,
all: bool, all: bool,
deref: bool, deref: bool,
exclude: Option<Spanned<String>>, exclude: Option<Spanned<String>>,
@ -97,6 +90,8 @@ impl Command for Du {
return Err(ShellError::NeedsPositiveValue(min_size.span)); return Err(ShellError::NeedsPositiveValue(min_size.span));
} }
} }
let current_dir = current_dir(engine_state, stack)?;
let args = DuArgs { let args = DuArgs {
path: call.opt(engine_state, stack, 0)?, path: call.opt(engine_state, stack, 0)?,
all: call.has_flag("all"), all: call.has_flag("all"),
@ -107,38 +102,22 @@ impl Command for Du {
}; };
let exclude = args.exclude.map_or(Ok(None), move |x| { let exclude = args.exclude.map_or(Ok(None), move |x| {
Pattern::new(&x.item).map(Some).map_err(|e| { Pattern::new(&x.item)
ShellError::GenericError( .map(Some)
"glob error".to_string(), .map_err(|e| ShellError::InvalidGlobPattern(e.msg.to_string(), x.span))
e.msg.to_string(),
Some(x.span),
None,
Vec::new(),
)
})
})?; })?;
let include_files = args.all; let include_files = args.all;
let mut paths = match args.path { let mut paths = match args.path {
Some(p) => { Some(p) => arg_glob(&p, &current_dir)?,
let item = p.item.to_str().expect("Why isn't this encoded properly?");
match nu_glob::glob_with(item, GLOB_PARAMS) {
// Convert the PatternError to a ShellError, preserving the span
// of the inputted glob.
Err(e) => {
return Err(ShellError::GenericError(
"glob error".to_string(),
e.msg.to_string(),
Some(p.span),
None,
Vec::new(),
))
}
Ok(path) => path,
}
}
// The * pattern should never fail. // The * pattern should never fail.
None => nu_glob::glob_with("*", GLOB_PARAMS).expect("du: * pattern failed to glob"), None => arg_glob(
&Spanned {
item: "*".into(),
span: Span::unknown(),
},
&current_dir,
)?,
} }
.filter(move |p| { .filter(move |p| {
if include_files { if include_files {

View File

@ -203,7 +203,7 @@ fn errors_if_source_doesnt_exist() {
cwd: dirs.test(), cwd: dirs.test(),
"mv non-existing-file test_folder/" "mv non-existing-file test_folder/"
); );
assert!(actual.err.contains("File(s) not found")); assert!(actual.err.contains("file not found"));
}) })
} }

View File

@ -6,6 +6,7 @@ use nu_test_support::fs::{
use nu_test_support::nu; use nu_test_support::nu;
use nu_test_support::playground::Playground; use nu_test_support::playground::Playground;
use rstest::rstest;
use std::path::Path; use std::path::Path;
#[cfg(not(target_os = "windows"))] #[cfg(not(target_os = "windows"))]
@ -968,3 +969,46 @@ fn test_cp_with_vars() {
assert!(dirs.test().join("target").exists()); assert!(dirs.test().join("target").exists());
}); });
} }
#[rstest]
#[case(r#"'a]c'"#)]
#[case(r#"'a[c'"#)]
#[case(r#"'a[bc]d'"#)]
#[case(r#"'a][c'"#)]
fn copies_files_with_glob_metachars(#[case] src_name: &str) {
Playground::setup("ucp_test_34", |dirs, sandbox| {
sandbox.with_files(vec![FileWithContent(
src_name,
"What is the sound of one hand clapping?",
)]);
let src = dirs.test().join(src_name);
// -- open command doesn't like file name
//// Get the hash of the file content to check integrity after copy.
//let src_hash = get_file_hash(src.display());
let actual = nu!(
cwd: dirs.test(),
"cp {} {}",
src.display(),
TEST_HELLO_WORLD_DEST
);
assert!(actual.err.is_empty());
assert!(dirs.test().join(TEST_HELLO_WORLD_DEST).exists());
//// Get the hash of the copied file content to check against first_hash.
//let after_cp_hash = get_file_hash(dirs.test().join(TEST_HELLO_WORLD_DEST).display());
//assert_eq!(src_hash, after_cp_hash);
});
}
#[cfg(not(windows))]
#[rstest]
#[case(r#"'a]?c'"#)]
#[case(r#"'a*.?c'"#)]
// windows doesn't allow filename with `*`.
fn copies_files_with_glob_metachars_nw(#[case] src_name: &str) {
copies_files_with_glob_metachars(src_name);
}

View File

@ -103,6 +103,19 @@ pub struct Paths {
scope: Option<PathBuf>, scope: Option<PathBuf>,
} }
impl Paths {
/// An iterator representing a single path.
pub fn single(path: &Path, relative_to: &Path) -> Self {
Paths {
dir_patterns: vec![Pattern::new("*").expect("hard coded pattern")],
require_dir: false,
options: MatchOptions::default(),
todo: vec![Ok((path.to_path_buf(), 0))],
scope: Some(relative_to.into()),
}
}
}
/// Return an iterator that produces all the `Path`s that match the given /// Return an iterator that produces all the `Path`s that match the given
/// pattern using default match options, which may be absolute or relative to /// pattern using default match options, which may be absolute or relative to
/// the current working directory. /// the current working directory.
@ -110,7 +123,7 @@ pub struct Paths {
/// This may return an error if the pattern is invalid. /// This may return an error if the pattern is invalid.
/// ///
/// This method uses the default match options and is equivalent to calling /// This method uses the default match options and is equivalent to calling
/// `glob_with(pattern, MatchOptions::new())`. Use `glob_with` directly if you /// `glob_with(pattern, MatchOptions::default())`. Use `glob_with` directly if you
/// want to use non-default match options. /// want to use non-default match options.
/// ///
/// When iterating, each result is a `GlobResult` which expresses the /// When iterating, each result is a `GlobResult` which expresses the
@ -273,6 +286,32 @@ pub fn glob_with(pattern: &str, options: MatchOptions) -> Result<Paths, PatternE
}) })
} }
/// Return an iterator that produces all the `Path`s that match the given
/// pattern relative to a specified parent directory and using specified match options.
/// Paths may be absolute or relative to the current working directory.
///
/// This is provided primarily for testability, so multithreaded test runners can
/// test pattern matches in different test directories at the same time without
/// having to append the parent to the pattern under test.
pub fn glob_with_parent(
pattern: &str,
options: MatchOptions,
parent: &Path,
) -> Result<Paths, PatternError> {
match glob_with(pattern, options) {
Ok(mut p) => {
p.scope = match p.scope {
None => Some(parent.to_path_buf()),
Some(s) if &s.to_string_lossy() == "." => Some(parent.to_path_buf()),
Some(s) => Some(s),
};
Ok(p)
}
Err(e) => Err(e),
}
}
/// A glob iteration error. /// A glob iteration error.
/// ///
/// This is typically returned when a particular path cannot be read /// This is typically returned when a particular path cannot be read
@ -347,9 +386,12 @@ impl Iterator for Paths {
// Shouldn't happen, but we're using -1 as a special index. // Shouldn't happen, but we're using -1 as a special index.
assert!(self.dir_patterns.len() < !0); assert!(self.dir_patterns.len() < !0);
// if there's one prefilled result, take it, otherwise fill the todo buffer
if self.todo.len() != 1 {
fill_todo(&mut self.todo, &self.dir_patterns, 0, &scope, self.options); fill_todo(&mut self.todo, &self.dir_patterns, 0, &scope, self.options);
} }
} }
}
loop { loop {
if self.dir_patterns.is_empty() || self.todo.is_empty() { if self.dir_patterns.is_empty() || self.todo.is_empty() {
@ -704,7 +746,7 @@ impl Pattern {
} }
/// Return if the given `str` matches this `Pattern` using the default /// Return if the given `str` matches this `Pattern` using the default
/// match options (i.e. `MatchOptions::new()`). /// match options (i.e. `MatchOptions::default()`).
/// ///
/// # Examples /// # Examples
/// ///
@ -720,7 +762,7 @@ impl Pattern {
} }
/// Return if the given `Path`, when converted to a `str`, matches this /// Return if the given `Path`, when converted to a `str`, matches this
/// `Pattern` using the default match options (i.e. `MatchOptions::new()`). /// `Pattern` using the default match options (i.e. `MatchOptions::default()`).
pub fn matches_path(&self, path: &Path) -> bool { pub fn matches_path(&self, path: &Path) -> bool {
// FIXME (#9639): This needs to handle non-utf8 paths // FIXME (#9639): This needs to handle non-utf8 paths
path.to_str().map_or(false, |s| self.matches(s)) path.to_str().map_or(false, |s| self.matches(s))

View File

@ -1128,6 +1128,31 @@ This is an internal Nushell error, please file an issue https://github.com/nushe
help("Help messages are currently not supported to be constants.") help("Help messages are currently not supported to be constants.")
)] )]
NotAConstHelp(#[label = "Cannot get help message at parse time."] Span), NotAConstHelp(#[label = "Cannot get help message at parse time."] Span),
/// Invalid glob pattern
///
/// ## Resolution
///
/// Correct glob pattern
#[error("Invalid glob pattern")]
#[diagnostic(
code(nu::shell::invalid_glob_pattern),
help("Refer to xxx for help on nushell glob patterns.")
)]
InvalidGlobPattern(String, #[label = "{0}"] Span),
/// Error expanding glob pattern
///
/// ## Resolution
///
/// Correct glob pattern or file access issue
#[error("Error expanding glob pattern")]
#[diagnostic(
code(nu::shell::error_expanding_glob),
help("Correct glob pattern or file access issue")
)]
//todo: add error detail
ErrorExpandingGlob(String, #[label = "{0}"] Span),
} }
// TODO: Implement as From trait // TODO: Implement as From trait