fish-shell/src/expand.rs

1582 lines
62 KiB
Rust
Raw Normal View History

//! String expansion functions. These functions perform several kinds of parameter expansion. There
//! are a lot of issues with regards to memory allocation. Overall, these functions would benefit
//! from using a more clever memory allocation scheme, perhaps an evil combination of talloc,
//! string buffers and reference counting.
use crate::builtins::shared::{
STATUS_CMD_ERROR, STATUS_CMD_UNKNOWN, STATUS_EXPAND_ERROR, STATUS_ILLEGAL_CMD,
STATUS_INVALID_ARGS, STATUS_NOT_EXECUTABLE, STATUS_READ_TOO_MUCH, STATUS_UNMATCHED_WILDCARD,
};
use crate::common::{
char_offset, charptr2wcstring, escape, escape_string, escape_string_for_double_quotes,
unescape_string, valid_var_name_char, wcs2zstring, EscapeFlags, EscapeStringStyle,
UnescapeFlags, UnescapeStringStyle, EXPAND_RESERVED_BASE, EXPAND_RESERVED_END,
};
use crate::complete::{CompleteFlags, Completion, CompletionList, CompletionReceiver};
use crate::env::{EnvVar, Environment};
use crate::exec::exec_subshell_for_expand;
use crate::future_feature_flags::{feature_test, FeatureFlag};
use crate::history::{history_session_id, History};
use crate::operation_context::OperationContext;
use crate::parse_constants::{ParseError, ParseErrorCode, ParseErrorList, SOURCE_LOCATION_UNKNOWN};
use crate::parse_util::{
parse_util_expand_variable_error, parse_util_locate_cmdsubst_range, MaybeParentheses,
};
use crate::path::path_apply_working_directory;
use crate::util::wcsfilecmp_glob;
2023-08-09 06:16:04 +08:00
use crate::wchar::prelude::*;
use crate::wcstringutil::{join_strings, trim};
use crate::wildcard::{wildcard_expand_string, wildcard_has_internal};
use crate::wildcard::{WildcardResult, ANY_CHAR, ANY_STRING, ANY_STRING_RECURSIVE};
use crate::wutil::{normalize_path, wcstoi_partial, Options};
use bitflags::bitflags;
2023-03-11 10:47:41 +08:00
bitflags! {
/// Set of flags controlling expansions.
#[derive(Copy, Clone, Default)]
pub struct ExpandFlags : u16 {
/// Fail expansion if there is a command substitution.
const FAIL_ON_CMDSUBST = 1 << 0;
builtin commandline: -x for expanded tokens, supplanting -o Issue #10194 reports Cobra completions do set -l args (commandline -opc) eval $args[1] __complete $args[2..] (commandline -ct | string escape) The intent behind "eval" is to expand variables and tildes in "$args". Fair enough. Several of our own completions do the same, see the next commit. The problem with "commandline -o" + "eval" is that the former already removes quotes that are relevant for "eval". This becomes a problem if $args contains quoted () or {}, for example this command will wrongly execute a command substituion: git --work-tree='(launch-missiles)' <TAB> It is possible to escape the string the tokens before running eval, but then there will be no expansion of variables etc. The problem is that "commandline -o" only unescapes tokens so they end up in a weird state somewhere in-between what the user typed and the expanded version. Remove the need for "eval" by introducing "commandline -x" which expands things like variables and braces. This enables custom completion scripts to be aware of shell variables without eval, see the added test for completions to "make -C $var/some/dir ". This means that essentially all third party scripts should migrate from "commandline -o" to "commandline -x". For example set -l tokens if commandline -x >/dev/null 2>&1 set tokens (commandline -xpc) else set tokens (commandline -opc) end Since this is mainly used for completions, the expansion skips command substitutions. They are passed through as-is (instead of cancelling or expanding to nothing) to make custom completion scripts work reasonably well in the common case. Of course there are cases where we would want to expand command substitutions here, so I'm not sure.
2024-01-06 15:45:33 +08:00
/// Skip command substitutions.
const SKIP_CMDSUBST = 1 << 14;
/// Skip variable expansion.
const SKIP_VARIABLES = 1 << 1;
/// Skip wildcard expansion.
const SKIP_WILDCARDS = 1 << 2;
/// The expansion is being done for tab or auto completions. Returned completions may have the
/// wildcard as a prefix instead of a match.
const FOR_COMPLETIONS = 1 << 3;
/// Only match files that are executable by the current user.
const EXECUTABLES_ONLY = 1 << 4;
/// Only match directories.
const DIRECTORIES_ONLY = 1 << 5;
/// Generate descriptions, stored in the description field of completions.
const GEN_DESCRIPTIONS = 1 << 6;
/// Un-expand home directories to tildes after.
const PRESERVE_HOME_TILDES = 1 << 7;
/// Allow fuzzy matching.
const FUZZY_MATCH = 1 << 8;
/// Disallow directory abbreviations like /u/l/b for /usr/local/bin. Only applicable if
/// fuzzy_match is set.
const NO_FUZZY_DIRECTORIES = 1 << 9;
/// Allows matching a leading dot even if the wildcard does not contain one.
/// By default, wildcards only match a leading dot literally; this is why e.g. '*' does not
/// match hidden files.
const ALLOW_NONLITERAL_LEADING_DOT = 1 << 10;
/// Do expansions specifically to support cd. This means using CDPATH as a list of potential
/// working directories, and to use logical instead of physical paths.
const SPECIAL_FOR_CD = 1 << 11;
/// Do expansions specifically for cd autosuggestion. This is to differentiate between cd
/// completions and cd autosuggestions.
const SPECIAL_FOR_CD_AUTOSUGGESTION = 1 << 12;
/// Do expansions specifically to support external command completions. This means using PATH as
/// a list of potential working directories.
const SPECIAL_FOR_COMMAND = 1 << 13;
}
}
/// Character representing a home directory.
pub const HOME_DIRECTORY: char = char_offset(EXPAND_RESERVED_BASE, 0);
/// Character representing process expansion for %self.
pub const PROCESS_EXPAND_SELF: char = char_offset(EXPAND_RESERVED_BASE, 1);
/// Character representing variable expansion.
pub const VARIABLE_EXPAND: char = char_offset(EXPAND_RESERVED_BASE, 2);
/// Character representing variable expansion into a single element.
pub const VARIABLE_EXPAND_SINGLE: char = char_offset(EXPAND_RESERVED_BASE, 3);
/// Character representing the start of a bracket expansion.
pub const BRACE_BEGIN: char = char_offset(EXPAND_RESERVED_BASE, 4);
/// Character representing the end of a bracket expansion.
pub const BRACE_END: char = char_offset(EXPAND_RESERVED_BASE, 5);
/// Character representing separation between two bracket elements.
pub const BRACE_SEP: char = char_offset(EXPAND_RESERVED_BASE, 6);
/// Character that takes the place of any whitespace within non-quoted text in braces
pub const BRACE_SPACE: char = char_offset(EXPAND_RESERVED_BASE, 7);
/// Separate subtokens in a token with this character.
pub const INTERNAL_SEPARATOR: char = char_offset(EXPAND_RESERVED_BASE, 8);
/// Character representing an empty variable expansion. Only used transitively while expanding
/// variables.
pub const VARIABLE_EXPAND_EMPTY: char = char_offset(EXPAND_RESERVED_BASE, 9);
2023-03-11 10:47:41 +08:00
const _: () = assert!(
EXPAND_RESERVED_END as u32 > VARIABLE_EXPAND_EMPTY as u32,
2023-03-11 10:47:41 +08:00
"Characters used in expansions must stay within private use area"
);
impl ExpandResult {
pub fn new(result: ExpandResultCode) -> Self {
Self { result, status: 0 }
}
pub fn ok() -> Self {
Self::new(ExpandResultCode::ok)
}
/// Make an error value with the given status.
pub fn make_error(status: libc::c_int) -> Self {
assert!(status != 0, "status cannot be 0 for an error result");
Self {
result: ExpandResultCode::error,
status,
}
}
}
impl PartialEq<ExpandResultCode> for ExpandResult {
fn eq(&self, other: &ExpandResultCode) -> bool {
self.result == *other
}
}
/// The string represented by PROCESS_EXPAND_SELF
pub const PROCESS_EXPAND_SELF_STR: &wstr = L!("%self");
/// Perform various forms of expansion on in, such as tilde expansion (\~USER becomes the users home
/// directory), variable expansion (\$VAR_NAME becomes the value of the environment variable
/// VAR_NAME), cmdsubst expansion and wildcard expansion. The results are inserted into the list
/// out.
///
/// If the parameter does not need expansion, it is copied into the list out.
///
/// \param input The parameter to expand
/// \param output The list to which the result will be appended.
/// \param ctx The parser, variables, and cancellation checker for this operation. The parser may
/// be null. \param errors Resulting errors, or nullptr to ignore
///
/// Return An expand_result_t.
/// wildcard_no_match and wildcard_match are normal exit conditions used only on
/// strings containing wildcards to tell if the wildcard produced any matches.
pub fn expand_string(
input: WString,
out_completions: &mut CompletionList,
flags: ExpandFlags,
ctx: &OperationContext,
errors: Option<&mut ParseErrorList>,
) -> ExpandResult {
let mut completions = vec![];
std::mem::swap(&mut completions, out_completions);
let mut recv = CompletionReceiver::from_list(completions, ctx.expansion_limit);
let result = expand_to_receiver(input, &mut recv, flags, ctx, errors);
*out_completions = recv.take();
result
}
/// Variant of string that inserts its results into a completion_receiver_t.
pub fn expand_to_receiver(
input: WString,
out_completions: &mut CompletionReceiver,
flags: ExpandFlags,
ctx: &OperationContext,
errors: Option<&mut ParseErrorList>,
) -> ExpandResult {
Expander::expand_string(input, out_completions, flags, ctx, errors)
}
/// expand_one is identical to expand_string, except it will fail if in expands to more than one
/// string. This is used for expanding command names.
///
/// \param inout_str The parameter to expand in-place
/// \param ctx The parser, variables, and cancellation checker for this operation. The parser may be
/// null.
/// \param errors Resulting errors, or nullptr to ignore
///
/// Return Whether expansion succeeded.
pub fn expand_one(
s: &mut WString,
flags: ExpandFlags,
ctx: &OperationContext,
errors: Option<&mut ParseErrorList>,
) -> bool {
let mut completions = CompletionList::new();
if !flags.contains(ExpandFlags::FOR_COMPLETIONS) && expand_is_clean(s) {
return true;
}
let mut tmp = WString::new();
std::mem::swap(s, &mut tmp);
if expand_string(tmp, &mut completions, flags, ctx, errors) == ExpandResultCode::ok
&& completions.len() == 1
{
std::mem::swap(s, &mut completions[0].completion);
return true;
}
false
}
/// Expand a command string like $HOME/bin/cmd into a command and list of arguments.
/// Return the command and arguments by reference.
/// If the expansion resulted in no or an empty command, the command will be an empty string. Note
/// that API does not distinguish between expansion resulting in an empty command (''), and
/// expansion resulting in no command (e.g. unset variable).
/// If `skip_wildcards` is true, then do not do wildcard expansion
/// Return an expand error.
pub fn expand_to_command_and_args(
instr: &wstr,
ctx: &OperationContext<'_>,
out_cmd: &mut WString,
mut out_args: Option<&mut Vec<WString>>,
errors: Option<&mut ParseErrorList>,
skip_wildcards: bool,
) -> ExpandResult {
// Fast path.
if expand_is_clean(instr) {
*out_cmd = instr.to_owned();
return ExpandResult::ok();
}
let mut eflags = ExpandFlags::FAIL_ON_CMDSUBST;
if skip_wildcards {
eflags |= ExpandFlags::SKIP_WILDCARDS;
}
let mut completions = CompletionList::new();
let expand_err = expand_string(instr.to_owned(), &mut completions, eflags, ctx, errors);
if expand_err == ExpandResultCode::ok {
// The first completion is the command, any remaining are arguments.
let mut completions = completions.into_iter();
if let Some(comp) = completions.next() {
*out_cmd = comp.completion;
}
if let Some(ref mut out_args) = out_args {
for comp in completions {
out_args.push(comp.completion);
}
}
}
expand_err
}
/// Convert the variable value to a human readable form, i.e. escape things, handle arrays, etc.
/// Suitable for pretty-printing.
pub fn expand_escape_variable(var: &EnvVar) -> WString {
let mut buff = WString::new();
let lst = var.as_list();
for el in lst {
if !buff.is_empty() {
buff.push_str(" ");
}
// We want to use quotes if we have more than one string, or the string contains a space.
let prefer_quotes = lst.len() > 1 || el.contains(' ');
if prefer_quotes && is_quotable(el) {
buff.push('\'');
buff.push_utfstr(el);
buff.push('\'');
} else {
buff.push_utfstr(&escape(el));
}
}
buff
}
/// Convert a string value to a human readable form, i.e. escape things, handle arrays, etc.
/// Suitable for pretty-printing.
pub fn expand_escape_string(el: &wstr) -> WString {
let mut buff = WString::new();
let prefer_quotes = el.contains(' ');
if prefer_quotes && is_quotable(el) {
buff.push('\'');
buff.push_utfstr(el);
buff.push('\'');
} else {
buff.push_utfstr(&escape(el));
}
buff
}
2023-04-09 20:33:20 +08:00
/// Perform tilde expansion and nothing else on the specified string, which is modified in place.
///
/// \param input the string to tilde expand
pub fn expand_tilde(input: &mut WString, vars: &dyn Environment) {
2023-04-09 20:33:20 +08:00
if input.chars().next() == Some('~') {
input.replace_range(0..1, wstr::from_char_slice(&[HOME_DIRECTORY]));
expand_home_directory(input, vars);
}
}
/// Perform the opposite of tilde expansion on the string, which is modified in place.
pub fn replace_home_directory_with_tilde(s: &wstr, vars: &dyn Environment) -> WString {
let mut result = s.to_owned();
// Only absolute paths get this treatment.
if result.starts_with(L!("/")) {
let mut home_directory = L!("~").to_owned();
expand_tilde(&mut home_directory, vars);
// If we can't get a home directory, don't replace anything.
// This is the case e.g. with --no-execute
if home_directory.is_empty() {
return result;
}
if !home_directory.ends_with(L!("/")) {
home_directory.push('/');
}
// Now check if the home_directory prefixes the string.
if result.starts_with(&home_directory) {
// Success
result.replace_range(0..home_directory.len(), L!("~/"));
}
}
result
}
/// Characters which make a string unclean if they are the first character of the string. See \c
/// expand_is_clean().
const UNCLEAN_FIRST: &wstr = L!("~%");
/// Unclean characters. See \c expand_is_clean().
const UNCLEAN: &wstr = L!("$*?\\\"'({})");
/// Test if the specified argument is clean, i.e. it does not contain any tokens which need to be
/// expanded or otherwise altered. Clean strings can be passed through expand_string and expand_one
/// without changing them. About two thirds of all strings are clean, so skipping expansion on them
/// actually does save a small amount of time, since it avoids multiple memory allocations during
/// the expansion process.
///
/// \param in the string to test
fn expand_is_clean(input: &wstr) -> bool {
if input.is_empty() {
return true;
}
// Test characters that have a special meaning in the first character position.
if UNCLEAN_FIRST.contains(input.as_char_slice()[0]) {
return false;
}
// Test characters that have a special meaning in any character position.
!input.chars().any(|c| UNCLEAN.contains(c))
}
/// Append a syntax error to the given error list.
macro_rules! append_syntax_error {
(
$errors:expr, $source_start:expr,
$fmt:expr $(, $arg:expr )* $(,)?
) => {
if let Some(ref mut errors) = $errors {
let mut error = ParseError::default();
error.source_start = $source_start;
error.source_length = 0;
error.code = ParseErrorCode::syntax;
error.text = wgettext_maybe_fmt!($fmt $(, $arg)*);
errors.push(error);
}
}
}
/// Append a cmdsub error to the given error list. But only do so if the error hasn't already been
/// recorded. This is needed because command substitution is a recursive process and some errors
/// could consequently be recorded more than once.
macro_rules! append_cmdsub_error {
(
$errors:expr, $source_start:expr, $source_end:expr,
$fmt:expr $(, $arg:expr )* $(,)?
) => {
append_cmdsub_error_formatted!(
$errors, $source_start, $source_end,
wgettext_maybe_fmt!($fmt $(, $arg)*));
}
}
macro_rules! append_cmdsub_error_formatted {
(
$errors:expr, $source_start:expr, $source_end:expr,
$text:expr $(,)?
) => {
if let Some(ref mut errors) = $errors {
let mut error = ParseError::default();
error.source_start = $source_start;
error.source_length = $source_end - $source_start + 1;
error.code = ParseErrorCode::cmdsubst;
error.text = $text;
if !errors.iter().any(|e| e.text == error.text) {
errors.push(error);
}
}
};
}
/// Append an overflow error, when expansion produces too much data.
fn append_overflow_error(
errors: &mut Option<&mut ParseErrorList>,
source_start: Option<usize>,
) -> ExpandResult {
if let Some(ref mut errors) = errors {
let mut error = ParseError::default();
error.source_start = source_start.unwrap_or(SOURCE_LOCATION_UNKNOWN);
error.source_length = 0;
error.code = ParseErrorCode::generic;
error.text = wgettext!("Expansion produced too many results").to_owned();
errors.push(error);
}
ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap())
}
/// Test if the specified string does not contain character which can not be used inside a quoted
/// string.
fn is_quotable(s: &wstr) -> bool {
!s.chars().any(|c| "\n\t\r\x08\x1B".contains(c))
}
enum ParseSliceError {
zero_index,
invalid_index,
}
/// Parse an array slicing specification Returns 0 on success. If a parse error occurs, returns the
/// index of the bad token. Note that 0 can never be a bad index because the string always starts
/// with [.
fn parse_slice(
input: &wstr,
idx: &mut Vec<i64>,
array_size: usize,
) -> Result<usize, (usize, ParseSliceError)> {
let size = i64::try_from(array_size).unwrap();
let mut pos = 1; // skip past the opening square bracket
loop {
while input.char_at(pos).is_whitespace() || input.char_at(pos) == INTERNAL_SEPARATOR {
pos += 1;
}
if input.char_at(pos) == ']' {
pos += 1;
break;
}
let tmp = if idx.is_empty() && input.char_at(pos) == '.' && input.char_at(pos + 1) == '.' {
// If we are at the first index expression, a missing start-index means the range starts
// at the first item.
1 // first index
} else {
let mut consumed = 0;
match wcstoi_partial(&input[pos..], Options::default(), &mut consumed) {
Ok(tmp) => {
if tmp == 0 {
// Explicitly refuse $foo[0] as valid syntax, regardless of whether or
// not we're going to show an error if the index ultimately evaluates
// to zero. This will help newcomers to fish avoid a common off-by-one
// error. See #4862.
return Err((pos, ParseSliceError::zero_index));
}
pos += consumed;
// Skip trailing whitespace.
pos += input[pos..]
.chars()
.take_while(|c| c.is_whitespace())
.count();
tmp
}
Err(_error) => {
// We don't test `*end` as is typically done because we expect it to not
// be the null char. Ignore the case of errno==-1 because it means the end
// char wasn't the null char.
return Err((pos, ParseSliceError::invalid_index));
}
}
};
let mut i1 = if tmp > -1 { tmp } else { size + tmp + 1 };
while input.char_at(pos) == INTERNAL_SEPARATOR {
pos += 1;
}
if input.char_at(pos) == '.' && input.char_at(pos + 1) == '.' {
pos += 2;
while input.char_at(pos) == INTERNAL_SEPARATOR {
pos += 1;
}
while input.char_at(pos).is_whitespace() {
pos += 1; // Allow the space in "[.. ]".
}
// If we are at the last index range expression then a missing end-index means the
// range spans until the last item.
let tmp1 = if input.char_at(pos) == ']' {
-1 // last index
} else {
let mut consumed = 0;
match wcstoi_partial(&input[pos..], Options::default(), &mut consumed) {
Ok(tmp) => {
if tmp == 0 {
return Err((pos, ParseSliceError::zero_index));
}
pos += consumed;
// Skip trailing whitespace.
pos += input[pos..]
.chars()
.take_while(|c| c.is_whitespace())
.count();
tmp
}
Err(_error) => {
return Err((pos, ParseSliceError::invalid_index));
}
}
};
let mut i2 = if tmp1 > -1 { tmp1 } else { size + tmp1 + 1 };
// Skip sequences that are entirely outside.
// This means "17..18" expands to nothing if there are less than 17 elements.
if i1 > size && i2 > size {
continue;
}
let mut direction = if i2 < i1 { -1 } else { 1 };
// If only the beginning is negative, always go reverse.
// If only the end, always go forward.
// Prevents `[x..-1]` from going reverse if less than x elements are there.
if (tmp1 > -1) != (tmp > -1) {
direction = if tmp1 > -1 { -1 } else { 1 };
} else {
// Clamp to array size when not forcing direction
// - otherwise "2..-1" clamps both to 1 and then becomes "1..1".
i1 = i1.min(size);
i2 = i2.min(size);
}
let mut jjj = i1;
while jjj * direction <= i2 * direction {
idx.push(jjj);
jjj += direction;
}
continue;
}
idx.push(i1);
}
Ok(pos)
}
/// Expand all environment variables in the string *ptr.
///
/// This function is slow, fragile and complicated. There are lots of little corner cases, like
/// $$foo should do a double expansion, $foo$bar should not double expand bar, etc.
///
/// This function operates on strings backwards, starting at last_idx.
///
/// Note: last_idx is considered to be where it previously finished processing. This means it
/// actually starts operating on last_idx-1. As such, to process a string fully, pass string.size()
/// as last_idx instead of string.size()-1.
///
/// Return the result of expansion.
fn expand_variables(
instr: WString,
out: &mut CompletionReceiver,
last_idx: usize,
vars: &dyn Environment,
errors: &mut Option<&mut ParseErrorList>,
) -> ExpandResult {
// last_idx may be 1 past the end of the string, but no further.
assert!(last_idx <= instr.len(), "Invalid last_idx");
if last_idx == 0 {
if !out.add(instr) {
return append_overflow_error(errors, None);
}
return ExpandResult::ok();
}
// Locate the last VARIABLE_EXPAND or VARIABLE_EXPAND_SINGLE
let mut is_single = false;
let mut varexp_char_idx = last_idx;
loop {
let done = varexp_char_idx == 0;
varexp_char_idx = varexp_char_idx.wrapping_sub(1);
if done {
break;
}
let c = instr.as_char_slice()[varexp_char_idx];
if [VARIABLE_EXPAND, VARIABLE_EXPAND_SINGLE].contains(&c) {
is_single = c == VARIABLE_EXPAND_SINGLE;
break;
}
}
if varexp_char_idx == usize::MAX {
// No variable expand char, we're done.
if !out.add(instr) {
return append_overflow_error(errors, None);
}
return ExpandResult::ok();
}
// Get the variable name.
let var_name_start = varexp_char_idx + 1;
let mut var_name_stop = var_name_start;
while var_name_stop < instr.len() {
let nc = instr.as_char_slice()[var_name_stop];
if nc == VARIABLE_EXPAND_EMPTY {
var_name_stop += 1;
break;
}
if !valid_var_name_char(nc) {
break;
}
var_name_stop += 1;
}
assert!(
var_name_stop >= var_name_start,
"Bogus variable name indexes"
);
// Get the variable name as a string, then try to get the variable from env.
let var_name = &instr[var_name_start..var_name_stop];
// It's an error if the name is empty.
if var_name.is_empty() {
if let Some(ref mut errors) = errors {
parse_util_expand_variable_error(
&instr,
0, /* global_token_pos */
varexp_char_idx,
errors,
);
}
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
2023-04-09 20:33:20 +08:00
}
// Do a dirty hack to make sliced history fast (#4650). We expand from either a variable, or a
// history_t. Note that "history" is read only in env.cpp so it's safe to special-case it in
// this way (it cannot be shadowed, etc).
let mut history = None;
let mut var = None;
if var_name == "history" {
history = Some(History::with_name(&history_session_id(vars)));
} else if var_name.as_char_slice() != [VARIABLE_EXPAND_EMPTY] {
var = vars.get(var_name);
}
// Parse out any following slice.
// Record the end of the variable name and any following slice.
let mut var_name_and_slice_stop = var_name_stop;
let mut all_values = true;
let slice_start = var_name_stop;
let mut var_idx_list = vec![];
if instr.as_char_slice().get(slice_start) == Some(&'[') {
all_values = false;
// If a variable is missing, behave as though we have one value, so that $var[1] always
// works.
let mut effective_val_count = 1;
if let Some(ref var) = var {
effective_val_count = var.as_list().len();
} else if let Some(ref history) = history {
effective_val_count = history.size();
}
match parse_slice(
&instr[slice_start..],
&mut var_idx_list,
effective_val_count,
) {
Ok(offset) => {
var_name_and_slice_stop = slice_start + offset;
}
Err((bad_pos, error)) => {
match error {
ParseSliceError::zero_index => {
append_syntax_error!(
errors,
slice_start + bad_pos,
"array indices start at 1, not 0."
);
}
ParseSliceError::invalid_index => {
append_syntax_error!(errors, slice_start + bad_pos, "Invalid index value");
}
}
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
}
}
let var_idx_list = var_idx_list.iter().filter_map(|&n| n.try_into().ok());
if var.is_none() && history.is_none() {
// Expanding a non-existent variable.
if !is_single {
// Normal expansions of missing variables successfully expand to nothing.
return ExpandResult::ok();
} else {
// Expansion to single argument.
// Replace the variable name and slice with VARIABLE_EXPAND_EMPTY.
let mut res = instr[..varexp_char_idx].to_owned();
if res.as_char_slice().last() == Some(&VARIABLE_EXPAND_SINGLE) {
res.push(VARIABLE_EXPAND_EMPTY);
}
res.push_utfstr(&instr[var_name_and_slice_stop..]);
return expand_variables(res, out, varexp_char_idx, vars, errors);
}
}
// Ok, we have a variable or a history. Let's expand it.
// Start by respecting the sliced elements.
assert!(
var.is_some() || history.is_some(),
"Should have variable or history here",
);
let mut var_item_list = vec![];
if all_values {
var_item_list = if let Some(ref history) = history {
history.get_history()
} else {
var.as_ref().unwrap().as_list().to_vec()
};
} else {
// We have to respect the slice.
if let Some(ref history) = history {
// Ask history to map indexes to item strings.
// Note this may have missing entries for out-of-bounds.
let item_map = history.items_at_indexes(var_idx_list.clone());
for item_index in var_idx_list {
if let Some(item) = item_map.get(&item_index) {
var_item_list.push(item.clone());
}
}
} else {
let all_var_items = var.as_ref().unwrap().as_list();
for item_index in var_idx_list {
// Check that we are within array bounds. If not, skip the element. Note:
// Negative indices (`echo $foo[-1]`) are already converted to positive ones
// here, So tmp < 1 means it's definitely not in.
// Note we are 1-based.
if item_index >= 1 && item_index <= all_var_items.len() {
var_item_list.push(all_var_items[item_index - 1].to_owned());
}
}
}
}
if is_single {
// Quoted expansion. Here we expect the variable's delimiter.
// Note history always has a space delimiter.
let delimit = if history.is_some() {
' '
} else {
var.as_ref().unwrap().get_delimiter()
};
let mut res = instr[..varexp_char_idx].to_owned();
if !res.is_empty() {
if res.as_char_slice().last() != Some(&VARIABLE_EXPAND_SINGLE) {
res.push(INTERNAL_SEPARATOR);
} else if var_item_list.is_empty() || var_item_list[0].is_empty() {
// First expansion is empty, but we need to recursively expand.
res.push(VARIABLE_EXPAND_EMPTY);
}
}
// Append all entries in var_item_list, separated by the delimiter.
res.push_utfstr(&join_strings(&var_item_list, delimit));
res.push_utfstr(&instr[var_name_and_slice_stop..]);
return expand_variables(res, out, varexp_char_idx, vars, errors);
} else {
// Normal cartesian-product expansion.
for item in var_item_list {
if varexp_char_idx == 0 && var_name_and_slice_stop == instr.len() {
if !out.add(item) {
return append_overflow_error(errors, None);
}
} else {
let mut new_in = instr[..varexp_char_idx].to_owned();
if !new_in.is_empty() {
if new_in.as_char_slice().last() != Some(&VARIABLE_EXPAND) {
new_in.push(INTERNAL_SEPARATOR);
} else if item.is_empty() {
new_in.push(VARIABLE_EXPAND_EMPTY);
}
}
new_in.push_utfstr(&item);
new_in.push_utfstr(&instr[var_name_and_slice_stop..]);
let res = expand_variables(new_in, out, varexp_char_idx, vars, errors);
if res.result != ExpandResultCode::ok {
return res;
}
}
}
}
ExpandResult::ok()
}
/// Perform brace expansion, placing the expanded strings into `out`.
fn expand_braces(
input: WString,
flags: ExpandFlags,
out: &mut CompletionReceiver,
errors: &mut Option<&mut ParseErrorList>,
) -> ExpandResult {
let mut syntax_error = false;
let mut brace_count = 0;
let mut brace_begin = None;
let mut brace_end = None;
let mut last_sep = None;
// Locate the first non-nested brace pair.
for (pos, c) in input.chars().enumerate() {
match c {
BRACE_BEGIN => {
if brace_count == 0 {
brace_begin = Some(pos);
}
brace_count += 1;
}
BRACE_END => {
brace_count -= 1;
#[allow(clippy::comparison_chain)]
if brace_count < 0 {
syntax_error = true;
} else if brace_count == 0 {
brace_end = Some(pos);
}
}
BRACE_SEP => {
if brace_count == 1 {
last_sep = Some(pos);
}
}
_ => {
// we ignore all other characters here
}
}
}
if brace_count > 0 {
if !flags.contains(ExpandFlags::FOR_COMPLETIONS) {
syntax_error = true;
} else {
// The user hasn't typed an end brace yet; make one up and append it, then expand
// that.
let mut synth = WString::new();
if let Some(last_sep) = last_sep {
synth.push_utfstr(&input[..brace_begin.unwrap() + 1]);
synth.push_utfstr(&input[last_sep + 1..]);
synth.push(BRACE_END);
} else {
synth.push_utfstr(&input);
synth.push(BRACE_END);
}
// Note: this code looks very fishy, apparently it has never worked.
return expand_braces(synth, ExpandFlags::FAIL_ON_CMDSUBST, out, errors);
}
}
if syntax_error {
append_syntax_error!(errors, SOURCE_LOCATION_UNKNOWN, "Mismatched braces");
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
let Some(brace_begin) = brace_begin else {
// No more brace expansions left; we can return the value as-is.
if !out.add(input) {
return append_overflow_error(errors, None);
}
return ExpandResult::ok();
};
let brace_end = brace_end.unwrap();
let length_preceding_braces = brace_begin;
let length_following_braces = input.len() - brace_end - 1;
let tot_len = length_preceding_braces + length_following_braces;
let mut item_begin = brace_begin + 1;
for (pos, c) in input.chars().enumerate().skip(brace_begin + 1) {
if brace_count == 0 && (c == BRACE_SEP || pos == brace_end) {
assert!(pos >= item_begin);
let item_len = pos - item_begin;
let item = input[item_begin..pos].to_owned();
let mut item = trim(item, Some(wstr::from_char_slice(&[BRACE_SPACE, '\0'])));
for c in item.as_char_slice_mut() {
if *c == BRACE_SPACE {
*c = ' ';
}
}
// `whole_item` is a whitespace- and brace-stripped member of a single pass of brace
// expansion, e.g. in `{ alpha , b,{c, d }}`, `alpha`, `b`, and `c, d` will, in the
// first round of expansion, each in turn be a `whole_item` (with recursive commas
// replaced by special placeholders).
// We recursively call `expand_braces` with each item until it's been fully expanded.
let mut whole_item = WString::new();
whole_item.reserve(tot_len + item_len + 2);
whole_item.push_utfstr(&input[..length_preceding_braces]);
whole_item.push_utfstr(&item);
whole_item.push_utfstr(&input[brace_end + 1..]);
let _ = expand_braces(whole_item, flags, out, errors);
item_begin = pos + 1;
if pos == brace_end {
break;
}
}
if c == BRACE_BEGIN {
brace_count += 1;
}
if c == BRACE_END {
brace_count -= 1;
}
}
ExpandResult::ok()
}
/// Expand a command substitution `input`, executing on `ctx`, and inserting the results into
/// `out_list`, or any errors into `errors`. Return an expand result.
pub fn expand_cmdsubst(
input: WString,
ctx: &OperationContext,
out: &mut CompletionReceiver,
errors: &mut Option<&mut ParseErrorList>,
) -> ExpandResult {
let mut cursor = 0;
let mut is_quoted = false;
let mut has_dollar = false;
let parens = match parse_util_locate_cmdsubst_range(
&input,
&mut cursor,
false,
Some(&mut is_quoted),
Some(&mut has_dollar),
) {
MaybeParentheses::Error => {
append_syntax_error!(errors, SOURCE_LOCATION_UNKNOWN, "Mismatched parenthesis");
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
MaybeParentheses::None => {
if !out.add(input) {
return append_overflow_error(errors, None);
}
return ExpandResult::ok();
}
MaybeParentheses::CommandSubstitution(parens) => parens,
};
let mut sub_res = vec![];
let job_group = ctx.job_group.clone();
let subshell_status = exec_subshell_for_expand(
&input[parens.command()],
ctx.parser(),
job_group.as_ref(),
&mut sub_res,
);
if subshell_status != 0 {
// TODO: Ad-hoc switch, how can we enumerate the possible errors more safely?
let err = match subshell_status {
_ if subshell_status == STATUS_READ_TOO_MUCH.unwrap() => {
wgettext!("Too much data emitted by command substitution so it was discarded")
}
// TODO: STATUS_CMD_ERROR is overused and too generic. We shouldn't have to test things
// to figure out what error to show after we've already been given an error code.
_ if subshell_status == STATUS_CMD_ERROR.unwrap() => {
if ctx.parser().is_eval_depth_exceeded() {
wgettext!("Unable to evaluate string substitution")
} else {
wgettext!("Too many active file descriptors")
}
}
_ if subshell_status == STATUS_CMD_UNKNOWN.unwrap() => {
wgettext!("Unknown command")
}
_ if subshell_status == STATUS_ILLEGAL_CMD.unwrap() => {
wgettext!("Commandname was invalid")
}
_ if subshell_status == STATUS_NOT_EXECUTABLE.unwrap() => {
wgettext!("Command not executable")
}
_ if subshell_status == STATUS_INVALID_ARGS.unwrap() => {
// TODO: Also overused
// This is sent for:
// invalid redirections or pipes (like `<&foo`),
// invalid variables (invalid name or read-only) for for-loops,
// switch $foo if $foo expands to more than one argument
// time in a background job.
wgettext!("Invalid arguments")
}
_ if subshell_status == STATUS_EXPAND_ERROR.unwrap() => {
// Sent in `for $foo in ...` if $foo expands to more than one word
wgettext!("Expansion error")
}
_ if subshell_status == STATUS_UNMATCHED_WILDCARD.unwrap() => {
// Sent in `for $foo in ...` if $foo expands to more than one word
wgettext!("Unmatched wildcard")
}
_ => {
wgettext!("Unknown error while evaluating command substitution")
}
};
append_cmdsub_error_formatted!(errors, parens.start(), parens.end() - 1, err.to_owned());
return ExpandResult::make_error(subshell_status);
}
// Expand slices like (cat /var/words)[1]
let mut tail_begin = parens.end();
if input.as_char_slice().get(tail_begin) == Some(&'[') {
let mut slice_idx = vec![];
let slice_begin = tail_begin;
let slice_end = match parse_slice(&input[slice_begin..], &mut slice_idx, sub_res.len()) {
Ok(offset) => slice_begin + offset,
Err((bad_pos, error)) => {
match error {
ParseSliceError::zero_index => {
append_syntax_error!(
errors,
slice_begin + bad_pos,
"array indices start at 1, not 0."
);
}
ParseSliceError::invalid_index => {
append_syntax_error!(errors, slice_begin + bad_pos, "Invalid index value");
}
}
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
};
let mut sub_res2 = vec![];
tail_begin = slice_end;
for idx in slice_idx {
if idx as usize > sub_res.len() || idx < 1 {
continue;
}
// -1 to convert from 1-based slice index to 0-based vector index.
sub_res2.push(sub_res[idx as usize - 1].to_owned());
}
sub_res = sub_res2;
}
// Recursively call ourselves to expand any remaining command substitutions. The result of this
// recursive call using the tail of the string is inserted into the tail_expand array list
let mut tail_expand_recv = out.subreceiver();
let mut tail = input[tail_begin..].to_owned();
// A command substitution inside double quotes magically closes the quoted string.
// Reopen the quotes just after the command substitution.
if is_quoted {
tail.insert(0, '"');
}
let _ = expand_cmdsubst(tail, ctx, &mut tail_expand_recv, errors); // TODO: offset error locations
let tail_expand = tail_expand_recv.take();
// Combine the result of the current command substitution with the result of the recursive tail
// expansion.
if is_quoted {
// Awkwardly reconstruct the command output.
let approx_size = sub_res.iter().map(|sub_item| sub_item.len() + 1).sum();
let mut sub_res_joined = WString::new();
sub_res_joined.reserve(approx_size);
for line in sub_res {
sub_res_joined.push_utfstr(&escape_string_for_double_quotes(&line));
sub_res_joined.push('\n');
}
// Mimic POSIX shells by stripping all trailing newlines.
if !sub_res_joined.is_empty() {
let mut i = sub_res_joined.len();
while i > 0 && sub_res_joined.as_char_slice()[i - 1] == '\n' {
i -= 1;
}
sub_res_joined.truncate(i);
}
// Instead of performing cartesian product expansion, we directly insert the command
// substitution output into the current expansion results.
for tail_item in tail_expand {
let mut whole_item = WString::new();
whole_item.reserve(
parens.start() + 1 + sub_res_joined.len() + 1 + tail_item.completion.len(),
);
whole_item.push_utfstr(&input[..parens.start() - if has_dollar { 1 } else { 0 }]);
whole_item.push(INTERNAL_SEPARATOR);
whole_item.push_utfstr(&sub_res_joined);
whole_item.push(INTERNAL_SEPARATOR);
whole_item.push_utfstr(&tail_item.completion["\"".len()..]);
if !out.add(whole_item) {
return append_overflow_error(errors, None);
}
}
return ExpandResult::ok();
}
for sub_item in sub_res {
let sub_item2 = escape_string(&sub_item, EscapeStringStyle::Script(EscapeFlags::COMMA));
for tail_item in &*tail_expand {
let mut whole_item = WString::new();
whole_item
.reserve(parens.start() + 1 + sub_item2.len() + 1 + tail_item.completion.len());
whole_item.push_utfstr(&input[..parens.start() - if has_dollar { 1 } else { 0 }]);
whole_item.push(INTERNAL_SEPARATOR);
whole_item.push_utfstr(&sub_item2);
whole_item.push(INTERNAL_SEPARATOR);
whole_item.push_utfstr(&tail_item.completion);
if !out.add(whole_item) {
return append_overflow_error(errors, None);
}
}
}
ExpandResult::ok()
}
// Given that input[0] is HOME_DIRECTORY or tilde (ugh), return the user's name. Return the empty
// string if it is just a tilde. Also return by reference the index of the first character of the
// remaining part of the string (e.g. the subsequent slash).
fn get_home_directory_name<'a>(input: &'a wstr, out_tail_idx: &mut usize) -> &'a wstr {
assert!([HOME_DIRECTORY, '~'].contains(&input.as_char_slice()[0]));
// We get the position of the /, but we need to remove it as well.
if let Some(pos) = input.chars().position(|c| c == '/') {
*out_tail_idx = pos;
&input[1..pos]
} else {
*out_tail_idx = input.len();
&input[1..]
}
}
/// Attempts tilde expansion of the string specified, modifying it in place.
fn expand_home_directory(input: &mut WString, vars: &dyn Environment) {
if input.as_char_slice().first() != Some(&HOME_DIRECTORY) {
return;
}
let mut tail_idx = usize::MAX;
let username = get_home_directory_name(input, &mut tail_idx);
let mut home = None;
if username.is_empty() {
// Current users home directory.
match vars.get_unless_empty(L!("HOME")) {
None => {
input.clear();
return;
}
Some(home_var) => {
home = Some(home_var.as_string());
tail_idx = 1;
}
};
} else {
// Some other user's home directory.
let name_cstr = wcs2zstring(username);
let mut userinfo: libc::passwd = unsafe { std::mem::zeroed() };
let mut result: *mut libc::passwd = std::ptr::null_mut();
let mut buf = [0 as libc::c_char; 8192];
let retval = unsafe {
libc::getpwnam_r(
name_cstr.as_ptr(),
&mut userinfo,
&mut buf[0],
std::mem::size_of_val(&buf),
&mut result,
)
};
if retval == 0 && !result.is_null() {
home = Some(charptr2wcstring(userinfo.pw_dir));
}
}
if let Some(home) = home {
input.replace_range(..tail_idx, &normalize_path(&home, true));
} else {
input.replace_range(0..1, L!("~"));
}
}
/// Expand the %self escape. Note this can only come at the beginning of the string.
fn expand_percent_self(input: &mut WString) {
if input.as_char_slice().first() == Some(&PROCESS_EXPAND_SELF) {
input.replace_range(0..1, &unsafe { libc::getpid() }.to_wstring());
}
}
/// Remove any internal separators. Also optionally convert wildcard characters to regular
/// equivalents. This is done to support skip_wildcards.
fn remove_internal_separator(s: &mut WString, conv: bool) {
// Remove all instances of INTERNAL_SEPARATOR.
s.retain(|c| c != INTERNAL_SEPARATOR);
// If conv is true, replace all instances of ANY_STRING with '*',
// ANY_STRING_RECURSIVE with '*'.
if conv {
for idx in s.as_char_slice_mut() {
match *idx {
ANY_CHAR => {
*idx = '?';
}
ANY_STRING | ANY_STRING_RECURSIVE => {
*idx = '*';
}
_ => {
// we ignore all other characters
}
}
}
}
}
/// A type that knows how to perform expansions.
struct Expander<'a, 'b, 'c> {
/// Operation context for this expansion.
ctx: &'c OperationContext<'b>,
/// Flags to use during expansion.
flags: ExpandFlags,
/// List to receive any errors generated during expansion, or null to ignore errors.
errors: &'c mut Option<&'a mut ParseErrorList>,
}
impl<'a, 'b, 'c> Expander<'a, 'b, 'c> {
fn new(
ctx: &'c OperationContext<'b>,
flags: ExpandFlags,
errors: &'c mut Option<&'a mut ParseErrorList>,
) -> Self {
Self { ctx, flags, errors }
}
fn expand_string(
input: WString,
out_completions: &'a mut CompletionReceiver,
flags: ExpandFlags,
ctx: &'a OperationContext<'b>,
mut errors: Option<&'a mut ParseErrorList>,
) -> ExpandResult {
assert!(
flags.contains(ExpandFlags::FAIL_ON_CMDSUBST) || ctx.has_parser(),
"Must have a parser if not skipping command substitutions"
);
// Early out. If we're not completing, and there's no magic in the input, we're done.
if !flags.contains(ExpandFlags::FOR_COMPLETIONS) && expand_is_clean(&input) {
if !out_completions.add(input) {
return append_overflow_error(&mut errors, None);
}
return ExpandResult::ok();
}
let mut expand = Expander::new(ctx, flags, &mut errors);
// Our expansion stages.
// An expansion stage is a member function pointer.
// It accepts the input string (transferring ownership) and returns the list of output
// completions by reference. It may return an error, which halts expansion.
let stages = [
Expander::stage_cmdsubst,
Expander::stage_variables,
Expander::stage_braces,
Expander::stage_home_and_self,
Expander::stage_wildcards,
];
// Load up our single initial completion.
let mut completions = vec![Completion::from_completion(input.clone())];
let mut total_result = ExpandResult::ok();
let mut output_storage = out_completions.subreceiver();
for stage in stages {
for comp in completions {
if expand.ctx.check_cancel() {
total_result = ExpandResult::new(ExpandResultCode::cancel);
break;
}
let this_result = (stage)(&mut expand, comp.completion, &mut output_storage);
total_result = this_result;
if total_result == ExpandResultCode::error {
break;
}
}
// Output becomes our next stage's input.
completions = output_storage.take();
if total_result == ExpandResultCode::error {
break;
}
}
// This is a little tricky: if one wildcard failed to match but we still got output, it
// means that a previous expansion resulted in multiple strings. For example:
// set dirs ./a ./b
// echo $dirs/*.txt
// Here if ./a/*.txt matches and ./b/*.txt does not, then we don't want to report a failed
// wildcard. So swallow failed-wildcard errors if we got any output.
if total_result == ExpandResultCode::wildcard_no_match && !completions.is_empty() {
total_result = ExpandResult::ok();
}
if total_result == ExpandResultCode::ok {
// Unexpand tildes if we want to preserve them (see #647).
if flags.contains(ExpandFlags::PRESERVE_HOME_TILDES) {
expand.unexpand_tildes(&input, &mut completions);
}
if !out_completions.extend(completions) {
total_result = append_overflow_error(expand.errors, None);
}
}
total_result
}
fn stage_cmdsubst(&mut self, input: WString, out: &mut CompletionReceiver) -> ExpandResult {
builtin commandline: -x for expanded tokens, supplanting -o Issue #10194 reports Cobra completions do set -l args (commandline -opc) eval $args[1] __complete $args[2..] (commandline -ct | string escape) The intent behind "eval" is to expand variables and tildes in "$args". Fair enough. Several of our own completions do the same, see the next commit. The problem with "commandline -o" + "eval" is that the former already removes quotes that are relevant for "eval". This becomes a problem if $args contains quoted () or {}, for example this command will wrongly execute a command substituion: git --work-tree='(launch-missiles)' <TAB> It is possible to escape the string the tokens before running eval, but then there will be no expansion of variables etc. The problem is that "commandline -o" only unescapes tokens so they end up in a weird state somewhere in-between what the user typed and the expanded version. Remove the need for "eval" by introducing "commandline -x" which expands things like variables and braces. This enables custom completion scripts to be aware of shell variables without eval, see the added test for completions to "make -C $var/some/dir ". This means that essentially all third party scripts should migrate from "commandline -o" to "commandline -x". For example set -l tokens if commandline -x >/dev/null 2>&1 set tokens (commandline -xpc) else set tokens (commandline -opc) end Since this is mainly used for completions, the expansion skips command substitutions. They are passed through as-is (instead of cancelling or expanding to nothing) to make custom completion scripts work reasonably well in the common case. Of course there are cases where we would want to expand command substitutions here, so I'm not sure.
2024-01-06 15:45:33 +08:00
if self.flags.contains(ExpandFlags::SKIP_CMDSUBST) {
if !out.add(input) {
return append_overflow_error(self.errors, None);
}
return ExpandResult::ok();
}
if self.flags.contains(ExpandFlags::FAIL_ON_CMDSUBST) {
let mut cursor = 0;
match parse_util_locate_cmdsubst_range(&input, &mut cursor, true, None, None) {
MaybeParentheses::Error => {
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
MaybeParentheses::None => {
if !out.add(input) {
return append_overflow_error(self.errors, None);
}
return ExpandResult::ok();
}
MaybeParentheses::CommandSubstitution(parens) => {
append_cmdsub_error!(
self.errors,
parens.start(),
parens.end()-1,
"command substitutions not allowed in command position. Try var=(your-cmd) $var ..."
);
return ExpandResult::make_error(STATUS_EXPAND_ERROR.unwrap());
}
}
} else {
assert!(
self.ctx.has_parser(),
"Must have a parser to expand command substitutions"
);
expand_cmdsubst(input, self.ctx, out, self.errors)
}
}
// We pass by value to match other stages. NOLINTNEXTLINE(performance-unnecessary-value-param)
fn stage_variables(&mut self, input: WString, out: &mut CompletionReceiver) -> ExpandResult {
// We accept incomplete strings here, since complete uses expand_string to expand incomplete
// strings from the commandline.
let mut next = unescape_string(
&input,
UnescapeStringStyle::Script(UnescapeFlags::SPECIAL | UnescapeFlags::INCOMPLETE),
)
.unwrap_or_default();
if self.flags.contains(ExpandFlags::SKIP_VARIABLES) {
for i in next.as_char_slice_mut() {
if [VARIABLE_EXPAND, VARIABLE_EXPAND_SINGLE].contains(i) {
*i = '$';
}
}
if !out.add(next) {
return append_overflow_error(self.errors, None);
}
ExpandResult::ok()
} else {
let size = next.len();
expand_variables(next, out, size, self.ctx.vars(), self.errors)
}
}
fn stage_braces(&mut self, input: WString, out: &mut CompletionReceiver) -> ExpandResult {
expand_braces(input, self.flags, out, self.errors)
}
fn stage_home_and_self(
&mut self,
mut input: WString,
out: &mut CompletionReceiver,
) -> ExpandResult {
expand_home_directory(&mut input, self.ctx.vars());
if !feature_test(FeatureFlag::remove_percent_self) {
expand_percent_self(&mut input);
}
if !out.add(input) {
return append_overflow_error(self.errors, None);
}
ExpandResult::ok()
}
fn stage_wildcards(
&mut self,
mut path_to_expand: WString,
out: &mut CompletionReceiver,
) -> ExpandResult {
let mut result = ExpandResult::ok();
remove_internal_separator(
&mut path_to_expand,
self.flags.contains(ExpandFlags::SKIP_WILDCARDS),
);
let has_wildcard = wildcard_has_internal(&path_to_expand); // e.g. ANY_STRING
let for_completions = self.flags.contains(ExpandFlags::FOR_COMPLETIONS);
let skip_wildcards = self.flags.contains(ExpandFlags::SKIP_WILDCARDS);
if has_wildcard && self.flags.contains(ExpandFlags::EXECUTABLES_ONLY) {
// don't do wildcard expansion for executables, see issue #785
} else if (for_completions && !skip_wildcards) || has_wildcard {
// We either have a wildcard, or we don't have a wildcard but we're doing completion
// expansion (so we want to get the completion of a file path). Note that if
// skip_wildcards is set, we stomped wildcards in remove_internal_separator above, so
// there actually aren't any.
//
// So we're going to treat this input as a file path. Compute the "working directories",
// which may be CDPATH if the special flag is set.
let working_dir = self.ctx.vars().get_pwd_slash();
let mut effective_working_dirs = vec![];
let for_cd = self.flags.contains(ExpandFlags::SPECIAL_FOR_CD);
let for_command = self.flags.contains(ExpandFlags::SPECIAL_FOR_COMMAND);
if !for_cd && !for_command {
// Common case.
effective_working_dirs.push(working_dir);
} else {
// Either special_for_command or special_for_cd. We can handle these
// mostly the same. There's the following differences:
//
// 1. An empty CDPATH should be treated as '.', but an empty PATH should be left empty
// (no commands can be found). Also, an empty element in either is treated as '.' for
// consistency with POSIX shells. Note that we rely on the latter by having called
// `munge_colon_delimited_array()` for these special env vars. Thus we do not
// special-case them here.
//
// 2. PATH is only "one level," while CDPATH is multiple levels. That is, input like
// 'foo/bar' should resolve against CDPATH, but not PATH.
//
// In either case, we ignore the path if we start with ./ or /. Also ignore it if we are
// doing command completion and we contain a slash, per IEEE 1003.1, chapter 8 under
// PATH.
if path_to_expand.starts_with(L!("/"))
|| path_to_expand.starts_with(L!("./"))
|| path_to_expand.starts_with(L!("../"))
|| (for_command && path_to_expand.contains('/'))
{
effective_working_dirs.push(working_dir);
} else {
// Get the PATH/CDPATH and CWD. Perhaps these should be passed in. An empty CDPATH
// implies just the current directory, while an empty PATH is left empty.
let mut paths = self
.ctx
.vars()
.get(if for_cd { L!("CDPATH") } else { L!("PATH") })
.map(|var| var.as_list().to_owned())
.unwrap_or_default();
// The current directory is always valid.
paths.push(if for_cd { L!(".") } else { L!("") }.to_owned());
for next_path in paths {
effective_working_dirs
.push(path_apply_working_directory(&next_path, &working_dir));
}
}
}
result = ExpandResult::new(ExpandResultCode::wildcard_no_match);
let mut expanded_recv = out.subreceiver();
for effective_working_dir in effective_working_dirs {
let expand_res = wildcard_expand_string(
&path_to_expand,
&effective_working_dir,
self.flags,
&*self.ctx.cancel_checker,
&mut expanded_recv,
);
match expand_res {
WildcardResult::Match => result = ExpandResult::ok(),
WildcardResult::NoMatch => (),
WildcardResult::Overflow => return append_overflow_error(self.errors, None),
WildcardResult::Cancel => return ExpandResult::new(ExpandResultCode::cancel),
}
}
let mut expanded = expanded_recv.take();
expanded.sort_by(|a, b| wcsfilecmp_glob(&a.completion, &b.completion));
if !out.extend(expanded) {
result = ExpandResult::new(ExpandResultCode::error);
}
} else {
// Can't fully justify this check. I think it's that SKIP_WILDCARDS is used when completing
// to mean don't do file expansions, so if we're not doing file expansions, just drop this
// completion on the floor.
#[allow(clippy::collapsible_if)]
if !self.flags.contains(ExpandFlags::FOR_COMPLETIONS) {
if !out.add(path_to_expand) {
return append_overflow_error(self.errors, None);
}
}
}
result
}
// Given an original input string, if it starts with a tilde, "unexpand" the expanded home
// directory. Note this may be just a tilde or a user name like ~foo/.
fn unexpand_tildes(&self, input: &wstr, completions: &mut CompletionList) {
// If input begins with tilde, then try to replace the corresponding string in each completion
// with the tilde. If it does not, there's nothing to do.
if input.as_char_slice().first() != Some(&'~') {
return;
}
// This is a subtle kludge. We need to decide whether to unexpand tildes for all
// completions, or only those which replace their tokens. The problem is that we're sloppy
// about setting the COMPLETE_REPLACES_TOKEN flag, except when we're completing in the
// wildcard stage, because no other clients of string expansion care. Example:
// HOME=/foo
// mkdir ~/foo # makes /foo/foo
// cd ~/<tab>
// Here we are likely to get a completion 'foo' which may match $HOME, but it extends its token
// instead of replacing it, so we don't modify it (it will just be appended to the original ~/).
//
// However if we are not completing, just expanding, then expansion just produces the full paths
// so we should unconditionally unexpand tildes.
let only_replacers = self.flags.contains(ExpandFlags::FOR_COMPLETIONS);
// Helper to decide whether to process a completion.
let should_process = |c: &Completion| !only_replacers || c.replaces_token();
// Early out if none qualify.
if !completions.iter().any(should_process) {
return;
}
// Get the username_with_tilde (like ~bert) and expand it into a home directory.
let mut tail_idx = usize::MAX;
let username_with_tilde =
WString::from_str("~") + get_home_directory_name(input, &mut tail_idx);
let mut home = username_with_tilde.clone();
expand_tilde(&mut home, self.ctx.vars());
// Now for each completion that starts with home, replace it with the username_with_tilde.
for comp in completions {
if should_process(comp) && comp.completion.starts_with(&home) {
comp.completion
.replace_range(..home.len(), &username_with_tilde);
// And mark that our tilde is literal, so it doesn't try to escape it.
comp.flags |= CompleteFlags::DONT_ESCAPE_TILDES;
}
}
}
}
#[derive(Copy, Clone, Debug, Eq, PartialEq)]
pub enum ExpandResultCode {
/// There was an error, for example, unmatched braces.
error,
/// Expansion succeeded.
ok,
/// Expansion was cancelled (e.g. control-C).
cancel,
/// Expansion succeeded, but a wildcard in the string matched no files,
/// so the output is empty.
wildcard_no_match,
}
/// These are the possible return values for expand_string.
#[must_use]
#[derive(Debug)]
pub struct ExpandResult {
/// The result of expansion.
pub result: ExpandResultCode,
/// If expansion resulted in an error, this is an appropriate value with which to populate
/// $status.
// todo!("should be c_int?");
pub status: i32,
2023-04-09 20:33:20 +08:00
}