[codex] simplify memory read metrics (#28164)

## Why

Memory read telemetry currently reconstructs the executable shell
command after a tool call finishes. That duplicates shell, login-policy,
and cwd resolution owned by the tool handlers, and can diverge from the
environment-specific command that unified exec actually ran.

## What changed

- Expose the existing restricted shell-script parser directly for raw
script text.
- Parse `shell_command` and `exec_command` input into plain command argv
before classifying memory reads.
- Preserve all-or-nothing safe-command validation for multi-command
scripts.
- Remove cwd resolution, shell selection, and the unnecessary async
boundary from memory read metric emission.

## Testing

- `just test -p codex-shell-command`
- `cargo check -p codex-core`
This commit is contained in:
pakrym-oai
2026-06-15 08:28:02 -07:00
committed by GitHub
Unverified
parent bbcfed8ac2
commit 42ad752f36
6 changed files with 36 additions and 58 deletions
+8 -5
View File
@@ -94,6 +94,12 @@ pub fn try_parse_word_only_commands_sequence(tree: &Tree, src: &str) -> Option<V
Some(commands)
}
/// Parses a shell script consisting only of plain commands joined by safe operators.
pub fn parse_shell_script_into_commands(script: &str) -> Option<Vec<Vec<String>>> {
let tree = try_parse_shell(script)?;
try_parse_word_only_commands_sequence(&tree, script)
}
pub fn extract_bash_command(command: &[String]) -> Option<(&str, &str)> {
let [shell, flag, script] = command else {
return None;
@@ -114,9 +120,7 @@ pub fn extract_bash_command(command: &[String]) -> Option<(&str, &str)> {
/// joined by safe operators.
pub fn parse_shell_lc_plain_commands(command: &[String]) -> Option<Vec<Vec<String>>> {
let (_, script) = extract_bash_command(command)?;
let tree = try_parse_shell(script)?;
try_parse_word_only_commands_sequence(&tree, script)
parse_shell_script_into_commands(script)
}
/// Returns the parsed argv for a single shell command in a here-doc style
@@ -322,8 +326,7 @@ mod tests {
use pretty_assertions::assert_eq;
fn parse_seq(src: &str) -> Option<Vec<Vec<String>>> {
let tree = try_parse_shell(src)?;
try_parse_word_only_commands_sequence(&tree, src)
parse_shell_script_into_commands(src)
}
#[test]
+9 -5
View File
@@ -1818,7 +1818,11 @@ fn parse_find_query_and_path(tail: &[String]) -> (Option<String>, Option<String>
fn parse_shell_lc_commands(original: &[String]) -> Option<Vec<ParsedCommand>> {
// Only handle bash/zsh here; PowerShell is stripped separately without bash parsing.
let (_, script) = extract_bash_command(original)?;
Some(parse_shell_script(script))
}
/// Parses command metadata from a Bash-compatible shell script.
pub fn parse_shell_script(script: &str) -> Vec<ParsedCommand> {
if let Some(tree) = try_parse_shell(script)
&& let Some(all_commands) = try_parse_word_only_commands_sequence(&tree, script)
&& !all_commands.is_empty()
@@ -1831,9 +1835,9 @@ fn parse_shell_lc_commands(original: &[String]) -> Option<Vec<ParsedCommand>> {
// Commands arrive in source order; drop formatting helpers while preserving it.
let filtered_commands = drop_small_formatting_commands(all_commands);
if filtered_commands.is_empty() {
return Some(vec![ParsedCommand::Unknown {
return vec![ParsedCommand::Unknown {
cmd: script.to_string(),
}]);
}];
}
// Build parsed commands, tracking `cd` segments to compute effective file paths.
let mut commands: Vec<ParsedCommand> = Vec::new();
@@ -1939,11 +1943,11 @@ fn parse_shell_lc_commands(original: &[String]) -> Option<Vec<ParsedCommand>> {
})
.collect();
}
return Some(commands);
return commands;
}
Some(vec![ParsedCommand::Unknown {
vec![ParsedCommand::Unknown {
cmd: script.to_string(),
}])
}]
}
/// Return true if this looks like a small formatting helper in a pipeline.