fix: move inline codex-rs/core unit tests into sibling files (#14444)

## Why
PR #13783 moved the `codex.rs` unit tests into `codex_tests.rs`. This
applies the same extraction pattern across the rest of `codex-rs/core`
so the production modules stay focused on runtime code instead of large
inline test blocks.

Keeping the tests in sibling files also makes follow-up edits easier to
review because product changes no longer have to share a file with
hundreds or thousands of lines of test scaffolding.

## What changed
- replaced each inline `mod tests { ... }` in `codex-rs/core/src/**`
with a path-based module declaration
- moved each extracted unit test module into a sibling `*_tests.rs`
file, using `mod_tests.rs` for `mod.rs` modules
- preserved the existing `cfg(...)` guards and module-local structure so
the refactor remains structural rather than behavioral

## Testing
- `cargo test -p codex-core --lib` (`1653 passed; 0 failed; 5 ignored`)
- `just fix -p codex-core`
- `cargo fmt --check`
- `cargo shear`
This commit is contained in:
Michael Bolin
2026-03-12 08:16:36 -07:00
committed by GitHub
Unverified
parent 7f2ca502f5
commit 0c8a36676a
252 changed files with 40158 additions and 40383 deletions
@@ -251,40 +251,5 @@ async fn resolve_aggregated_output(
}
#[cfg(test)]
mod tests {
use super::split_valid_utf8_prefix_with_max;
use pretty_assertions::assert_eq;
#[test]
fn split_valid_utf8_prefix_respects_max_bytes_for_ascii() {
let mut buf = b"hello word!".to_vec();
let first = split_valid_utf8_prefix_with_max(&mut buf, 5).expect("expected prefix");
assert_eq!(first, b"hello".to_vec());
assert_eq!(buf, b" word!".to_vec());
let second = split_valid_utf8_prefix_with_max(&mut buf, 5).expect("expected prefix");
assert_eq!(second, b" word".to_vec());
assert_eq!(buf, b"!".to_vec());
}
#[test]
fn split_valid_utf8_prefix_avoids_splitting_utf8_codepoints() {
// "é" is 2 bytes in UTF-8. With a max of 3 bytes, we should only emit 1 char (2 bytes).
let mut buf = "ééé".as_bytes().to_vec();
let first = split_valid_utf8_prefix_with_max(&mut buf, 3).expect("expected prefix");
assert_eq!(std::str::from_utf8(&first).unwrap(), "é");
assert_eq!(buf, "éé".as_bytes().to_vec());
}
#[test]
fn split_valid_utf8_prefix_makes_progress_on_invalid_utf8() {
let mut buf = vec![0xff, b'a', b'b'];
let first = split_valid_utf8_prefix_with_max(&mut buf, 2).expect("expected prefix");
assert_eq!(first, vec![0xff]);
assert_eq!(buf, b"ab".to_vec());
}
}
#[path = "async_watcher_tests.rs"]
mod tests;
@@ -0,0 +1,35 @@
use super::split_valid_utf8_prefix_with_max;
use pretty_assertions::assert_eq;
#[test]
fn split_valid_utf8_prefix_respects_max_bytes_for_ascii() {
let mut buf = b"hello word!".to_vec();
let first = split_valid_utf8_prefix_with_max(&mut buf, 5).expect("expected prefix");
assert_eq!(first, b"hello".to_vec());
assert_eq!(buf, b" word!".to_vec());
let second = split_valid_utf8_prefix_with_max(&mut buf, 5).expect("expected prefix");
assert_eq!(second, b" word".to_vec());
assert_eq!(buf, b"!".to_vec());
}
#[test]
fn split_valid_utf8_prefix_avoids_splitting_utf8_codepoints() {
// "é" is 2 bytes in UTF-8. With a max of 3 bytes, we should only emit 1 char (2 bytes).
let mut buf = "ééé".as_bytes().to_vec();
let first = split_valid_utf8_prefix_with_max(&mut buf, 3).expect("expected prefix");
assert_eq!(std::str::from_utf8(&first).unwrap(), "é");
assert_eq!(buf, "éé".as_bytes().to_vec());
}
#[test]
fn split_valid_utf8_prefix_makes_progress_on_invalid_utf8() {
let mut buf = vec![0xff, b'a', b'b'];
let first = split_valid_utf8_prefix_with_max(&mut buf, 2).expect("expected prefix");
assert_eq!(first, vec![0xff]);
assert_eq!(buf, b"ab".to_vec());
}
@@ -179,94 +179,5 @@ impl HeadTailBuffer {
}
#[cfg(test)]
mod tests {
use super::HeadTailBuffer;
use pretty_assertions::assert_eq;
#[test]
fn keeps_prefix_and_suffix_when_over_budget() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
assert_eq!(buf.omitted_bytes(), 0);
// Exceeds max by 2; we should keep head+tail and omit the middle.
buf.push_chunk(b"ab".to_vec());
assert!(buf.omitted_bytes() > 0);
let rendered = String::from_utf8_lossy(&buf.to_bytes()).to_string();
assert!(rendered.starts_with("01234"));
assert!(rendered.ends_with("89ab"));
}
#[test]
fn max_bytes_zero_drops_everything() {
let mut buf = HeadTailBuffer::new(0);
buf.push_chunk(b"abc".to_vec());
assert_eq!(buf.retained_bytes(), 0);
assert_eq!(buf.omitted_bytes(), 3);
assert_eq!(buf.to_bytes(), b"".to_vec());
assert_eq!(buf.snapshot_chunks(), Vec::<Vec<u8>>::new());
}
#[test]
fn head_budget_zero_keeps_only_last_byte_in_tail() {
let mut buf = HeadTailBuffer::new(1);
buf.push_chunk(b"abc".to_vec());
assert_eq!(buf.retained_bytes(), 1);
assert_eq!(buf.omitted_bytes(), 2);
assert_eq!(buf.to_bytes(), b"c".to_vec());
}
#[test]
fn draining_resets_state() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
buf.push_chunk(b"ab".to_vec());
let drained = buf.drain_chunks();
assert!(!drained.is_empty());
assert_eq!(buf.retained_bytes(), 0);
assert_eq!(buf.omitted_bytes(), 0);
assert_eq!(buf.to_bytes(), b"".to_vec());
}
#[test]
fn chunk_larger_than_tail_budget_keeps_only_tail_end() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
// Tail budget is 5 bytes. This chunk should replace the tail and keep only its last 5 bytes.
buf.push_chunk(b"ABCDEFGHIJK".to_vec());
let out = String::from_utf8_lossy(&buf.to_bytes()).to_string();
assert!(out.starts_with("01234"));
assert!(out.ends_with("GHIJK"));
assert!(buf.omitted_bytes() > 0);
}
#[test]
fn fills_head_then_tail_across_multiple_chunks() {
let mut buf = HeadTailBuffer::new(10);
// Fill the 5-byte head budget across multiple chunks.
buf.push_chunk(b"01".to_vec());
buf.push_chunk(b"234".to_vec());
assert_eq!(buf.to_bytes(), b"01234".to_vec());
// Then fill the 5-byte tail budget.
buf.push_chunk(b"567".to_vec());
buf.push_chunk(b"89".to_vec());
assert_eq!(buf.to_bytes(), b"0123456789".to_vec());
assert_eq!(buf.omitted_bytes(), 0);
// One more byte causes the tail to drop its oldest byte.
buf.push_chunk(b"a".to_vec());
assert_eq!(buf.to_bytes(), b"012346789a".to_vec());
assert_eq!(buf.omitted_bytes(), 1);
}
}
#[path = "head_tail_buffer_tests.rs"]
mod tests;
@@ -0,0 +1,89 @@
use super::HeadTailBuffer;
use pretty_assertions::assert_eq;
#[test]
fn keeps_prefix_and_suffix_when_over_budget() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
assert_eq!(buf.omitted_bytes(), 0);
// Exceeds max by 2; we should keep head+tail and omit the middle.
buf.push_chunk(b"ab".to_vec());
assert!(buf.omitted_bytes() > 0);
let rendered = String::from_utf8_lossy(&buf.to_bytes()).to_string();
assert!(rendered.starts_with("01234"));
assert!(rendered.ends_with("89ab"));
}
#[test]
fn max_bytes_zero_drops_everything() {
let mut buf = HeadTailBuffer::new(0);
buf.push_chunk(b"abc".to_vec());
assert_eq!(buf.retained_bytes(), 0);
assert_eq!(buf.omitted_bytes(), 3);
assert_eq!(buf.to_bytes(), b"".to_vec());
assert_eq!(buf.snapshot_chunks(), Vec::<Vec<u8>>::new());
}
#[test]
fn head_budget_zero_keeps_only_last_byte_in_tail() {
let mut buf = HeadTailBuffer::new(1);
buf.push_chunk(b"abc".to_vec());
assert_eq!(buf.retained_bytes(), 1);
assert_eq!(buf.omitted_bytes(), 2);
assert_eq!(buf.to_bytes(), b"c".to_vec());
}
#[test]
fn draining_resets_state() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
buf.push_chunk(b"ab".to_vec());
let drained = buf.drain_chunks();
assert!(!drained.is_empty());
assert_eq!(buf.retained_bytes(), 0);
assert_eq!(buf.omitted_bytes(), 0);
assert_eq!(buf.to_bytes(), b"".to_vec());
}
#[test]
fn chunk_larger_than_tail_budget_keeps_only_tail_end() {
let mut buf = HeadTailBuffer::new(10);
buf.push_chunk(b"0123456789".to_vec());
// Tail budget is 5 bytes. This chunk should replace the tail and keep only its last 5 bytes.
buf.push_chunk(b"ABCDEFGHIJK".to_vec());
let out = String::from_utf8_lossy(&buf.to_bytes()).to_string();
assert!(out.starts_with("01234"));
assert!(out.ends_with("GHIJK"));
assert!(buf.omitted_bytes() > 0);
}
#[test]
fn fills_head_then_tail_across_multiple_chunks() {
let mut buf = HeadTailBuffer::new(10);
// Fill the 5-byte head budget across multiple chunks.
buf.push_chunk(b"01".to_vec());
buf.push_chunk(b"234".to_vec());
assert_eq!(buf.to_bytes(), b"01234".to_vec());
// Then fill the 5-byte tail budget.
buf.push_chunk(b"567".to_vec());
buf.push_chunk(b"89".to_vec());
assert_eq!(buf.to_bytes(), b"0123456789".to_vec());
assert_eq!(buf.omitted_bytes(), 0);
// One more byte causes the tail to drop its oldest byte.
buf.push_chunk(b"a".to_vec());
assert_eq!(buf.to_bytes(), b"012346789a".to_vec());
assert_eq!(buf.omitted_bytes(), 1);
}
+2 -347
View File
@@ -169,350 +169,5 @@ pub(crate) fn generate_chunk_id() -> String {
#[cfg(test)]
#[cfg(unix)]
mod tests {
use super::head_tail_buffer::HeadTailBuffer;
use super::*;
use crate::codex::Session;
use crate::codex::TurnContext;
use crate::codex::make_session_and_context;
use crate::protocol::AskForApproval;
use crate::protocol::SandboxPolicy;
use crate::tools::context::ExecCommandToolOutput;
use crate::unified_exec::ExecCommandRequest;
use crate::unified_exec::WriteStdinRequest;
use core_test_support::skip_if_sandbox;
use std::sync::Arc;
use tokio::time::Duration;
async fn test_session_and_turn() -> (Arc<Session>, Arc<TurnContext>) {
let (session, mut turn) = make_session_and_context().await;
turn.approval_policy
.set(AskForApproval::Never)
.expect("test setup should allow updating approval policy");
turn.sandbox_policy
.set(SandboxPolicy::DangerFullAccess)
.expect("test setup should allow updating sandbox policy");
turn.file_system_sandbox_policy =
codex_protocol::permissions::FileSystemSandboxPolicy::from(turn.sandbox_policy.get());
turn.network_sandbox_policy =
codex_protocol::permissions::NetworkSandboxPolicy::from(turn.sandbox_policy.get());
(Arc::new(session), Arc::new(turn))
}
async fn exec_command(
session: &Arc<Session>,
turn: &Arc<TurnContext>,
cmd: &str,
yield_time_ms: u64,
) -> Result<ExecCommandToolOutput, UnifiedExecError> {
let context =
UnifiedExecContext::new(Arc::clone(session), Arc::clone(turn), "call".to_string());
let process_id = session
.services
.unified_exec_manager
.allocate_process_id()
.await;
session
.services
.unified_exec_manager
.exec_command(
ExecCommandRequest {
command: vec!["bash".to_string(), "-lc".to_string(), cmd.to_string()],
process_id,
yield_time_ms,
max_output_tokens: None,
workdir: None,
network: None,
tty: true,
sandbox_permissions: SandboxPermissions::UseDefault,
additional_permissions: None,
additional_permissions_preapproved: false,
justification: None,
prefix_rule: None,
},
&context,
)
.await
}
async fn write_stdin(
session: &Arc<Session>,
process_id: i32,
input: &str,
yield_time_ms: u64,
) -> Result<ExecCommandToolOutput, UnifiedExecError> {
session
.services
.unified_exec_manager
.write_stdin(WriteStdinRequest {
process_id,
input,
yield_time_ms,
max_output_tokens: None,
})
.await
}
#[test]
fn push_chunk_preserves_prefix_and_suffix() {
let mut buffer = HeadTailBuffer::default();
buffer.push_chunk(vec![b'a'; UNIFIED_EXEC_OUTPUT_MAX_BYTES]);
buffer.push_chunk(vec![b'b']);
buffer.push_chunk(vec![b'c']);
assert_eq!(buffer.retained_bytes(), UNIFIED_EXEC_OUTPUT_MAX_BYTES);
let snapshot = buffer.snapshot_chunks();
let first = snapshot.first().expect("expected at least one chunk");
assert_eq!(first.first(), Some(&b'a'));
assert!(snapshot.iter().any(|chunk| chunk.as_slice() == b"b"));
assert_eq!(
snapshot
.last()
.expect("expected at least one chunk")
.as_slice(),
b"c"
);
}
#[test]
fn head_tail_buffer_default_preserves_prefix_and_suffix() {
let mut buffer = HeadTailBuffer::default();
buffer.push_chunk(vec![b'a'; UNIFIED_EXEC_OUTPUT_MAX_BYTES]);
buffer.push_chunk(b"bc".to_vec());
let rendered = buffer.to_bytes();
assert_eq!(rendered.first(), Some(&b'a'));
assert!(rendered.ends_with(b"bc"));
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn unified_exec_persists_across_requests() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process_id");
write_stdin(
&session,
process_id,
"export CODEX_INTERACTIVE_SHELL_VAR=codex\n",
2_500,
)
.await?;
let out_2 = write_stdin(
&session,
process_id,
"echo $CODEX_INTERACTIVE_SHELL_VAR\n",
2_500,
)
.await?;
assert!(
out_2.truncated_output().contains("codex"),
"expected environment variable output"
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn multi_unified_exec_sessions() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let shell_a = exec_command(&session, &turn, "bash -i", 2_500).await?;
let session_a = shell_a.process_id.expect("expected process id");
write_stdin(
&session,
session_a,
"export CODEX_INTERACTIVE_SHELL_VAR=codex\n",
2_500,
)
.await?;
let out_2 =
exec_command(&session, &turn, "echo $CODEX_INTERACTIVE_SHELL_VAR", 2_500).await?;
tokio::time::sleep(Duration::from_secs(2)).await;
assert!(
out_2.process_id.is_none(),
"short command should not report a process id if it exits quickly"
);
assert!(
!out_2.truncated_output().contains("codex"),
"short command should run in a fresh shell"
);
let out_3 = write_stdin(
&session,
shell_a.process_id.expect("expected process id"),
"echo $CODEX_INTERACTIVE_SHELL_VAR\n",
2_500,
)
.await?;
assert!(
out_3.truncated_output().contains("codex"),
"session should preserve state"
);
Ok(())
}
#[tokio::test]
async fn unified_exec_timeouts() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
const TEST_VAR_VALUE: &str = "unified_exec_var_123";
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process id");
write_stdin(
&session,
process_id,
format!("export CODEX_INTERACTIVE_SHELL_VAR={TEST_VAR_VALUE}\n").as_str(),
2_500,
)
.await?;
let out_2 = write_stdin(
&session,
process_id,
"sleep 5 && echo $CODEX_INTERACTIVE_SHELL_VAR\n",
10,
)
.await?;
assert!(
!out_2.truncated_output().contains(TEST_VAR_VALUE),
"timeout too short should yield incomplete output"
);
tokio::time::sleep(Duration::from_secs(7)).await;
let out_3 = write_stdin(&session, process_id, "", 100).await?;
assert!(
out_3.truncated_output().contains(TEST_VAR_VALUE),
"subsequent poll should retrieve output"
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn unified_exec_pause_blocks_yield_timeout() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
session.set_out_of_band_elicitation_pause_state(true);
let paused_session = Arc::clone(&session);
tokio::spawn(async move {
tokio::time::sleep(Duration::from_secs(2)).await;
paused_session.set_out_of_band_elicitation_pause_state(false);
});
let started = tokio::time::Instant::now();
let response =
exec_command(&session, &turn, "sleep 1 && echo unified-exec-done", 250).await?;
assert!(
started.elapsed() >= Duration::from_secs(2),
"pause should block the unified exec yield timeout"
);
assert!(
response.truncated_output().contains("unified-exec-done"),
"exec_command should wait for output after the pause lifts"
);
assert!(
response.process_id.is_none(),
"completed command should not leave a background process"
);
Ok(())
}
#[tokio::test]
#[ignore] // Ignored while we have a better way to test this.
async fn requests_with_large_timeout_are_capped() -> anyhow::Result<()> {
let (session, turn) = test_session_and_turn().await;
let result = exec_command(&session, &turn, "echo codex", 120_000).await?;
assert!(result.process_id.is_some());
assert!(result.truncated_output().contains("codex"));
Ok(())
}
#[tokio::test]
#[ignore] // Ignored while we have a better way to test this.
async fn completed_commands_do_not_persist_sessions() -> anyhow::Result<()> {
let (session, turn) = test_session_and_turn().await;
let result = exec_command(&session, &turn, "echo codex", 2_500).await?;
assert!(
result.process_id.is_some(),
"completed command should report a process id"
);
assert!(result.truncated_output().contains("codex"));
assert!(
session
.services
.unified_exec_manager
.process_store
.lock()
.await
.processes
.is_empty()
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn reusing_completed_process_returns_unknown_process() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process id");
write_stdin(&session, process_id, "exit\n", 2_500).await?;
tokio::time::sleep(Duration::from_millis(200)).await;
let err = write_stdin(&session, process_id, "", 100)
.await
.expect_err("expected unknown process error");
match err {
UnifiedExecError::UnknownProcessId { process_id: err_id } => {
assert_eq!(err_id, process_id, "process id should match request");
}
other => panic!("expected UnknownProcessId, got {other:?}"),
}
assert!(
session
.services
.unified_exec_manager
.process_store
.lock()
.await
.processes
.is_empty()
);
Ok(())
}
}
#[path = "mod_tests.rs"]
mod tests;
+343
View File
@@ -0,0 +1,343 @@
use super::head_tail_buffer::HeadTailBuffer;
use super::*;
use crate::codex::Session;
use crate::codex::TurnContext;
use crate::codex::make_session_and_context;
use crate::protocol::AskForApproval;
use crate::protocol::SandboxPolicy;
use crate::tools::context::ExecCommandToolOutput;
use crate::unified_exec::ExecCommandRequest;
use crate::unified_exec::WriteStdinRequest;
use core_test_support::skip_if_sandbox;
use std::sync::Arc;
use tokio::time::Duration;
async fn test_session_and_turn() -> (Arc<Session>, Arc<TurnContext>) {
let (session, mut turn) = make_session_and_context().await;
turn.approval_policy
.set(AskForApproval::Never)
.expect("test setup should allow updating approval policy");
turn.sandbox_policy
.set(SandboxPolicy::DangerFullAccess)
.expect("test setup should allow updating sandbox policy");
turn.file_system_sandbox_policy =
codex_protocol::permissions::FileSystemSandboxPolicy::from(turn.sandbox_policy.get());
turn.network_sandbox_policy =
codex_protocol::permissions::NetworkSandboxPolicy::from(turn.sandbox_policy.get());
(Arc::new(session), Arc::new(turn))
}
async fn exec_command(
session: &Arc<Session>,
turn: &Arc<TurnContext>,
cmd: &str,
yield_time_ms: u64,
) -> Result<ExecCommandToolOutput, UnifiedExecError> {
let context =
UnifiedExecContext::new(Arc::clone(session), Arc::clone(turn), "call".to_string());
let process_id = session
.services
.unified_exec_manager
.allocate_process_id()
.await;
session
.services
.unified_exec_manager
.exec_command(
ExecCommandRequest {
command: vec!["bash".to_string(), "-lc".to_string(), cmd.to_string()],
process_id,
yield_time_ms,
max_output_tokens: None,
workdir: None,
network: None,
tty: true,
sandbox_permissions: SandboxPermissions::UseDefault,
additional_permissions: None,
additional_permissions_preapproved: false,
justification: None,
prefix_rule: None,
},
&context,
)
.await
}
async fn write_stdin(
session: &Arc<Session>,
process_id: i32,
input: &str,
yield_time_ms: u64,
) -> Result<ExecCommandToolOutput, UnifiedExecError> {
session
.services
.unified_exec_manager
.write_stdin(WriteStdinRequest {
process_id,
input,
yield_time_ms,
max_output_tokens: None,
})
.await
}
#[test]
fn push_chunk_preserves_prefix_and_suffix() {
let mut buffer = HeadTailBuffer::default();
buffer.push_chunk(vec![b'a'; UNIFIED_EXEC_OUTPUT_MAX_BYTES]);
buffer.push_chunk(vec![b'b']);
buffer.push_chunk(vec![b'c']);
assert_eq!(buffer.retained_bytes(), UNIFIED_EXEC_OUTPUT_MAX_BYTES);
let snapshot = buffer.snapshot_chunks();
let first = snapshot.first().expect("expected at least one chunk");
assert_eq!(first.first(), Some(&b'a'));
assert!(snapshot.iter().any(|chunk| chunk.as_slice() == b"b"));
assert_eq!(
snapshot
.last()
.expect("expected at least one chunk")
.as_slice(),
b"c"
);
}
#[test]
fn head_tail_buffer_default_preserves_prefix_and_suffix() {
let mut buffer = HeadTailBuffer::default();
buffer.push_chunk(vec![b'a'; UNIFIED_EXEC_OUTPUT_MAX_BYTES]);
buffer.push_chunk(b"bc".to_vec());
let rendered = buffer.to_bytes();
assert_eq!(rendered.first(), Some(&b'a'));
assert!(rendered.ends_with(b"bc"));
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn unified_exec_persists_across_requests() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process_id");
write_stdin(
&session,
process_id,
"export CODEX_INTERACTIVE_SHELL_VAR=codex\n",
2_500,
)
.await?;
let out_2 = write_stdin(
&session,
process_id,
"echo $CODEX_INTERACTIVE_SHELL_VAR\n",
2_500,
)
.await?;
assert!(
out_2.truncated_output().contains("codex"),
"expected environment variable output"
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn multi_unified_exec_sessions() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let shell_a = exec_command(&session, &turn, "bash -i", 2_500).await?;
let session_a = shell_a.process_id.expect("expected process id");
write_stdin(
&session,
session_a,
"export CODEX_INTERACTIVE_SHELL_VAR=codex\n",
2_500,
)
.await?;
let out_2 = exec_command(&session, &turn, "echo $CODEX_INTERACTIVE_SHELL_VAR", 2_500).await?;
tokio::time::sleep(Duration::from_secs(2)).await;
assert!(
out_2.process_id.is_none(),
"short command should not report a process id if it exits quickly"
);
assert!(
!out_2.truncated_output().contains("codex"),
"short command should run in a fresh shell"
);
let out_3 = write_stdin(
&session,
shell_a.process_id.expect("expected process id"),
"echo $CODEX_INTERACTIVE_SHELL_VAR\n",
2_500,
)
.await?;
assert!(
out_3.truncated_output().contains("codex"),
"session should preserve state"
);
Ok(())
}
#[tokio::test]
async fn unified_exec_timeouts() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
const TEST_VAR_VALUE: &str = "unified_exec_var_123";
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process id");
write_stdin(
&session,
process_id,
format!("export CODEX_INTERACTIVE_SHELL_VAR={TEST_VAR_VALUE}\n").as_str(),
2_500,
)
.await?;
let out_2 = write_stdin(
&session,
process_id,
"sleep 5 && echo $CODEX_INTERACTIVE_SHELL_VAR\n",
10,
)
.await?;
assert!(
!out_2.truncated_output().contains(TEST_VAR_VALUE),
"timeout too short should yield incomplete output"
);
tokio::time::sleep(Duration::from_secs(7)).await;
let out_3 = write_stdin(&session, process_id, "", 100).await?;
assert!(
out_3.truncated_output().contains(TEST_VAR_VALUE),
"subsequent poll should retrieve output"
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn unified_exec_pause_blocks_yield_timeout() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
session.set_out_of_band_elicitation_pause_state(true);
let paused_session = Arc::clone(&session);
tokio::spawn(async move {
tokio::time::sleep(Duration::from_secs(2)).await;
paused_session.set_out_of_band_elicitation_pause_state(false);
});
let started = tokio::time::Instant::now();
let response = exec_command(&session, &turn, "sleep 1 && echo unified-exec-done", 250).await?;
assert!(
started.elapsed() >= Duration::from_secs(2),
"pause should block the unified exec yield timeout"
);
assert!(
response.truncated_output().contains("unified-exec-done"),
"exec_command should wait for output after the pause lifts"
);
assert!(
response.process_id.is_none(),
"completed command should not leave a background process"
);
Ok(())
}
#[tokio::test]
#[ignore] // Ignored while we have a better way to test this.
async fn requests_with_large_timeout_are_capped() -> anyhow::Result<()> {
let (session, turn) = test_session_and_turn().await;
let result = exec_command(&session, &turn, "echo codex", 120_000).await?;
assert!(result.process_id.is_some());
assert!(result.truncated_output().contains("codex"));
Ok(())
}
#[tokio::test]
#[ignore] // Ignored while we have a better way to test this.
async fn completed_commands_do_not_persist_sessions() -> anyhow::Result<()> {
let (session, turn) = test_session_and_turn().await;
let result = exec_command(&session, &turn, "echo codex", 2_500).await?;
assert!(
result.process_id.is_some(),
"completed command should report a process id"
);
assert!(result.truncated_output().contains("codex"));
assert!(
session
.services
.unified_exec_manager
.process_store
.lock()
.await
.processes
.is_empty()
);
Ok(())
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn reusing_completed_process_returns_unknown_process() -> anyhow::Result<()> {
skip_if_sandbox!(Ok(()));
let (session, turn) = test_session_and_turn().await;
let open_shell = exec_command(&session, &turn, "bash -i", 2_500).await?;
let process_id = open_shell.process_id.expect("expected process id");
write_stdin(&session, process_id, "exit\n", 2_500).await?;
tokio::time::sleep(Duration::from_millis(200)).await;
let err = write_stdin(&session, process_id, "", 100)
.await
.expect_err("expected unknown process error");
match err {
UnifiedExecError::UnknownProcessId { process_id: err_id } => {
assert_eq!(err_id, process_id, "process id should match request");
}
other => panic!("expected UnknownProcessId, got {other:?}"),
}
assert!(
session
.services
.unified_exec_manager
.process_store
.lock()
.await
.processes
.is_empty()
);
Ok(())
}
@@ -832,104 +832,5 @@ enum ProcessStatus {
}
#[cfg(test)]
mod tests {
use super::*;
use pretty_assertions::assert_eq;
use tokio::time::Duration;
use tokio::time::Instant;
#[test]
fn unified_exec_env_injects_defaults() {
let env = apply_unified_exec_env(HashMap::new());
let expected = HashMap::from([
("NO_COLOR".to_string(), "1".to_string()),
("TERM".to_string(), "dumb".to_string()),
("LANG".to_string(), "C.UTF-8".to_string()),
("LC_CTYPE".to_string(), "C.UTF-8".to_string()),
("LC_ALL".to_string(), "C.UTF-8".to_string()),
("COLORTERM".to_string(), String::new()),
("PAGER".to_string(), "cat".to_string()),
("GIT_PAGER".to_string(), "cat".to_string()),
("GH_PAGER".to_string(), "cat".to_string()),
("CODEX_CI".to_string(), "1".to_string()),
]);
assert_eq!(env, expected);
}
#[test]
fn unified_exec_env_overrides_existing_values() {
let mut base = HashMap::new();
base.insert("NO_COLOR".to_string(), "0".to_string());
base.insert("PATH".to_string(), "/usr/bin".to_string());
let env = apply_unified_exec_env(base);
assert_eq!(env.get("NO_COLOR"), Some(&"1".to_string()));
assert_eq!(env.get("PATH"), Some(&"/usr/bin".to_string()));
}
#[test]
fn pruning_prefers_exited_processes_outside_recently_used() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), true),
(3, now - Duration::from_secs(20), false),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), false),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
assert_eq!(candidate, Some(2));
}
#[test]
fn pruning_falls_back_to_lru_when_no_exited() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), false),
(3, now - Duration::from_secs(20), false),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), false),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
assert_eq!(candidate, Some(1));
}
#[test]
fn pruning_protects_recent_processes_even_if_exited() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), false),
(3, now - Duration::from_secs(20), true),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), true),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
// (10) is exited but among the last 8; we should drop the LRU outside that set.
assert_eq!(candidate, Some(1));
}
}
#[path = "process_manager_tests.rs"]
mod tests;
@@ -0,0 +1,99 @@
use super::*;
use pretty_assertions::assert_eq;
use tokio::time::Duration;
use tokio::time::Instant;
#[test]
fn unified_exec_env_injects_defaults() {
let env = apply_unified_exec_env(HashMap::new());
let expected = HashMap::from([
("NO_COLOR".to_string(), "1".to_string()),
("TERM".to_string(), "dumb".to_string()),
("LANG".to_string(), "C.UTF-8".to_string()),
("LC_CTYPE".to_string(), "C.UTF-8".to_string()),
("LC_ALL".to_string(), "C.UTF-8".to_string()),
("COLORTERM".to_string(), String::new()),
("PAGER".to_string(), "cat".to_string()),
("GIT_PAGER".to_string(), "cat".to_string()),
("GH_PAGER".to_string(), "cat".to_string()),
("CODEX_CI".to_string(), "1".to_string()),
]);
assert_eq!(env, expected);
}
#[test]
fn unified_exec_env_overrides_existing_values() {
let mut base = HashMap::new();
base.insert("NO_COLOR".to_string(), "0".to_string());
base.insert("PATH".to_string(), "/usr/bin".to_string());
let env = apply_unified_exec_env(base);
assert_eq!(env.get("NO_COLOR"), Some(&"1".to_string()));
assert_eq!(env.get("PATH"), Some(&"/usr/bin".to_string()));
}
#[test]
fn pruning_prefers_exited_processes_outside_recently_used() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), true),
(3, now - Duration::from_secs(20), false),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), false),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
assert_eq!(candidate, Some(2));
}
#[test]
fn pruning_falls_back_to_lru_when_no_exited() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), false),
(3, now - Duration::from_secs(20), false),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), false),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
assert_eq!(candidate, Some(1));
}
#[test]
fn pruning_protects_recent_processes_even_if_exited() {
let now = Instant::now();
let meta = vec![
(1, now - Duration::from_secs(40), false),
(2, now - Duration::from_secs(30), false),
(3, now - Duration::from_secs(20), true),
(4, now - Duration::from_secs(19), false),
(5, now - Duration::from_secs(18), false),
(6, now - Duration::from_secs(17), false),
(7, now - Duration::from_secs(16), false),
(8, now - Duration::from_secs(15), false),
(9, now - Duration::from_secs(14), false),
(10, now - Duration::from_secs(13), true),
];
let candidate = UnifiedExecProcessManager::process_id_to_prune_from_meta(&meta);
// (10) is exited but among the last 8; we should drop the LRU outside that set.
assert_eq!(candidate, Some(1));
}