feat: allow PID to be excluded from process group (sort of)

src/procmon.bash

@@ -14,6 +14,10 @@ _pm_init_output() {
   mkdir -p "${_pm_output}"
 }
 
+# Starts the process monitor.
+# Returns:
+#   1 if the process monitor is already running
+#   0 otherwise
 pm_start() {
   _pm_assert_state_not "running" || return 1
   _pm_init_output
@@ -53,6 +57,17 @@ pm_start() {
   return 0
 }
 
+# Tracks the last job started by the shell as part of a monitored group.
+# If a tracked process dies:
+#  1. without an error code (e.g. it is killed);
+#  2. with a non-zero error code.
+# Then all processes in the process group and their descendants (with caveats)
+# are also killed. This makes sure that the harness does not leave processes
+# behind.
+#
+# Returns:
+#   1 if the process monitor is not running
+#   0 otherwise
 pm_track_last_job() {
   _pm_assert_state "running" || return 1
 
@@ -62,6 +77,29 @@ pm_track_last_job() {
   fi
 }
 
+# Stops tracking a given PID.
+# Arguments:
+#   $1: PID to stop tracking
+# Returns:
+#   1 if the process monitor is not running
+#   0 otherwise
+# Note:
+#   This function is flaky. The process monitor
+#   might still see the PID as tracked after this
+#   function returns for a short period of time,
+#   so do not rely too much on it.
+pm_stop_tracking() {
+  _pm_assert_state "running" || return 1
+
+  local pid=$1
+  rm -rf "${_pm_output}/${pid}.pid" || true
+}
+
+# Returns the list of PIDs being tracked by the process monitor by
+# setting the `result` variable.
+# Returns:
+#   1 if the process monitor is not running
+#   0 otherwise
 pm_known_pids() {
   _pm_assert_state "running" || return 1
 
@@ -114,10 +152,18 @@ _pm_halt() {
   pm_kill_rec "$_pm_pid"
 }
 
+# Stops the process monitor, killing the entire process group.
+# Returns:
+#   1 if the process monitor is not running
+#   0 otherwise
 pm_stop() {
   _pm_halt "halted"
 }
 
+pm_join() {
+  await "$_pm_pid" "$1"
+}
+
 pm_job_exit() {
   exit_code=$1
   echo "$exit_code" > "${_pm_output}/${BASHPID}.pid"

test/test_procmon.bats

@@ -130,7 +130,33 @@ setup() {
   await "$p1"
   await "$p2"
 
-  sleep 1
+  pm_join 3
 
   assert_equal $(pm_state) "halted_process_failure"
+}
+
+@test "should no longer track a process if requested" {
+  assert pm_start
+
+  (
+    while true; do
+      sleep 1
+    done
+    pm_job_exit 1
+  ) &
+  pid1=$!
+  pm_track_last_job
+
+  pm_stop_tracking $pid1
+  kill -SIGKILL $pid1
+  await "$pid1"
+
+  # Again, we need to allow time for the procmon
+  # to pick up on the kill.
+  sleep 1
+
+  pm_stop
+  pm_join 3
+
+  assert_equal $(pm_state) "halted"
 }