Revert the parallel test attempt

Author: PastelStorm

CHANGELOG.md

@@ -11,7 +11,6 @@
 * Add explicit dependencies for `backoff`, `pandas`, `psutil`, `pypdf`, and `requests` (previously only transitive via `unstructured[all-docs]`)
 * Pre-download NLTK models before parallel test runs to prevent race conditions
 * Parallelize Docker smoke tests by running one container per xdist worker on dedicated ports
-* Parallelize parallel-mode curl tests with background processes
 * Pin uv version in Dockerfile for reproducible builds
 * Remove `py3.12-pip` from Dockerfile (unused since uv migration)
 * Drop mypy from CI (ruff covers linting sufficiently)

scripts/parallel-mode-test.sh

@@ -1,21 +1,16 @@
 #!/usr/bin/env bash
 
 # parallel-mode-test.sh
-# Run each test case against two instances of the api (single mode vs parallel mode)
-# and diff the outputs to make sure parallel mode does not alter the response.
-# All test cases run concurrently for speed.
+# Iterate a list of curl commands, and run each one against two instances of the api
+# The smoke test will start one container with parallel mode and one without, and
+# diff the two outputs to make sure parallel mode does not alter the response.
 # Note the filepaths assume you ran this from the top level
 
 # shellcheck disable=SC2317  # Shellcheck complains that trap functions are unreachable...
 
-set -e
-
 base_url_1=$1
 base_url_2=$2
 
-tmpdir=$(mktemp -d)
-trap 'rm -rf "$tmpdir"' EXIT
-
 declare -a curl_params=(
     "-F files=@sample-docs/layout-parser-paper.pdf -F 'strategy=fast'"
     "-F files=@sample-docs/layout-parser-paper.pdf -F 'strategy=auto"
@@ -26,71 +21,42 @@ declare -a curl_params=(
     "-F files=@sample-docs/layout-parser-paper.pdf -F 'hi_res_model_name=yolox'"
 )
 
-run_test_case() {
-    local idx=$1
-    local params=$2
-    local single_output="$tmpdir/single_${idx}.json"
-    local parallel_output="$tmpdir/parallel_${idx}.json"
-
-    curl_command="curl $base_url_1/general/v0/general $params"
-    echo "[$idx] Testing: $curl_command"
-
-    # Run in single mode
-    # Note(austin): Parallel mode screws up hierarchy! While we deal with that,
-    # let's ignore parent_id fields in the results
-    $curl_command 2> /dev/null | jq -S 'del(..|.parent_id?)' > "$single_output"
-
-    # Stop if curl didn't work
-    if [ ! -s "$single_output" ]; then
-        echo "[$idx] Single mode command failed!"
-        $curl_command
-        return 1
-    fi
-
-    # Run in parallel mode
-    curl_command="curl $base_url_2/general/v0/general $params"
-    $curl_command 2> /dev/null | jq -S 'del(..|.parent_id?)' > "$parallel_output"
-
-    # Stop if curl didn't work
-    if [ ! -s "$parallel_output" ]; then
-        echo "[$idx] Parallel mode command failed!"
-        $curl_command
-        return 1
-    fi
-
-    local original_length
-    local parallel_length
-    original_length=$(jq 'length' "$single_output")
-    parallel_length=$(jq 'length' "$parallel_output")
-
-    if [[ "$original_length" != "$parallel_length" ]]; then
-        echo "[$idx] Parallel mode returned a different number of elements! ($original_length vs $parallel_length)"
-        echo "[$idx] Params: $params"
-        return 1
-    fi
-
-    echo "[$idx] PASSED ($original_length elements)"
-}
-
-# Launch all test cases concurrently
-pids=()
-for i in "${!curl_params[@]}"; do
-    run_test_case "$i" "${curl_params[$i]}" &
-    pids+=($!)
+for params in "${curl_params[@]}"
+do
+   curl_command="curl $base_url_1/general/v0/general $params"
+   echo Testing: "$curl_command"
+
+   # Run in single mode
+   # Note(austin): Parallel mode screws up hierarchy! While we deal with that,
+   # let's ignore parent_id fields in the results
+   $curl_command 2> /dev/null | jq -S 'del(..|.parent_id?)' > output.json
+   original_length=$(jq 'length' output.json)
+
+   # Stop if curl didn't work
+   if [ ! -s output.json ]; then
+       echo Command failed!
+       $curl_command
+       exit 1
+   fi
+
+   # Run in parallel mode
+   curl_command="curl $base_url_2/general/v0/general $params"
+   $curl_command 2> /dev/null | jq -S 'del(..|.parent_id?)' > parallel_output.json
+   parallel_length=$(jq 'length' parallel_output.json)
+
+   # Stop if curl didn't work
+   if [ ! -s parallel_output.json ]; then
+       echo Command failed!
+       $curl_command
+       exit 1
+   fi
+
+   if ! [[ "$original_length" == "$parallel_length" ]]; then
+       echo Parallel mode returned a different number of elements!
+       echo Params: "$params"
+       exit 1
+   fi
+
+   rm -f output.json parallel_output.json
+   echo
 done
-
-# Wait for all and collect failures
-failed=0
-for i in "${!pids[@]}"; do
-    if ! wait "${pids[$i]}"; then
-        echo "Test case $i failed!"
-        failed=1
-    fi
-done
-
-if [ "$failed" -ne 0 ]; then
-    echo "Some parallel mode tests failed!"
-    exit 1
-fi
-
-echo "All parallel mode tests passed!"