Simplify: run generation hooks first, then trailing-whitespace cleanup

Instead of manually stripping trailing whitespace in each generation
script, reorder pre-commit hooks so:
1. Generation hooks run FIRST (generate-manifests, update-values, etc.)
2. trailing-whitespace hook runs AFTER to clean up generated files

This removes the strip_trailing_whitespace() function and all its calls,
making the code simpler and letting pre-commit handle formatting.

Author: nammn

.pre-commit-config.yaml

@@ -13,7 +13,54 @@
 
 repos:
   # =============================================================================
-  # Standard pre-commit hooks
+  # Generation hooks (run FIRST so formatters can clean up generated files)
+  # =============================================================================
+  - repo: local
+    hooks:
+      - id: generate-manifests
+        name: Generate CRD manifests
+        entry: scripts/dev/generate_files.sh generate_manifests
+        language: script
+        files: ^api/.*\.go$
+        pass_filenames: false
+
+      - id: update-release-json
+        name: Update release.json
+        entry: scripts/dev/generate_files.sh update_release
+        language: script
+        files: (release\.json|scripts/evergreen/release/)
+        pass_filenames: false
+
+      - id: update-values-yaml
+        name: Update Helm values files
+        entry: scripts/dev/generate_files.sh update_values
+        language: script
+        files: ^helm_chart/values.*\.yaml$
+        pass_filenames: false
+
+      - id: generate-standalone-yaml
+        name: Generate standalone YAML files
+        entry: scripts/dev/generate_files.sh generate_standalone_yaml
+        language: script
+        files: ^helm_chart/
+        pass_filenames: false
+
+      - id: regenerate-mco-tests
+        name: Regenerate MCO evergreen tests
+        entry: scripts/dev/generate_files.sh update_mco_tests
+        language: script
+        files: ^scripts/evergreen/e2e/mco/
+        pass_filenames: false
+
+      - id: regenerate-multicluster-rbac
+        name: Regenerate multi-cluster RBAC samples
+        entry: scripts/dev/regenerate_multicluster_rbac.sh
+        language: script
+        files: ^(cmd/kubectl-mongodb|pkg/kubectl-mongodb)/
+        pass_filenames: false
+
+  # =============================================================================
+  # Standard pre-commit hooks (run AFTER generation to clean up generated files)
   # =============================================================================
   - repo: https://github.com/pre-commit/pre-commit-hooks
     rev: v5.0.0
@@ -128,13 +175,10 @@ repos:
           )
 
   # =============================================================================
-  # Local/custom hooks for project-specific tasks
+  # Local/custom linting hooks
   # =============================================================================
   - repo: local
     hooks:
-      # -------------------------------------------------------------------------
-      # Linting hooks
-      # -------------------------------------------------------------------------
       - id: golangci-lint
         name: Go linting
         entry: scripts/evergreen/lint_code.sh
@@ -188,51 +232,6 @@ repos:
         files: ^helm_chart/
         pass_filenames: false
 
-      # -------------------------------------------------------------------------
-      # Generation hooks (run files generation and stage changes)
-      # -------------------------------------------------------------------------
-      - id: generate-manifests
-        name: Generate CRD manifests
-        entry: scripts/dev/generate_files.sh generate_manifests
-        language: script
-        files: ^api/.*\.go$
-        pass_filenames: false
-
-      - id: update-release-json
-        name: Update release.json
-        entry: scripts/dev/generate_files.sh update_release
-        language: script
-        files: (release\.json|scripts/evergreen/release/)
-        pass_filenames: false
-
-      - id: update-values-yaml
-        name: Update Helm values files
-        entry: scripts/dev/generate_files.sh update_values
-        language: script
-        files: ^helm_chart/values.*\.yaml$
-        pass_filenames: false
-
-      - id: generate-standalone-yaml
-        name: Generate standalone YAML files
-        entry: scripts/dev/generate_files.sh generate_standalone_yaml
-        language: script
-        files: ^helm_chart/
-        pass_filenames: false
-
-      - id: regenerate-mco-tests
-        name: Regenerate MCO evergreen tests
-        entry: scripts/dev/generate_files.sh update_mco_tests
-        language: script
-        files: ^scripts/evergreen/e2e/mco/
-        pass_filenames: false
-
-      - id: regenerate-multicluster-rbac
-        name: Regenerate multi-cluster RBAC samples
-        entry: scripts/dev/regenerate_multicluster_rbac.sh
-        language: script
-        files: ^(cmd/kubectl-mongodb|pkg/kubectl-mongodb)/
-        pass_filenames: false
-
 # CI mode settings (set via environment variable PRE_COMMIT_FROM_REF/PRE_COMMIT_TO_REF)
 # In CI, pre-commit will check all files: pre-commit run --all-files
 ci:

scripts/dev/generate_files.sh

@@ -1,30 +1,16 @@
 #!/usr/bin/env bash
 #
 # File generation script for pre-commit hooks.
-# This script generates various files (YAML configs, manifests, etc.) and ensures
-# they are properly formatted without trailing whitespace.
+# This script generates various files (YAML configs, manifests, etc.).
+# Trailing whitespace cleanup is handled by the trailing-whitespace hook
+# which runs AFTER these generation hooks.
 #
 
 set -Eeou pipefail
 
 source scripts/dev/set_env_context.sh
 source scripts/funcs/printing
 
-# Helper function to strip trailing whitespace from a file
-# This ensures generated files don't cause pre-commit hook cycles
-strip_trailing_whitespace() {
-  local file="$1"
-  if [[ -f "${file}" ]]; then
-    # Use sed to remove trailing whitespace from each line
-    # Note: sed -i '' (with space) is required for macOS, sed -i'' for Linux
-    if [[ "$(uname)" == "Darwin" ]]; then
-      sed -i '' 's/[[:space:]]*$//' "${file}"
-    else
-      sed -i 's/[[:space:]]*$//' "${file}"
-    fi
-  fi
-}
-
 if [ -f "${PROJECT_DIR}/venv/bin/activate" ]; then
   source "${PROJECT_DIR}/venv/bin/activate"
 fi
@@ -34,7 +20,6 @@ mkdir -p "$(go env GOPATH)/bin"
 update_mco_tests() {
   echo "Regenerating MCO evergreen tests configuration"
   python scripts/evergreen/e2e/mco/create_mco_tests.py >.evergreen-mco.yml
-  strip_trailing_whitespace .evergreen-mco.yml
   git add .evergreen-mco.yml
 }
 
@@ -82,29 +67,11 @@ generate_standalone_yaml() {
   rm -rf "${charttmpdir:?}"/*
   helm template --namespace mongodb -f helm_chart/values.yaml helm_chart --output-dir "${charttmpdir}" --values helm_chart/values-multi-cluster.yaml "$@"
   cat "${FILES[@]}" >public/mongodb-kubernetes-multi-cluster.yaml
-
-  # Strip trailing whitespace from all generated files
-  strip_trailing_whitespace public/mongodb-kubernetes.yaml
-  strip_trailing_whitespace public/mongodb-kubernetes-openshift.yaml
-  strip_trailing_whitespace public/mongodb-kubernetes-multi-cluster.yaml
-  strip_trailing_whitespace public/crds.yaml
-  strip_trailing_whitespace config/manager/manager.yaml
-  strip_trailing_whitespace config/rbac/database-roles.yaml
-  strip_trailing_whitespace config/rbac/operator-roles-base.yaml
-  strip_trailing_whitespace config/rbac/operator-roles-clustermongodbroles.yaml
-  strip_trailing_whitespace config/rbac/operator-roles-pvc-resize.yaml
-  strip_trailing_whitespace config/rbac/operator-roles-telemetry.yaml
 }
 
 generate_manifests() {
   make manifests
 
-  # Strip trailing whitespace from generated CRD files
-  for f in config/crd/bases/*.yaml helm_chart/crds/*.yaml; do
-    strip_trailing_whitespace "${f}"
-  done
-  strip_trailing_whitespace public/crds.yaml
-
   git add config/crd/bases
   git add helm_chart/crds
   git add public/crds.yaml
@@ -115,10 +82,6 @@ update_values_yaml_files() {
   # shellcheck disable=SC2154
   python scripts/evergreen/release/update_helm_values_files.py
 
-  # Strip trailing whitespace from generated files
-  strip_trailing_whitespace helm_chart/values.yaml
-  strip_trailing_whitespace helm_chart/values-openshift.yaml
-
   # commit any changes we made
   git add helm_chart/values.yaml
   git add helm_chart/values-openshift.yaml

scripts/dev/regenerate_multicluster_rbac.sh

@@ -1,7 +1,6 @@
 #!/usr/bin/env bash
 set -Eeou pipefail
 
-# Source the environment context to get go in PATH
 # shellcheck disable=SC1091
 source scripts/dev/set_env_context.sh
 

scripts/evergreen/check_precommit.sh

@@ -6,13 +6,9 @@ initial_index_state=$(git diff --name-only --cached --diff-filter=AM)
 
 export EVERGREEN_MODE=true
 
-# Source the environment context (sets up PATH for go, etc.)
-# This is created by clone->setup_context step in CI
 # shellcheck disable=SC1091
 source scripts/dev/set_env_context.sh
 
-# Activate the venv if it exists (CI creates this via python_venv setup step)
-# The venv contains pre-commit installed from requirements.txt
 if [[ -f "${PROJECT_DIR}/venv/bin/activate" ]]; then
   echo "Activating venv..."
   # shellcheck disable=SC1091

scripts/evergreen/lint_code.sh

@@ -1,7 +1,6 @@
 #!/usr/bin/env bash
 set -Eeou pipefail
 
-# Source the environment context to get go in PATH
 # shellcheck disable=SC1091
 source scripts/dev/set_env_context.sh