test: add 6 regression tests from real-world diffctx bug reports

Author: nikolay-e

tests/cases/diff/algorithm/algorithm_022_unchanged_unrelated_excluded_after.yaml

@@ -0,0 +1,142 @@
+argocd/apps/myapp/values.yaml: |
+  replicaCount: 3
+  image:
+    repository: ghcr.io/myorg/myapp
+    tag: main-def5678
+  service:
+    type: ClusterIP
+    port: 8080
+  ingress:
+    enabled: true
+    host: myapp.example.com
+    annotations:
+      nginx.ingress.kubernetes.io/proxy-body-size: "50m"
+  resources:
+    limits:
+      cpu: "1"
+      memory: 512Mi
+    requests:
+      cpu: 250m
+      memory: 256Mi
+  autoscaling:
+    enabled: true
+    minReplicas: 2
+    maxReplicas: 5
+    targetCPUUtilization: 80
+
+argocd/apps/myapp/Chart.yaml: |
+  apiVersion: v2
+  name: myapp
+  version: 0.2.0
+  appVersion: "1.1.0"
+  description: My application Helm chart
+  dependencies:
+    - name: postgresql
+      version: "12.x.x"
+      repository: "https://charts.bitnami.com/bitnami"
+      condition: postgresql.enabled
+
+ansible/scripts/music_consistency_checker.py: |
+  import os
+  from pathlib import Path
+  from typing import List, Dict, Optional
+
+  def scan_library(root_path: str) -> List[Dict]:
+      results = []
+      for entry in Path(root_path).rglob("*.flac"):
+          results.append({"path": str(entry), "format": "flac"})
+      for entry in Path(root_path).rglob("*.mp3"):
+          results.append({"path": str(entry), "format": "mp3"})
+      return results
+
+  def check_metadata(file_path: str) -> Dict[str, Optional[str]]:
+      return {
+          "title": None,
+          "artist": None,
+          "album": None,
+          "track_number": None,
+          "genre": None,
+      }
+
+  def validate_naming_convention(file_path: str) -> bool:
+      name = Path(file_path).stem
+      parts = name.split(" - ")
+      return len(parts) >= 2
+
+  def find_duplicates(library: List[Dict]) -> List[List[Dict]]:
+      seen = {}
+      duplicates = []
+      for item in library:
+          key = Path(item["path"]).stem
+          if key in seen:
+              duplicates.append([seen[key], item])
+          else:
+              seen[key] = item
+      return duplicates
+
+  def generate_report(library: List[Dict], duplicates: List[List[Dict]]) -> str:
+      lines = [f"Total files: {len(library)}", f"Duplicates: {len(duplicates)}"]
+      return "\n".join(lines)
+
+scripts/sops-new.sh: |
+  #!/bin/bash
+  set -euo pipefail
+
+  encrypt_file() {
+      local input_file="$1"
+      local output_file="$2"
+      sops --encrypt --age "$(cat ~/.sops/age-key.pub)" \
+          --input-type yaml --output-type yaml \
+          "$input_file" > "$output_file"
+  }
+
+  decrypt_file() {
+      local input_file="$1"
+      sops --decrypt --input-type yaml --output-type yaml "$input_file"
+  }
+
+  rotate_keys() {
+      local file="$1"
+      sops updatekeys -y "$file"
+  }
+
+  if [[ "${1:-}" == "encrypt" ]]; then
+      encrypt_file "${2:-}" "${3:-}"
+  elif [[ "${1:-}" == "decrypt" ]]; then
+      decrypt_file "${2:-}"
+  elif [[ "${1:-}" == "rotate" ]]; then
+      rotate_keys "${2:-}"
+  else
+      echo "Usage: $0 {encrypt|decrypt|rotate} [args...]"
+      exit 1
+  fi
+
+garbage_notifier.py: |
+  GARBAGE_ALG022_NOTIFIER_MARKER_A = "notifier"
+  GARBAGE_ALG022_CHANNEL_MARKER_B = True
+
+  class SlackNotifier:
+      GARBAGE_ALG022_SLACK_MARKER_C = "slack"
+
+      def send(self, channel: str, message: str) -> bool:
+          GARBAGE_ALG022_SEND_MARKER_D = "sent"
+          return True
+
+      def format_block(self, text: str) -> dict:
+          GARBAGE_ALG022_BLOCK_MARKER_E = "block"
+          return {"text": text}
+
+garbage_health.py: |
+  GARBAGE_ALG022_HEALTH_MARKER_F = "health"
+  GARBAGE_ALG022_PROBE_MARKER_G = 30
+
+  class HealthChecker:
+      GARBAGE_ALG022_CHECKER_MARKER_H = "checker"
+
+      def check_db(self) -> bool:
+          GARBAGE_ALG022_DB_MARKER_I = "ok"
+          return True
+
+      def check_redis(self) -> bool:
+          GARBAGE_ALG022_REDIS_MARKER_J = "ok"
+          return True

tests/cases/diff/algorithm/algorithm_022_unchanged_unrelated_excluded_before.yaml

@@ -0,0 +1,130 @@
+argocd/apps/myapp/values.yaml: |
+  replicaCount: 2
+  image:
+    repository: ghcr.io/myorg/myapp
+    tag: main-abc1234
+  service:
+    type: ClusterIP
+    port: 8080
+  ingress:
+    enabled: true
+    host: myapp.example.com
+  resources:
+    limits:
+      cpu: 500m
+      memory: 256Mi
+    requests:
+      cpu: 100m
+      memory: 128Mi
+
+argocd/apps/myapp/Chart.yaml: |
+  apiVersion: v2
+  name: myapp
+  version: 0.1.0
+  appVersion: "1.0.0"
+  description: My application Helm chart
+
+ansible/scripts/music_consistency_checker.py: |
+  import os
+  from pathlib import Path
+  from typing import List, Dict, Optional
+
+  def scan_library(root_path: str) -> List[Dict]:
+      results = []
+      for entry in Path(root_path).rglob("*.flac"):
+          results.append({"path": str(entry), "format": "flac"})
+      for entry in Path(root_path).rglob("*.mp3"):
+          results.append({"path": str(entry), "format": "mp3"})
+      return results
+
+  def check_metadata(file_path: str) -> Dict[str, Optional[str]]:
+      return {
+          "title": None,
+          "artist": None,
+          "album": None,
+          "track_number": None,
+          "genre": None,
+      }
+
+  def validate_naming_convention(file_path: str) -> bool:
+      name = Path(file_path).stem
+      parts = name.split(" - ")
+      return len(parts) >= 2
+
+  def find_duplicates(library: List[Dict]) -> List[List[Dict]]:
+      seen = {}
+      duplicates = []
+      for item in library:
+          key = Path(item["path"]).stem
+          if key in seen:
+              duplicates.append([seen[key], item])
+          else:
+              seen[key] = item
+      return duplicates
+
+  def generate_report(library: List[Dict], duplicates: List[List[Dict]]) -> str:
+      lines = [f"Total files: {len(library)}", f"Duplicates: {len(duplicates)}"]
+      return "\n".join(lines)
+
+scripts/sops-new.sh: |
+  #!/bin/bash
+  set -euo pipefail
+
+  encrypt_file() {
+      local input_file="$1"
+      local output_file="$2"
+      sops --encrypt --age "$(cat ~/.sops/age-key.pub)" \
+          --input-type yaml --output-type yaml \
+          "$input_file" > "$output_file"
+  }
+
+  decrypt_file() {
+      local input_file="$1"
+      sops --decrypt --input-type yaml --output-type yaml "$input_file"
+  }
+
+  rotate_keys() {
+      local file="$1"
+      sops updatekeys -y "$file"
+  }
+
+  if [[ "${1:-}" == "encrypt" ]]; then
+      encrypt_file "${2:-}" "${3:-}"
+  elif [[ "${1:-}" == "decrypt" ]]; then
+      decrypt_file "${2:-}"
+  elif [[ "${1:-}" == "rotate" ]]; then
+      rotate_keys "${2:-}"
+  else
+      echo "Usage: $0 {encrypt|decrypt|rotate} [args...]"
+      exit 1
+  fi
+
+garbage_notifier.py: |
+  GARBAGE_ALG022_NOTIFIER_MARKER_A = "notifier"
+  GARBAGE_ALG022_CHANNEL_MARKER_B = True
+
+  class SlackNotifier:
+      GARBAGE_ALG022_SLACK_MARKER_C = "slack"
+
+      def send(self, channel: str, message: str) -> bool:
+          GARBAGE_ALG022_SEND_MARKER_D = "sent"
+          return True
+
+      def format_block(self, text: str) -> dict:
+          GARBAGE_ALG022_BLOCK_MARKER_E = "block"
+          return {"text": text}
+
+garbage_health.py: |
+  GARBAGE_ALG022_HEALTH_MARKER_F = "health"
+  GARBAGE_ALG022_PROBE_MARKER_G = 30
+
+  class HealthChecker:
+      GARBAGE_ALG022_CHECKER_MARKER_H = "checker"
+
+      def check_db(self) -> bool:
+          GARBAGE_ALG022_DB_MARKER_I = "ok"
+          return True
+
+      def check_redis(self) -> bool:
+          GARBAGE_ALG022_REDIS_MARKER_J = "ok"
+          return True

tests/cases/diff/algorithm/algorithm_022_unchanged_unrelated_excluded_diffctx.yaml

@@ -0,0 +1,26 @@
+must_include_files:
+- argocd/apps/myapp/values.yaml
+- argocd/apps/myapp/Chart.yaml
+must_not_include:
+- music_consistency_checker.py
+- scan_library
+- check_metadata
+- validate_naming_convention
+- find_duplicates
+- generate_report
+- sops-new.sh
+- encrypt_file
+- decrypt_file
+- rotate_keys
+- GARBAGE_ALG022_NOTIFIER_MARKER_A
+- GARBAGE_ALG022_CHANNEL_MARKER_B
+- GARBAGE_ALG022_SLACK_MARKER_C
+- GARBAGE_ALG022_SEND_MARKER_D
+- GARBAGE_ALG022_BLOCK_MARKER_E
+- GARBAGE_ALG022_HEALTH_MARKER_F
+- GARBAGE_ALG022_PROBE_MARKER_G
+- GARBAGE_ALG022_CHECKER_MARKER_H
+- GARBAGE_ALG022_DB_MARKER_I
+- GARBAGE_ALG022_REDIS_MARKER_J
+- garbage_notifier.py
+- garbage_health.py

tests/cases/diff/cicd/cicd_016_workflow_yaml_changes_fragmented_after.yaml

@@ -0,0 +1,82 @@
+.github/workflows/cd.yml: |
+  name: CD
+  on:
+    push:
+      branches: [main]
+
+  env:
+    REGISTRY: ghcr.io
+    IMAGE_NAME: ${{ github.repository }}
+
+  jobs:
+    deploy:
+      runs-on: ubuntu-latest
+      permissions:
+        contents: read
+        packages: write
+      steps:
+        - uses: actions/checkout@v4
+
+        - name: Login to GHCR
+          uses: docker/login-action@v3
+          with:
+            registry: ghcr.io
+            username: ${{ github.actor }}
+            password: ${{ secrets.GITHUB_TOKEN }}
+
+        - name: Generate image tags
+          id: tags
+          run: |
+            PRIMARY_TAG="main-$(git rev-parse --short HEAD)"
+            echo "primary_tag=${PRIMARY_TAG}" >> "$GITHUB_OUTPUT"
+
+        - name: Set up Docker Buildx
+          uses: docker/setup-buildx-action@v3
+
+        - name: Build and push
+          uses: docker/build-push-action@v6
+          with:
+            context: .
+            push: true
+            tags: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{ steps.tags.outputs.primary_tag }}
+            build-args: |
+              GIT_SHA=${{ steps.tags.outputs.primary_tag }}
+
+pyproject.toml: |
+  [project]
+  name = "myapp"
+  version = "1.0.0"
+  requires-python = ">=3.11"
+
+  [project.scripts]
+  myapp = "myapp.cli:main"
+
+garbage_lint.py: |
+  GARBAGE_CICD016_LINT_MARKER_A = "lint module"
+  GARBAGE_CICD016_CHECK_MARKER_B = True
+
+  class LintRunner:
+      GARBAGE_CICD016_RUNNER_MARKER_C = "runner"
+
+      def run_checks(self, path: str) -> bool:
+          GARBAGE_CICD016_RESULT_MARKER_D = "passed"
+          return True
+
+      def format_report(self, results: list) -> str:
+          GARBAGE_CICD016_REPORT_MARKER_E = "report"
+          return ""
+
+garbage_migrate.py: |
+  GARBAGE_CICD016_MIGRATE_MARKER_F = "migrate module"
+  GARBAGE_CICD016_SCHEMA_MARKER_G = 42
+
+  class MigrationRunner:
+      GARBAGE_CICD016_MIGRATION_MARKER_H = "migration"
+
+      def apply(self, version: int) -> None:
+          GARBAGE_CICD016_APPLY_MARKER_I = "applied"
+          pass
+
+      def rollback(self, version: int) -> None:
+          GARBAGE_CICD016_ROLLBACK_MARKER_J = "rolled back"
+          pass