Add Rust JSON serialization/deserialization benchmark

Adds a benchmark testing JSON parsing and serialization performance using
serde_json. The benchmark processes ~1.3MB of JSON data representing 100
user records with nested structures (profiles, settings, posts).

Tests both deserialization (JSON → Rust structs) and serialization
(Rust structs → JSON) in a single benchmark cycle.

Author: posborne

benchmarks/rust-compression/benchmark.wasm

@@ -0,0 +1 @@
+../Dockerfile.rust

benchmarks/rust-html-rewriter/benchmark.wasm

@@ -0,0 +1,29 @@
+# Rust JSON Serialization/Deserialization Benchmark
+
+This benchmark tests JSON parsing and serialization performance using the `serde_json` crate, which is the de facto standard for JSON processing in Rust.
+
+## What it tests
+
+The benchmark performs both deserialization (JSON string → Rust structs) and serialization (Rust structs → JSON string) on a realistic dataset representing API response data with nested structures:
+
+- 100 user records
+- Each user has a profile, settings, and multiple posts
+- Nested data structures including hashmaps, vectors, and optional fields
+- Total input size: ~1.3 MB of JSON data
+
+## Input Data
+
+The `default.input` file contains 100 user records with:
+- User metadata (id, username, email, timestamps)
+- Profile information (bio, avatar, location, social links)
+- User settings (theme, language, notifications, preferences)
+- Multiple posts per user (5-15 posts with content, tags, likes)
+
+The input file can be regenerated using `generate_input.py`.
+
+## Implementation
+
+Uses:
+- `serde` 1.0 with derive macros for struct serialization
+- `serde_json` 1.0 for JSON processing
+- Measures full round-trip: parse JSON → serialize back to JSON

benchmarks/rust-json/Dockerfile

@@ -0,0 +1,2 @@
+[rust-json] processed 100 users
+[rust-json] serialized size: 1175589 bytes

benchmarks/rust-json/README.md

@@ -0,0 +1,62 @@
+#!/usr/bin/env python3
+"""Generate a realistic JSON input file for benchmarking."""
+
+import json
+import random
+from datetime import datetime, timedelta
+
+def generate_user(user_id):
+    """Generate a single user with realistic data."""
+    return {
+        "id": user_id,
+        "username": f"user_{user_id}",
+        "email": f"user{user_id}@example.com",
+        "full_name": f"User {user_id} Name",
+        "is_active": random.choice([True, False]),
+        "created_at": (datetime.now() - timedelta(days=random.randint(1, 1000))).isoformat(),
+        "updated_at": (datetime.now() - timedelta(days=random.randint(0, 100))).isoformat(),
+        "profile": {
+            "bio": f"This is a bio for user {user_id}. " * random.randint(1, 5),
+            "avatar_url": f"https://example.com/avatars/{user_id}.jpg",
+            "location": random.choice(["New York", "London", "Tokyo", "Berlin", "Paris", "Sydney"]),
+            "website": f"https://user{user_id}.example.com" if random.random() > 0.3 else None,
+            "social_links": [
+                f"https://twitter.com/user{user_id}",
+                f"https://github.com/user{user_id}",
+                f"https://linkedin.com/in/user{user_id}"
+            ]
+        },
+        "settings": {
+            "theme": random.choice(["light", "dark", "auto"]),
+            "language": random.choice(["en", "es", "fr", "de", "ja", "zh"]),
+            "notifications_enabled": random.choice([True, False]),
+            "privacy_level": random.choice(["public", "friends", "private"]),
+            "preferences": {
+                f"pref_{i}": f"value_{i}" for i in range(random.randint(3, 8))
+            }
+        },
+        "posts": [
+            {
+                "id": user_id * 1000 + post_id,
+                "title": f"Post {post_id} by user {user_id}",
+                "content": f"This is the content of post {post_id}. " * random.randint(10, 50),
+                "tags": [f"tag{random.randint(1, 20)}" for _ in range(random.randint(1, 5))],
+                "likes": random.randint(0, 1000),
+                "comments_count": random.randint(0, 100),
+                "published_at": (datetime.now() - timedelta(days=random.randint(0, 365))).isoformat()
+            }
+            for post_id in range(random.randint(5, 15))
+        ]
+    }
+
+def main():
+    # Generate 100 users for a reasonably sized benchmark
+    users = [generate_user(i) for i in range(1, 101)]
+
+    with open("default.input", "w") as f:
+        json.dump(users, f, indent=2)
+
+    print(f"Generated JSON with {len(users)} users")
+
+if __name__ == "__main__":
+    main()

benchmarks/rust-json/benchmark.stderr.expected

@@ -0,0 +1 @@
+target