Parse katanemo config using serde/yaml package (#6)

* Parse katanemo config using serde/yaml package

- load yaml file into typed classes
- pass katanemo config to plugin using envoy wasm plugin config
- add tests in configuration.rs file

envoyfilter/katanemo-config.yaml

@@ -0,0 +1,32 @@
+katanemo-prompt-config:
+  default-prompt-endpoint: "127.0.0.1"
+  load-balancing: "round-robin"
+  timeout-ms: 5000
+
+  embedding-provider:
+    name: "SentenceTransformer"
+    model: "all-MiniLM-L6-v2"
+
+  llm-providers:
+
+    - name: "open-ai-gpt-4"
+      api-key: "$OPEN_AI_API_KEY"
+      model: gpt-4
+
+  system-prompt: |
+    You are a helpful weather forecaster. Please following following guidelines when responding to user queries:
+    - Use farenheight for temperature
+    - Use miles per hour for wind speed
+
+  prompt-targets:
+
+    - type: context-resolver
+      name: weather-forecast
+      few-shot-examples:
+        - what is the weather in New York?
+      endpoint: "POST:$WEATHER_FORECAST_API_ENDPOINT"
+      cache-response: true
+      cache-response-settings:
+        - cache-ttl-secs: 3600 # cache expiry in seconds
+        - cache-max-size: 1000 # in number of items
+        - cache-eviction-strategy: LRU