Make common parse lenient (google-gemini#179)

Per [b/346812151](https://b.corp.google.com/issues/346812151),

This updates the json parser in common to be lenient in its decoding.
Otherwise, it can not properly decode json literals such as:
```json
{
   "result": true
}
```

A test has been added to track this change as well.

.changes/common/distribution-doll-drop-bridge.json

@@ -0,0 +1 @@
+{"type":"PATCH","changes":["Make JSON decoding lenient. This should fix issues with JSON literals."]}

common/src/main/kotlin/com/google/ai/client/generativeai/common/APIController.kt

@@ -52,6 +52,7 @@ import kotlinx.serialization.json.Json
 internal val JSON = Json {
   ignoreUnknownKeys = true
   prettyPrint = false
+  isLenient = true
 }
 
 /**

common/src/test/java/com/google/ai/client/generativeai/common/UnarySnapshotTests.kt

@@ -24,6 +24,7 @@ import com.google.ai.client.generativeai.common.shared.FunctionCallPart
 import com.google.ai.client.generativeai.common.shared.HarmCategory
 import com.google.ai.client.generativeai.common.shared.TextPart
 import com.google.ai.client.generativeai.common.util.goldenUnaryFile
+import com.google.ai.client.generativeai.common.util.shouldNotBeNullOrEmpty
 import io.kotest.assertions.throwables.shouldThrow
 import io.kotest.matchers.collections.shouldNotBeEmpty
 import io.kotest.matchers.nulls.shouldNotBeNull
@@ -313,4 +314,21 @@ internal class UnarySnapshotTests {
         callPart.functionCall.args["season"] shouldBe null
       }
     }
+
+  @Test
+  fun `function call contains json literal`() =
+    goldenUnaryFile("success-function-call-json-literal.json") {
+      withTimeout(testTimeout) {
+        val response = apiController.generateContent(textGenerateContentRequest("prompt"))
+        val content = response.candidates.shouldNotBeNullOrEmpty().first().content
+        val callPart =
+          content.let {
+            it.shouldNotBeNull()
+            it.parts.shouldNotBeEmpty()
+            it.parts.first().shouldBeInstanceOf<FunctionCallPart>()
+          }
+
+        callPart.functionCall.args["current"] shouldBe "true"
+      }
+    }
 }

common/src/test/kotlin/dev/shreyaspatil/ai/client/generativeai/common/util/tests.kt

@@ -25,6 +25,8 @@ import dev.shreyaspatil.ai.client.generativeai.common.RequestOptions
 import dev.shreyaspatil.ai.client.generativeai.common.server.Candidate
 import dev.shreyaspatil.ai.client.generativeai.common.shared.Content
 import dev.shreyaspatil.ai.client.generativeai.common.shared.TextPart
+import io.kotest.matchers.collections.shouldNotBeEmpty
+import io.kotest.matchers.nulls.shouldNotBeNull
 import io.ktor.client.engine.mock.MockEngine
 import io.ktor.client.engine.mock.respond
 import io.ktor.http.HttpHeaders
@@ -183,3 +185,14 @@ internal fun loadGoldenFile(path: String): File = loadResourceFile("golden-files
 
 /** Loads a file from the test resources directory. */
 internal fun loadResourceFile(path: String) = File("src/test/resources/$path")
+
+/**
+ * Ensures that a collection is neither null or empty.
+ *
+ * Syntax sugar for [shouldNotBeNull] and [shouldNotBeEmpty].
+ */
+inline fun <reified T : Any> Collection<T>?.shouldNotBeNullOrEmpty(): Collection<T> {
+  shouldNotBeNull()
+  shouldNotBeEmpty()
+  return this
+}

common/src/test/resources/golden-files/unary/success-function-call-json-literal.json

@@ -0,0 +1,45 @@
+{
+  "candidates": [
+    {
+      "content": {
+        "parts": [
+          {
+            "functionCall": {
+              "name": "functionName",
+              "args": {
+                "original_title": "String",
+                "current": true
+              }
+            }
+          }
+        ],
+        "role": "model"
+      },
+      "finishReason": "STOP",
+      "index": 0,
+      "safetyRatings": [
+        {
+          "category": "HARM_CATEGORY_HARASSMENT",
+          "probability": "NEGLIGIBLE"
+        },
+        {
+          "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT",
+          "probability": "NEGLIGIBLE"
+        },
+        {
+          "category": "HARM_CATEGORY_DANGEROUS_CONTENT",
+          "probability": "NEGLIGIBLE"
+        },
+        {
+          "category": "HARM_CATEGORY_HATE_SPEECH",
+          "probability": "NEGLIGIBLE"
+        }
+      ]
+    }
+  ],
+  "usageMetadata": {
+    "promptTokenCount": 774,
+    "candidatesTokenCount": 4176,
+    "totalTokenCount": 4950
+  }
+}