1.文章介绍

Retrofit2+Rxjava2+Glide+MVP架构正是火热，以学习角度出发，分模块来熟悉下这种架构的各个模块源码以及实现方式。
什么情况下使用Retrofit2+Rxjava2呢？Retrofit2+Rxjava2的主要优势是它的解耦性，在项目越复杂时体现出的优势越明显。

2.干货

继续上一篇的分析，在本次分析中创建的是ExecutorCallAdapterFactory：

final class ExecutorCallAdapterFactory extends CallAdapter.Factory {
  final Executor callbackExecutor;

  ExecutorCallAdapterFactory(Executor callbackExecutor) {
    this.callbackExecutor = callbackExecutor;
  }

  @Override public @Nullable CallAdapter<?, ?> get(
      Type returnType, Annotation[] annotations, Retrofit retrofit) {
    // returnType在文章之前分析过 是Call<DataModel> ，getRawType返回是Call
    if (getRawType(returnType) != Call.class) {
      return null;
    }
    //这里根据returnType本文返回为Call<DataModel> ，那么responseType就是DataModel
    final Type responseType = Utils.getCallResponseType(returnType);
    return new CallAdapter<Object, Call<?>>() {
      @Override public Type responseType() {
        return responseType;
      }

      @Override public Call<Object> adapt(Call<Object> call) {
        return new ExecutorCallbackCall<>(callbackExecutor, call);
      }
    };
  }
  ...
}

传入的参数是Call<DataModel>，以下相当于通过反射获取泛型的参数DataModel：

 static Type getCallResponseType(Type returnType) {
    if (!(returnType instanceof ParameterizedType)) {
      throw new IllegalArgumentException(
          "Call return type must be parameterized as Call<Foo> or Call<? extends Foo>");
    }
    return getParameterUpperBound(0, (ParameterizedType) returnType);
  }

  static Type getParameterUpperBound(int index, ParameterizedType type) {
    Type[] types = type.getActualTypeArguments();
    if (index < 0 || index >= types.length) {
      throw new IllegalArgumentException(
          "Index " + index + " not in range [0," + types.length + ") for " + type);
    }
    Type paramType = types[index];
    if (paramType instanceof WildcardType) {
      return ((WildcardType) paramType).getUpperBounds()[0];
    }
    return paramType;
  }

可以看到在Retrofit中，如果不主动添加CallAdapter，那么声明的方法的返回值必须是Call类。这里本应该分析到了adapt，把数据转换成这个Call类型，这个函数在文章后面会用到这里先不分析。
CallAdapter返回的参数responseType非常重要，就是泛型的实际类型，后面也会用得上。

网络请求适配后，还需要有数据格式的适配：

 //这里就把responseType用上了，记住是返回的实际类型，本文是DataModel
 Converter<ResponseBody, ResponseT> responseConverter =
        createResponseConverter(retrofit, method, responseType);

  private static <ResponseT> Converter<ResponseBody, ResponseT> createResponseConverter(
      Retrofit retrofit, Method method, Type responseType) {
    Annotation[] annotations = method.getAnnotations();
    try {
      return retrofit.responseBodyConverter(responseType, annotations);
    } catch (RuntimeException e) { // Wide exception range because factories are user code.
      throw methodError(method, e, "Unable to create converter for %s", responseType);
    }
  }

  public <T> Converter<ResponseBody, T> responseBodyConverter(Type type, Annotation[] annotations) {
    return nextResponseBodyConverter(null, type, annotations);
  }

  public <T> Converter<ResponseBody, T> nextResponseBodyConverter(
      @Nullable Converter.Factory skipPast, Type type, Annotation[] annotations) {
    checkNotNull(type, "type == null");
    checkNotNull(annotations, "annotations == null");

    int start = converterFactories.indexOf(skipPast) + 1;
    for (int i = start, count = converterFactories.size(); i < count; i++) {
      Converter<ResponseBody, ?> converter =
          converterFactories.get(i).responseBodyConverter(type, annotations, this);
      if (converter != null) {
        //noinspection unchecked
        return (Converter<ResponseBody, T>) converter;
      }
    }

    StringBuilder builder = new StringBuilder("Could not locate ResponseBody converter for ")
        .append(type)
        .append(".\n");
    if (skipPast != null) {
      builder.append("  Skipped:");
      for (int i = 0; i < start; i++) {
        builder.append("\n   * ").append(converterFactories.get(i).getClass().getName());
      }
      builder.append('\n');
    }
    builder.append("  Tried:");
    for (int i = start, count = converterFactories.size(); i < count; i++) {
      builder.append("\n   * ").append(converterFactories.get(i).getClass().getName());
    }
    throw new IllegalArgumentException(builder.toString());
  }

和CallAdapter类似的处理，在未声明Converter时，默认有一个new BuiltInConverters()和platform.defaultConverterFactories()：

  List<? extends Converter.Factory> defaultConverterFactories() {
    return emptyList();
  }

defaultConverterFactories返回的是一个空的List，那么BuiltInConverters是否是默认使用的呢?

final class BuiltInConverters extends Converter.Factory {
  /** Not volatile because we don't mind multiple threads discovering this. */
  private boolean checkForKotlinUnit = true;

  @Override public @Nullable Converter<ResponseBody, ?> responseBodyConverter(
      Type type, Annotation[] annotations, Retrofit retrofit) {
    //根据之前的分析 这里是Call泛型的实际类型 本文是DataModel
    if (type == ResponseBody.class) {
      return Utils.isAnnotationPresent(annotations, Streaming.class)
          ? StreamingResponseBodyConverter.INSTANCE
          : BufferingResponseBodyConverter.INSTANCE;
    }
    if (type == Void.class) {
      return VoidResponseBodyConverter.INSTANCE;
    }
    if (checkForKotlinUnit) {
      try {
        if (type == Unit.class) {
          return UnitResponseBodyConverter.INSTANCE;
        }
      } catch (NoClassDefFoundError ignored) {
        checkForKotlinUnit = false;
      }
    }
    return null;
  }
  ...
}

这里可以看到如果不配置Converter时只能对ResponseBody和Void做处理，显然本文的Demo必须要声明Converter能处理自定义的Class类型。
在本文中通过Gson把json数据映射为对象，按如下配置：

addConverterFactory(GsonConverterFactory.create())

public final class GsonConverterFactory extends Converter.Factory {
  /**
   * Create an instance using a default {@link Gson} instance for conversion. Encoding to JSON and
   * decoding from JSON (when no charset is specified by a header) will use UTF-8.
   */
  public static GsonConverterFactory create() {
    return create(new Gson());
  }

  /**
   * Create an instance using {@code gson} for conversion. Encoding to JSON and
   * decoding from JSON (when no charset is specified by a header) will use UTF-8.
   */
  @SuppressWarnings("ConstantConditions") // Guarding public API nullability.
  public static GsonConverterFactory create(Gson gson) {
    if (gson == null) throw new NullPointerException("gson == null");
    return new GsonConverterFactory(gson);
  }

  private final Gson gson;

  private GsonConverterFactory(Gson gson) {
    this.gson = gson;
  }
  ...
}

GsonConverterFactory就是对Gson的封装处理，数据格式的适配函数如下：

  @Override
  public Converter<ResponseBody, ?> responseBodyConverter(Type type, Annotation[] annotations,
      Retrofit retrofit) {
    TypeAdapter<?> adapter = gson.getAdapter(TypeToken.get(type));
    return new GsonResponseBodyConverter<>(gson, adapter);
  }

  public <T> TypeAdapter<T> getAdapter(TypeToken<T> type) {
    TypeAdapter<?> cached = typeTokenCache.get(type == null ? NULL_KEY_SURROGATE : type);
    if (cached != null) {
      return (TypeAdapter<T>) cached;
    }

    Map<TypeToken<?>, FutureTypeAdapter<?>> threadCalls = calls.get();
    boolean requiresThreadLocalCleanup = false;
    if (threadCalls == null) {
      threadCalls = new HashMap<TypeToken<?>, FutureTypeAdapter<?>>();
      calls.set(threadCalls);
      requiresThreadLocalCleanup = true;
    }

    // the key and value type parameters always agree
    FutureTypeAdapter<T> ongoingCall = (FutureTypeAdapter<T>) threadCalls.get(type);
    if (ongoingCall != null) {
      return ongoingCall;
    }

    try {
      FutureTypeAdapter<T> call = new FutureTypeAdapter<T>();
      threadCalls.put(type, call);

      for (TypeAdapterFactory factory : factories) {
        TypeAdapter<T> candidate = factory.create(this, type);
        if (candidate != null) {
          call.setDelegate(candidate);
          typeTokenCache.put(type, candidate);
          return candidate;
        }
      }
      throw new IllegalArgumentException("GSON cannot handle " + type);
    } finally {
      threadCalls.remove(type);

      if (requiresThreadLocalCleanup) {
        calls.remove();
      }
    }
  }

从上面的函数可以看到，从factories中去查找匹配的TypeAdapter。返回去看这个函数：

TypeAdapter<?> adapter = gson.getAdapter(TypeToken.get(type));

这里涉及到TypeToken：

  public static TypeToken<?> get(Type type) {
    return new TypeToken<Object>(type);
  }

  TypeToken(Type type) {
    //这里传入的参数是Call泛型的实际类型，本文是DataModel
    this.type = $Gson$Types.canonicalize($Gson$Preconditions.checkNotNull(type));
    this.rawType = (Class<? super T>) $Gson$Types.getRawType(this.type);
    this.hashCode = this.type.hashCode();
  }

  public static Type canonicalize(Type type) {
    // 很显然，本文可以通过这个TypeAdapter来转完成数据适配，这里type在本文中是DataModel
    if (type instanceof Class) {//原始类型
      Class<?> c = (Class<?>) type;
      return c.isArray() ? new GenericArrayTypeImpl(canonicalize(c.getComponentType())) : c;

    } else if (type instanceof ParameterizedType) {//参数化类型
      ParameterizedType p = (ParameterizedType) type;
      return new ParameterizedTypeImpl(p.getOwnerType(),
          p.getRawType(), p.getActualTypeArguments());

    } else if (type instanceof GenericArrayType) {//数组类型 
      GenericArrayType g = (GenericArrayType) type;
      return new GenericArrayTypeImpl(g.getGenericComponentType());

    } else if (type instanceof WildcardType) {//通配符类型
      WildcardType w = (WildcardType) type;
      return new WildcardTypeImpl(w.getUpperBounds(), w.getLowerBounds());

    } else {
      // type is either serializable as-is or unsupported
      return type;
    }
  }

  public static Class<?> getRawType(Type type) {
    // 很显然，本文直接返回DataModel
    if (type instanceof Class<?>) {
      // type is a normal class.
      return (Class<?>) type;

    } else if (type instanceof ParameterizedType) {
      ParameterizedType parameterizedType = (ParameterizedType) type;

      // I'm not exactly sure why getRawType() returns Type instead of Class.
      // Neal isn't either but suspects some pathological case related
      // to nested classes exists.
      Type rawType = parameterizedType.getRawType();
      checkArgument(rawType instanceof Class);
      return (Class<?>) rawType;

    } else if (type instanceof GenericArrayType) {
      Type componentType = ((GenericArrayType)type).getGenericComponentType();
      return Array.newInstance(getRawType(componentType), 0).getClass();

    } else if (type instanceof TypeVariable) {
      // we could use the variable's bounds, but that won't work if there are multiple.
      // having a raw type that's more general than necessary is okay
      return Object.class;

    } else if (type instanceof WildcardType) {
      return getRawType(((WildcardType) type).getUpperBounds()[0]);

    } else {
      String className = type == null ? "null" : type.getClass().getName();
      throw new IllegalArgumentException("Expected a Class, ParameterizedType, or "
          + "GenericArrayType, but <" + type + "> is of type " + className);
    }
  }

在Gson初始化时添加了各种TypeAdapters：

Gson(final Excluder excluder, final FieldNamingStrategy fieldNamingStrategy,
      final Map<Type, InstanceCreator<?>> instanceCreators, boolean serializeNulls,
      boolean complexMapKeySerialization, boolean generateNonExecutableGson, boolean htmlSafe,
      boolean prettyPrinting, boolean lenient, boolean serializeSpecialFloatingPointValues,
      LongSerializationPolicy longSerializationPolicy,
      List<TypeAdapterFactory> typeAdapterFactories) {
    this.constructorConstructor = new ConstructorConstructor(instanceCreators);
    this.excluder = excluder;
    this.fieldNamingStrategy = fieldNamingStrategy;
    this.serializeNulls = serializeNulls;
    this.generateNonExecutableJson = generateNonExecutableGson;
    this.htmlSafe = htmlSafe;
    this.prettyPrinting = prettyPrinting;
    this.lenient = lenient;

    List<TypeAdapterFactory> factories = new ArrayList<TypeAdapterFactory>();

    // built-in type adapters that cannot be overridden
    factories.add(TypeAdapters.JSON_ELEMENT_FACTORY);
    factories.add(ObjectTypeAdapter.FACTORY);

    // the excluder must precede all adapters that handle user-defined types
    factories.add(excluder);

    // user's type adapters
    factories.addAll(typeAdapterFactories);

    // type adapters for basic platform types
    factories.add(TypeAdapters.STRING_FACTORY);
    factories.add(TypeAdapters.INTEGER_FACTORY);
    factories.add(TypeAdapters.BOOLEAN_FACTORY);
    factories.add(TypeAdapters.BYTE_FACTORY);
    factories.add(TypeAdapters.SHORT_FACTORY);
    TypeAdapter<Number> longAdapter = longAdapter(longSerializationPolicy);
    factories.add(TypeAdapters.newFactory(long.class, Long.class, longAdapter));
    factories.add(TypeAdapters.newFactory(double.class, Double.class,
            doubleAdapter(serializeSpecialFloatingPointValues)));
    factories.add(TypeAdapters.newFactory(float.class, Float.class,
            floatAdapter(serializeSpecialFloatingPointValues)));
    factories.add(TypeAdapters.NUMBER_FACTORY);
    factories.add(TypeAdapters.ATOMIC_INTEGER_FACTORY);
    factories.add(TypeAdapters.ATOMIC_BOOLEAN_FACTORY);
    factories.add(TypeAdapters.newFactory(AtomicLong.class, atomicLongAdapter(longAdapter)));
    factories.add(TypeAdapters.newFactory(AtomicLongArray.class, atomicLongArrayAdapter(longAdapter)));
    factories.add(TypeAdapters.ATOMIC_INTEGER_ARRAY_FACTORY);
    factories.add(TypeAdapters.CHARACTER_FACTORY);
    factories.add(TypeAdapters.STRING_BUILDER_FACTORY);
    factories.add(TypeAdapters.STRING_BUFFER_FACTORY);
    factories.add(TypeAdapters.newFactory(BigDecimal.class, TypeAdapters.BIG_DECIMAL));
    factories.add(TypeAdapters.newFactory(BigInteger.class, TypeAdapters.BIG_INTEGER));
    factories.add(TypeAdapters.URL_FACTORY);
    factories.add(TypeAdapters.URI_FACTORY);
    factories.add(TypeAdapters.UUID_FACTORY);
    factories.add(TypeAdapters.CURRENCY_FACTORY);
    factories.add(TypeAdapters.LOCALE_FACTORY);
    factories.add(TypeAdapters.INET_ADDRESS_FACTORY);
    factories.add(TypeAdapters.BIT_SET_FACTORY);
    factories.add(DateTypeAdapter.FACTORY);
    factories.add(TypeAdapters.CALENDAR_FACTORY);
    factories.add(TimeTypeAdapter.FACTORY);
    factories.add(SqlDateTypeAdapter.FACTORY);
    factories.add(TypeAdapters.TIMESTAMP_FACTORY);
    factories.add(ArrayTypeAdapter.FACTORY);
    factories.add(TypeAdapters.CLASS_FACTORY);

    // type adapters for composite and user-defined types
    factories.add(new CollectionTypeAdapterFactory(constructorConstructor));
    factories.add(new MapTypeAdapterFactory(constructorConstructor, complexMapKeySerialization));
    this.jsonAdapterFactory = new JsonAdapterAnnotationTypeAdapterFactory(constructorConstructor);
    factories.add(jsonAdapterFactory);
    factories.add(TypeAdapters.ENUM_FACTORY);
    factories.add(new ReflectiveTypeAdapterFactory(
        constructorConstructor, fieldNamingStrategy, excluder, jsonAdapterFactory));

    this.factories = Collections.unmodifiableList(factories);
  }

查找匹配的TypeAdapters核心代码如下：

for (TypeAdapterFactory factory : factories) {
   TypeAdapter<T> candidate = factory.create(this, type);
   if (candidate != null) {
     call.setDelegate(candidate);
     typeTokenCache.put(type, candidate);
     return candidate;
   }
}

只有在TypeAdapter的create到匹配type时，才会使用该TypeAdapter。注意这里的type是声明泛型的实际类型，按本文的分析就是DataModel。
对factories举个例子，例如：

   factories.add(TypeAdapters.JSON_ELEMENT_FACTORY);

   public static final TypeAdapterFactory JSON_ELEMENT_FACTORY
      = newTypeHierarchyFactory(JsonElement.class, JSON_ELEMENT);

   /**
   * Returns a factory for all subtypes of {@code typeAdapter}. We do a runtime check to confirm
   * that the deserialized type matches the type requested.
   */
  public static <T1> TypeAdapterFactory newTypeHierarchyFactory(
      final Class<T1> clazz, final TypeAdapter<T1> typeAdapter) {
    return new TypeAdapterFactory() {
      @SuppressWarnings("unchecked")
      @Override public <T2> TypeAdapter<T2> create(Gson gson, TypeToken<T2> typeToken) {
        //从TypeToken源码可知，这里返回的就是实际类型，本文的DataModel
        final Class<? super T2> requestedType = typeToken.getRawType();
        if (!clazz.isAssignableFrom(requestedType)) {
          return null;
        }
        return (TypeAdapter<T2>) new TypeAdapter<T1>() {
          @Override public void write(JsonWriter out, T1 value) throws IOException {
            typeAdapter.write(out, value);
          }

          @Override public T1 read(JsonReader in) throws IOException {
            T1 result = typeAdapter.read(in);
            if (result != null && !requestedType.isInstance(result)) {
              throw new JsonSyntaxException("Expected a " + requestedType.getName()
                  + " but was " + result.getClass().getName());
            }
            return result;
          }
        };
      }
      @Override public String toString() {
        return "Factory[typeHierarchy=" + clazz.getName() + ",adapter=" + typeAdapter + "]";
      }
    };
  }

通过isAssignableFrom接口判断是否可以使用JSON_ELEMENT_FACTORY。很明显这里并不能，那么会使用哪个TypeAdapterFactory呢？继续分析，我们看到最后有一个ReflectiveTypeAdapterFactory：

  factories.add(new ReflectiveTypeAdapterFactory(
        constructorConstructor, fieldNamingStrategy, excluder, jsonAdapterFactory));

  public ReflectiveTypeAdapterFactory(ConstructorConstructor constructorConstructor,
      FieldNamingStrategy fieldNamingPolicy, Excluder excluder,
      JsonAdapterAnnotationTypeAdapterFactory jsonAdapterFactory) {
    this.constructorConstructor = constructorConstructor;
    this.fieldNamingPolicy = fieldNamingPolicy;
    this.excluder = excluder;
    this.jsonAdapterFactory = jsonAdapterFactory;
  }

  @Override public <T> TypeAdapter<T> create(Gson gson, final TypeToken<T> type) {
    Class<? super T> raw = type.getRawType();

    if (!Object.class.isAssignableFrom(raw)) {
      return null; // it's a primitive!
    }

    ObjectConstructor<T> constructor = constructorConstructor.get(type);
    return new Adapter<T>(constructor, getBoundFields(gson, type, raw));
  }

  private Map<String, BoundField> getBoundFields(Gson context, TypeToken<?> type, Class<?> raw) {
    Map<String, BoundField> result = new LinkedHashMap<String, BoundField>();
    if (raw.isInterface()) {
      return result;
    }

    Type declaredType = type.getType();
    while (raw != Object.class) {
      Field[] fields = raw.getDeclaredFields();
      for (Field field : fields) {
        boolean serialize = excludeField(field, true);
        boolean deserialize = excludeField(field, false);
        if (!serialize && !deserialize) {
          continue;
        }
        field.setAccessible(true);
        Type fieldType = $Gson$Types.resolve(type.getType(), raw, field.getGenericType());
        List<String> fieldNames = getFieldNames(field);
        BoundField previous = null;
        for (int i = 0, size = fieldNames.size(); i < size; ++i) {
          String name = fieldNames.get(i);
          if (i != 0) serialize = false; // only serialize the default name
          BoundField boundField = createBoundField(context, field, name,
              TypeToken.get(fieldType), serialize, deserialize);
          BoundField replaced = result.put(name, boundField);
          if (previous == null) previous = replaced;
        }
        if (previous != null) {
          throw new IllegalArgumentException(declaredType
              + " declares multiple JSON fields named " + previous.name);
        }
      }
      type = TypeToken.get($Gson$Types.resolve(type.getType(), raw, raw.getGenericSuperclass()));
      raw = type.getRawType();
    }
    return result;
  }

以上代码通过反射动态实现对Class类的映射，然后在read、write方法里面拆、组装数据：

public static final class Adapter<T> extends TypeAdapter<T> {
    private final ObjectConstructor<T> constructor;
    private final Map<String, BoundField> boundFields;

    Adapter(ObjectConstructor<T> constructor, Map<String, BoundField> boundFields) {
      this.constructor = constructor;
      this.boundFields = boundFields;
    }

    @Override public T read(JsonReader in) throws IOException {
      if (in.peek() == JsonToken.NULL) {
        in.nextNull();
        return null;
      }

      T instance = constructor.construct();

      try {
        in.beginObject();
        while (in.hasNext()) {
          String name = in.nextName();
          BoundField field = boundFields.get(name);
          if (field == null || !field.deserialized) {
            in.skipValue();
          } else {
            field.read(in, instance);
          }
        }
      } catch (IllegalStateException e) {
        throw new JsonSyntaxException(e);
      } catch (IllegalAccessException e) {
        throw new AssertionError(e);
      }
      in.endObject();
      return instance;
    }

    @Override public void write(JsonWriter out, T value) throws IOException {
      if (value == null) {
        out.nullValue();
        return;
      }

      out.beginObject();
      try {
        for (BoundField boundField : boundFields.values()) {
          if (boundField.writeField(value)) {
            out.name(boundField.name);
            boundField.write(out, value);
          }
        }
      } catch (IllegalAccessException e) {
        throw new AssertionError(e);
      }
      out.endObject();
    }
  }

Gson就是这样动态实现了数据适配。

当调用了loadWeatherInfo(***)这个接口后，经过动态代理继而中转到HttpServiceMethod：

  private HttpServiceMethod(RequestFactory requestFactory, okhttp3.Call.Factory callFactory,
      CallAdapter<ResponseT, ReturnT> callAdapter,
      Converter<ResponseBody, ResponseT> responseConverter) {
    this.requestFactory = requestFactory;
    this.callFactory = callFactory;
    this.callAdapter = callAdapter;
    this.responseConverter = responseConverter;
  }

  @Override ReturnT invoke(Object[] args) {
    return callAdapter.adapt(
        new OkHttpCall<>(requestFactory, args, callFactory, responseConverter));
  }

最后我们看到实际上是调用到我们之前配置的callAdapter的adapt接口，之前我还特意标注了一下adapt这个接口是作什么用，刚好这里给出解释了。
回去再看看adapt函数干什么了：

static final class ExecutorCallbackCall<T> implements Call<T> {
    final Executor callbackExecutor;
    final Call<T> delegate;

    ExecutorCallbackCall(Executor callbackExecutor, Call<T> delegate) {
      this.callbackExecutor = callbackExecutor;
      this.delegate = delegate;
    }

    @Override public void enqueue(final Callback<T> callback) {
      checkNotNull(callback, "callback == null");

      delegate.enqueue(new Callback<T>() {
        @Override public void onResponse(Call<T> call, final Response<T> response) {
          callbackExecutor.execute(new Runnable() {
            @Override public void run() {
              if (delegate.isCanceled()) {
                // Emulate OkHttp's behavior of throwing/delivering an IOException on cancellation.
                callback.onFailure(ExecutorCallbackCall.this, new IOException("Canceled"));
              } else {
                callback.onResponse(ExecutorCallbackCall.this, response);
              }
            }
          });
        }

        @Override public void onFailure(Call<T> call, final Throwable t) {
          callbackExecutor.execute(new Runnable() {
            @Override public void run() {
              callback.onFailure(ExecutorCallbackCall.this, t);
            }
          });
        }
      });
    }

    @Override public boolean isExecuted() {
      return delegate.isExecuted();
    }

    @Override public Response<T> execute() throws IOException {
      return delegate.execute();
    }

    @Override public void cancel() {
      delegate.cancel();
    }

    @Override public boolean isCanceled() {
      return delegate.isCanceled();
    }

    @SuppressWarnings("CloneDoesntCallSuperClone") // Performing deep clone.
    @Override public Call<T> clone() {
      return new ExecutorCallbackCall<>(callbackExecutor, delegate.clone());
    }

    @Override public Request request() {
      return delegate.request();
    }
  }

callbackExecutor.execute其实就是Handler.post，在ExecutorCallAdapterFactory的adapt方法里把Call代理给ExecutorCallbackCall了，ExecutorCallbackCall可以理解为Handler+Call的封装装，它的enqueue实际就是OkHttpCall.enqueue后把response通过主线程的Handler.post回调给应用。

至于为什么使用Handler，我自己的理解是：
1.借用Handler的消息队列来对任务排序
2.有某方法需要回调到主线程
至少我目前只想到这两个用途，因为真正异步涉及到的线程调度都在OkHttpCall内部的线程池了。

        Call<DataModel> call = RetrofitUtils.getDeafault(WeatherService.WEATHERSERVICE_URL)
                .create(WeatherService.class)
                .loadWeatherInfo("101190201");
        call.enqueue(new Callback<DataModel>() {
            @Override
            public void onResponse(Call<DataModel> call, Response<DataModel> response) {
                DataModel model = response.body();
                DataModel.WeatherInfoBean bean = model.getWeatherinfo();
                Log.d(TAG, "onResponse:" + bean.getCity());
            }

            @Override
            public void onFailure(Call<DataModel> call, Throwable t) {
                Log.d(TAG, "onFailure:" + t.getMessage());
            }
        });

以上源码中的call实际上是一个ExecutorCallbackCall。
而call.enqueue实际上就是OkHttpCall.enqueue处理，根据继承多态的原理：

  @Override public void enqueue(final Callback<T> callback) {
    checkNotNull(callback, "callback == null");

    okhttp3.Call call;
    Throwable failure;

    synchronized (this) {
      if (executed) throw new IllegalStateException("Already executed.");
      executed = true;

      call = rawCall;
      failure = creationFailure;
      if (call == null && failure == null) {
        try {
          call = rawCall = createRawCall();
        } catch (Throwable t) {
          throwIfFatal(t);
          failure = creationFailure = t;
        }
      }
    }

    if (failure != null) {
      callback.onFailure(this, failure);
      return;
    }

    if (canceled) {
      call.cancel();
    }

    call.enqueue(new okhttp3.Callback() {
      @Override public void onResponse(okhttp3.Call call, okhttp3.Response rawResponse) {
        Response<T> response;
        try {
          response = parseResponse(rawResponse);
        } catch (Throwable e) {
          throwIfFatal(e);
          callFailure(e);
          return;
        }

        try {
          callback.onResponse(OkHttpCall.this, response);
        } catch (Throwable t) {
          t.printStackTrace();
        }
      }

      @Override public void onFailure(okhttp3.Call call, IOException e) {
        callFailure(e);
      }

      private void callFailure(Throwable e) {
        try {
          callback.onFailure(OkHttpCall.this, e);
        } catch (Throwable t) {
          t.printStackTrace();
        }
      }
    });
  }

Retrofit2.0默认支持okhttp，可以看到Retrofit的enqueue实际上是依赖okhttp的enqueue实现的。通过okhttp创建了一个Call，并且把实际的网络请求让okhttp的队列去处理，最后把网络请求的结果通过callback返回给调用者。这里也可以感受到写代码不同的是，要么是在函数里面直接执行业务返回数据的同步方式，要么是先注册CallBack然后请求数据通过callback把数据回调回来的异步方式，而Retrofit是虽然调用了某个接口(比如本文的loadWeatherInfo)，但是并没有实际的去执行某个业务(比如从网络上去请求数据)，而是在execute或者enqueue的时候才真正的去执行实际业务(比如真正根据配置去创建一个网络请求)。

  private okhttp3.Call createRawCall() throws IOException {
    okhttp3.Call call = callFactory.newCall(requestFactory.create(args));
    if (call == null) {
      throw new NullPointerException("Call.Factory returned null.");
    }
    return call;
  }

requestFactory.create(args)就是创建一个Request，但是callFactory是在哪创建的呢？细心的人可能会在文章前面找到：

public Retrofit build() {
      if (baseUrl == null) {
        throw new IllegalStateException("Base URL required.");
      }

      okhttp3.Call.Factory callFactory = this.callFactory;
      if (callFactory == null) {
        callFactory = new OkHttpClient();
      }
      ...
}

如果在Retrofit初始化时没有创建callFactory，就会默认使用OkHttpClient，这种情况的函数流程如下：

  @Override public Call newCall(Request request) {
    return RealCall.newRealCall(this, request, false /* for web socket */);
  }

  private RealCall(OkHttpClient client, Request originalRequest, boolean forWebSocket) {
    this.client = client;
    this.originalRequest = originalRequest;
    this.forWebSocket = forWebSocket;
    this.retryAndFollowUpInterceptor = new RetryAndFollowUpInterceptor(client, forWebSocket);
    this.timeout = new AsyncTimeout() {
      @Override protected void timedOut() {
        cancel();
      }
    };
    this.timeout.timeout(client.callTimeoutMillis(), MILLISECONDS);
  }

  static RealCall newRealCall(OkHttpClient client, Request originalRequest, boolean forWebSocket) {
    // Safely publish the Call instance to the EventListener.
    RealCall call = new RealCall(client, originalRequest, forWebSocket);
    call.eventListener = client.eventListenerFactory().create(call);
    return call;
  }

那么在的OkHttpCall里面的call.enqueue调用在继承多态的特性下就是如下调用：

  @Override public void enqueue(Callback responseCallback) {
    synchronized (this) {
      if (executed) throw new IllegalStateException("Already Executed");
      executed = true;
    }
    captureCallStackTrace();
    // 很简单的事件回调 可以看出这个是在调用线程
    eventListener.callStart(this);
    // 真正的实现异步处理
    client.dispatcher().enqueue(new AsyncCall(responseCallback));
  }

揭开异步机制的面纱:

  void enqueue(AsyncCall call) {
    synchronized (this) {
      readyAsyncCalls.add(call);
    }
    promoteAndExecute();
  }

  private boolean promoteAndExecute() {
    assert (!Thread.holdsLock(this));

    List<AsyncCall> executableCalls = new ArrayList<>();
    boolean isRunning;
    synchronized (this) {
      for (Iterator<AsyncCall> i = readyAsyncCalls.iterator(); i.hasNext(); ) {
        AsyncCall asyncCall = i.next();

        if (runningAsyncCalls.size() >= maxRequests) break; // Max capacity.
        if (runningCallsForHost(asyncCall) >= maxRequestsPerHost) continue; // Host max capacity.

        i.remove();
        executableCalls.add(asyncCall);
        runningAsyncCalls.add(asyncCall);
      }
      isRunning = runningCallsCount() > 0;
    }

    for (int i = 0, size = executableCalls.size(); i < size; i++) {
      AsyncCall asyncCall = executableCalls.get(i);
      asyncCall.executeOn(executorService());
    }

    return isRunning;
  }

Dispatcher默认的线程池：

  public synchronized ExecutorService executorService() {
    if (executorService == null) {
      executorService = new ThreadPoolExecutor(0, Integer.MAX_VALUE, 60, TimeUnit.SECONDS,
          new SynchronousQueue<Runnable>(), Util.threadFactory("OkHttp Dispatcher", false));
    }
    return executorService;
  }

在Dispatcher内部先统计了正在执行的异步任务，来确保执行效率，然后把任务放在了线程池里执行。

    void executeOn(ExecutorService executorService) {
      assert (!Thread.holdsLock(client.dispatcher()));
      boolean success = false;
      try {
        executorService.execute(this);
        success = true;
      } catch (RejectedExecutionException e) {
        InterruptedIOException ioException = new InterruptedIOException("executor rejected");
        ioException.initCause(e);
        eventListener.callFailed(RealCall.this, ioException);
        responseCallback.onFailure(RealCall.this, ioException);
      } finally {
        if (!success) {
          client.dispatcher().finished(this); // This call is no longer running!
        }
      }
    }

AsyncCall继承于NamedRunnable：

  final class AsyncCall extends NamedRunnable {
    private final Callback responseCallback;

    AsyncCall(Callback responseCallback) {
      super("OkHttp %s", redactedUrl());
      this.responseCallback = responseCallback;
    }
    ...
    @Override protected void execute() {
      boolean signalledCallback = false;
      timeout.enter();
      try {
        Response response = getResponseWithInterceptorChain();
        if (retryAndFollowUpInterceptor.isCanceled()) {
          signalledCallback = true;
          responseCallback.onFailure(RealCall.this, new IOException("Canceled"));
        } else {
          signalledCallback = true;
          responseCallback.onResponse(RealCall.this, response);
        }
      } catch (IOException e) {
        e = timeoutExit(e);
        if (signalledCallback) {
          // Do not signal the callback twice!
          Platform.get().log(INFO, "Callback failure for " + toLoggableString(), e);
        } else {
          eventListener.callFailed(RealCall.this, e);
          responseCallback.onFailure(RealCall.this, e);
        }
      } finally {
        client.dispatcher().finished(this);
      }
    }
}

public abstract class NamedRunnable implements Runnable {
  protected final String name;

  public NamedRunnable(String format, Object... args) {
    this.name = Util.format(format, args);
  }

  @Override public final void run() {
    String oldName = Thread.currentThread().getName();
    Thread.currentThread().setName(name);
    try {
      execute();
    } finally {
      Thread.currentThread().setName(oldName);
    }
  }

  protected abstract void execute();
}

可以看到实际处理是调用execute，根据Java的继承多态特性，会调用到AsyncCall的execute，这里已经能看到数据的请求了，来看一下如何实现的：

  Response getResponseWithInterceptorChain() throws IOException {
    // Build a full stack of interceptors.
    List<Interceptor> interceptors = new ArrayList<>();
    interceptors.addAll(client.interceptors());
    interceptors.add(retryAndFollowUpInterceptor);
    interceptors.add(new BridgeInterceptor(client.cookieJar()));
    interceptors.add(new CacheInterceptor(client.internalCache()));
    interceptors.add(new ConnectInterceptor(client));
    if (!forWebSocket) {
      interceptors.addAll(client.networkInterceptors());
    }
    interceptors.add(new CallServerInterceptor(forWebSocket));

    Interceptor.Chain chain = new RealInterceptorChain(interceptors, null, null, null, 0,
        originalRequest, this, eventListener, client.connectTimeoutMillis(),
        client.readTimeoutMillis(), client.writeTimeoutMillis());

    return chain.proceed(originalRequest);
  }

在这里添加了interceptors并且创建了一个RealInterceptorChain：

  @Override public Response proceed(Request request) throws IOException {
    return proceed(request, streamAllocation, httpCodec, connection);
  }

  public Response proceed(Request request, StreamAllocation streamAllocation, HttpCodec httpCodec,
      RealConnection connection) throws IOException {
    if (index >= interceptors.size()) throw new AssertionError();

    calls++;

    // If we already have a stream, confirm that the incoming request will use it.
    if (this.httpCodec != null && !this.connection.supportsUrl(request.url())) {
      throw new IllegalStateException("network interceptor " + interceptors.get(index - 1)
          + " must retain the same host and port");
    }

    // If we already have a stream, confirm that this is the only call to chain.proceed().
    if (this.httpCodec != null && calls > 1) {
      throw new IllegalStateException("network interceptor " + interceptors.get(index - 1)
          + " must call proceed() exactly once");
    }
    
    // Call the next interceptor in the chain.
    // index首次是0，然后index + 1实现链式调用
    RealInterceptorChain next = new RealInterceptorChain(interceptors, streamAllocation, httpCodec,
        connection, index + 1, request, call, eventListener, connectTimeout, readTimeout,
        writeTimeout);
    // 这里在之前添加过client.interceptors()、BridgeInterceptor等
    Interceptor interceptor = interceptors.get(index);
    // 通过各种interceptor对数据做处理
    Response response = interceptor.intercept(next);

    // Confirm that the next interceptor made its required call to chain.proceed().
    if (httpCodec != null && index + 1 < interceptors.size() && next.calls != 1) {
      throw new IllegalStateException("network interceptor " + interceptor
          + " must call proceed() exactly once");
    }

    // Confirm that the intercepted response isn't null.
    if (response == null) {
      throw new NullPointerException("interceptor " + interceptor + " returned null");
    }

    if (response.body() == null) {
      throw new IllegalStateException(
          "interceptor " + interceptor + " returned a response with no body");
    }

    return response;
  }

注意client.interceptors()是在最开始就处理的，比如声明的Log信息拦截器：

        OkHttpClient.Builder builder = new OkHttpClient.Builder();
        OkHttpClient okHttpClient = builder.build();
        if (BuildConfig.DEBUG) {
            HttpLoggingInterceptor loggingInterceptor = new HttpLoggingInterceptor();
            loggingInterceptor.setLevel(HttpLoggingInterceptor.Level.BODY);
            builder.addInterceptor(loggingInterceptor);
        }

通过一系列的interceptor，最终我们需要获取到网络请求到的数据，在最后一个interceptor CallServerInterceptor里：

 @Override public Response intercept(Chain chain) throws IOException {
    RealInterceptorChain realChain = (RealInterceptorChain) chain;
    HttpCodec httpCodec = realChain.httpStream();
    StreamAllocation streamAllocation = realChain.streamAllocation();
    RealConnection connection = (RealConnection) realChain.connection();
    Request request = realChain.request();

    long sentRequestMillis = System.currentTimeMillis();

    realChain.eventListener().requestHeadersStart(realChain.call());
    httpCodec.writeRequestHeaders(request);
    realChain.eventListener().requestHeadersEnd(realChain.call(), request);

    Response.Builder responseBuilder = null;
    if (HttpMethod.permitsRequestBody(request.method()) && request.body() != null) {
      // If there's a "Expect: 100-continue" header on the request, wait for a "HTTP/1.1 100
      // Continue" response before transmitting the request body. If we don't get that, return
      // what we did get (such as a 4xx response) without ever transmitting the request body.
      if ("100-continue".equalsIgnoreCase(request.header("Expect"))) {
        httpCodec.flushRequest();
        realChain.eventListener().responseHeadersStart(realChain.call());
        responseBuilder = httpCodec.readResponseHeaders(true);
      }

      if (responseBuilder == null) {
        // Write the request body if the "Expect: 100-continue" expectation was met.
        realChain.eventListener().requestBodyStart(realChain.call());
        long contentLength = request.body().contentLength();
        CountingSink requestBodyOut =
            new CountingSink(httpCodec.createRequestBody(request, contentLength));
        BufferedSink bufferedRequestBody = Okio.buffer(requestBodyOut);

        request.body().writeTo(bufferedRequestBody);
        bufferedRequestBody.close();
        realChain.eventListener()
            .requestBodyEnd(realChain.call(), requestBodyOut.successfulCount);
      } else if (!connection.isMultiplexed()) {
        // If the "Expect: 100-continue" expectation wasn't met, prevent the HTTP/1 connection
        // from being reused. Otherwise we're still obligated to transmit the request body to
        // leave the connection in a consistent state.
        streamAllocation.noNewStreams();
      }
    }

    httpCodec.finishRequest();

    if (responseBuilder == null) {
      realChain.eventListener().responseHeadersStart(realChain.call());
      responseBuilder = httpCodec.readResponseHeaders(false);
    }

    Response response = responseBuilder
        .request(request)
        .handshake(streamAllocation.connection().handshake())
        .sentRequestAtMillis(sentRequestMillis)
        .receivedResponseAtMillis(System.currentTimeMillis())
        .build();

    int code = response.code();
    if (code == 100) {
      // server sent a 100-continue even though we did not request one.
      // try again to read the actual response
      responseBuilder = httpCodec.readResponseHeaders(false);

      response = responseBuilder
              .request(request)
              .handshake(streamAllocation.connection().handshake())
              .sentRequestAtMillis(sentRequestMillis)
              .receivedResponseAtMillis(System.currentTimeMillis())
              .build();

      code = response.code();
    }

    realChain.eventListener()
            .responseHeadersEnd(realChain.call(), response);

    if (forWebSocket && code == 101) {
      // Connection is upgrading, but we need to ensure interceptors see a non-null response body.
      response = response.newBuilder()
          .body(Util.EMPTY_RESPONSE)
          .build();
    } else {
      response = response.newBuilder()
          .body(httpCodec.openResponseBody(response))
          .build();
    }

    if ("close".equalsIgnoreCase(response.request().header("Connection"))
        || "close".equalsIgnoreCase(response.header("Connection"))) {
      streamAllocation.noNewStreams();
    }

    if ((code == 204 || code == 205) && response.body().contentLength() > 0) {
      throw new ProtocolException(
          "HTTP " + code + " had non-zero Content-Length: " + response.body().contentLength());
    }

    return response;
  }

获得的数据response就交给了new okhttp3.Callback来处理：

      @Override public void onResponse(okhttp3.Call call, okhttp3.Response rawResponse) {
        Response<T> response;
        try {
          response = parseResponse(rawResponse);
        } catch (Throwable e) {
          throwIfFatal(e);
          callFailure(e);
          return;
        }

        try {
          callback.onResponse(OkHttpCall.this, response);
        } catch (Throwable t) {
          t.printStackTrace();
        }
      }

既然可以从网络上获取数据了，也就是说网络的适配介绍差不多了，那么从网络上获取的数据还需要转换为应用需要的数据，是如何处理的呢？接着往下分析：

  Response<T> parseResponse(okhttp3.Response rawResponse) throws IOException {
    ResponseBody rawBody = rawResponse.body();

    // Remove the body's source (the only stateful object) so we can pass the response along.
    rawResponse = rawResponse.newBuilder()
        .body(new NoContentResponseBody(rawBody.contentType(), rawBody.contentLength()))
        .build();

    int code = rawResponse.code();
    if (code < 200 || code >= 300) {
      try {
        // Buffer the entire body to avoid future I/O.
        ResponseBody bufferedBody = Utils.buffer(rawBody);
        return Response.error(bufferedBody, rawResponse);
      } finally {
        rawBody.close();
      }
    }

    if (code == 204 || code == 205) {
      rawBody.close();
      return Response.success(null, rawResponse);
    }

    ExceptionCatchingResponseBody catchingBody = new ExceptionCatchingResponseBody(rawBody);
    try {
      //很重要 这里使用了配置的Converter来转换数据
      T body = responseConverter.convert(catchingBody);
      return Response.success(body, rawResponse);
    } catch (RuntimeException e) {
      // If the underlying source threw an exception, propagate that rather than indicating it was
      // a runtime exception.
      catchingBody.throwIfCaught();
      throw e;
    }
  }

从代码中可以看到，正常情况下Retrofit把从网络上获取的数据通过配置的Converter来转换成应用层需要的格式，还记得responseConverter是什么吗，回到文章里面去找找吧，对于本文的分析就是GsonResponseBodyConverter：

  @Override public T convert(ResponseBody value) throws IOException {
    JsonReader jsonReader = gson.newJsonReader(value.charStream());
    try {
      // 这里的adapter从哪来？读者是否还记得
      T result = adapter.read(jsonReader);
      if (jsonReader.peek() != JsonToken.END_DOCUMENT) {
        throw new JsonIOException("JSON document was not fully consumed.");
      }
      return result;
    } finally {
      value.close();
    }
  }

本文的前部分就介绍过TypeAdapter，其中ReflectiveTypeAdapterFactory是本文使用到的Adapter，它的read是通过反射来实现类型的动态映射：

@Override public T read(JsonReader in) throws IOException {
      if (in.peek() == JsonToken.NULL) {
        in.nextNull();
        return null;
      }

      T instance = constructor.construct();

      try {
        in.beginObject();
        while (in.hasNext()) {
          String name = in.nextName();
          BoundField field = boundFields.get(name);
          if (field == null || !field.deserialized) {
            in.skipValue();
          } else {
            field.read(in, instance);
          }
        }
      } catch (IllegalStateException e) {
        throw new JsonSyntaxException(e);
      } catch (IllegalAccessException e) {
        throw new AssertionError(e);
      }
      in.endObject();
      return instance;
    }

处理完毕后的类型T就是Call泛型的实际类型，在本文中就是DataModel，数据处理成功后再通过应用层的回调把实例对象返回到应用：

try {
  //这个response是Response<T>，泛型T已经转换为实际类型
  callback.onResponse(OkHttpCall.this, response);
} catch (Throwable t) {
  t.printStackTrace();
}

回调到应用层就是如下形式，以下源码中的new Callback<T>与上文的callback一致：

      delegate.enqueue(new Callback<T>() {
        @Override public void onResponse(Call<T> call, final Response<T> response) {
          callbackExecutor.execute(new Runnable() {
            @Override public void run() {
              if (delegate.isCanceled()) {
                // Emulate OkHttp's behavior of throwing/delivering an IOException on cancellation.
                callback.onFailure(ExecutorCallbackCall.this, new IOException("Canceled"));
              } else {
                callback.onResponse(ExecutorCallbackCall.this, response);
              }
            }
          });
        }

中间经过了一层ExecutorCallbackCall的调度，最终还是调度到应用层的Callback：

        call.enqueue(new Callback<DataModel>() {
            @Override
            public void onResponse(Call<DataModel> call, Response<DataModel> response) {
                DataModel model = response.body();
                DataModel.WeatherInfoBean bean = model.getWeatherinfo();
                Log.d(TAG, "onResponse:" + bean.getCity());
            }

            @Override
            public void onFailure(Call<DataModel> call, Throwable t) {
                Log.d(TAG, "onFailure:" + t.getMessage());
            }
        });

3.结束语

比较简陋的把Retrofit的源码摸了一遍，了解了其中部分实现机制，发现确实解耦的比较好，目前只是初试Retrofit源码，有个框架上的认识而已，通过积累会完善这个系列。
Rxjava的分析会是下一步的计划，反正都是难啃的骨头，一起加油！文章的分析流程不确保完全正确，有分析的不对的地方欢迎指正。