play framework2學習之旅Body parsers

Body parsers（body解析器）

What is a body parser?（還是那句話，認真讀第一段內容介紹！）

An HTTP request (at least for those using the POST and PUT operations) contains a body. This body can be formatted with any format specified in the Content-Type header.A body parser transforms this request body into a Java value.（這是最關鍵的一句話！！！！）

Note: You can’t write BodyParser implementation directly using Java. Because a Play BodyParser must handle the body content incrementaly using anIteratee[Array[Byte], A] it must be implemented in Scala.

However Play provides default BodyParsers that should fit most use cases (parsing Json, Xml, Text, uploading files). And you can reuse these default parsers to create your own directly in Java; for example you can provide an RDF parsers based on the Text one.

The BodyParser Java API

In the Java API, all body parsers must generate a play.mvc.Http.RequestBodyvalue. This value computed by the body parser can then be retrieved viarequest().body():

public static Result index() {
  RequestBody body = request().body();
  return ok("Got body: " + body);
}

You can specify the BodyParser to use for a particular action using [e-mail] annotation:

@BodyParser.Of(BodyParser.Json.class)
public static Result index() {
  RequestBody body = request().body();
  return ok("Got json: " + body.asJson());
}

The Http.RequestBody API

As we just said all body parsers in the Java API will give you aplay.mvc.Http.RequestBody value. From this body object you can retrieve the request body content in the most appropriate Java type.

這點要十分注意！！！這個是可以背過的O(∩_∩)O

Note: The RequestBody methods like asText() or asJson() will return null if the parser used to compute this request body doesn’t support this content type. For example in an action method annotated [e-mail](BodyParser.Json.class), calling asXml() on the generated body will retun null.

Some parsers can provide a most specific type than Http.RequestBody (ie. a subclass of Http.RequestBody). You can automatically cast the request body into another type using the as(...) helper method:

@BodyParser.Of(BodyLengthParser.class)
pulic static Result index() {
  BodyLength body = request().body().as(BodyLength.class);
  ok("Request body length: " + body.getLength());
}

Default body parser: AnyContent（這個遇到了去查吧，留個印象，瀏覽帶過）

If you don’t specify your own body parser, Play will use the default one guessing the most appropriate content type from the Content-Type header:

• text/plain: String, accessible via asText()
• application/json: JsonNode, accessible via asJson()
• text/xml: org.w3c.Document, accessible via asXml()
• application/form-url-encoded: Map<String, String[]>, accessible viaasFormUrlEncoded()
• multipart/form-data: Http.MultipartFormData, accessible viaasMultipartFormData()
• Any other content type: Http.RawBuffer, accessible via asRaw()

Example:

pulic static Result save() {
  RequestBody body = request().body();
  String textBody = body.asText();
  
  if(textBody != null) {
    ok("Got: " + text);
  } else {
    badRequest("Expecting text/plain request body");
  }
}

Max content length（知道有這個東西，知道在哪去改即可）

Text based body parsers (such as text, json, xml or formUrlEncoded) use a max content length because they have to load all the content into memory.

There is a default content length (the default is 100KB).

Tip: The default content size can be defined in application.conf:

parsers.text.maxLength=128K

You can also specify a maximum content length via the @BodyParser.Of annotation:

// Accept only 10KB of data.
@BodyParser.Of(value = BodyParser.Text.class, maxLength = 10 * 1024)
pulic static Result index() {
  if(request().body().isMaxSizeExceeded()) {
    return badRequest("Too much data!");
  } else {
    ok("Got body: " + request().body().asText()); 
  }
}