java - 作为 byte[] 发送的字符串内容将其内容转义为查询字符串

Question

我需要通过网络发送一个字符串，并且我在 Spring Boot 的 REST Controller 的另一端获取它。

我使用以下几行发送数据:

byte[] content = sw.toString().getBytes(StandardCharsets.UTF_8); //sw is StringWriter
httpUrlConnection = (HttpURLConnection)new URL("http://127.0.0.1:8080/upload").openConnection();
httpUrlConnection.setDoInput(false);
httpUrlConnection.setDoOutput(true);
httpUrlConnection.setRequestMethod("POST");
httpUrlConnection.setFixedLengthStreamingMode(content.length);
httpUrlConnection.connect();
httpUrlConnection.getOutputStream().write(content);
httpUrlConnection.getOutputStream().flush();

我收到的是这样的:

@RequestMapping(value="/upload", method = RequestMethod.POST)
public ResponseEntity<String> upload(@RequestBody byte[] uploadedData)
    throws NoSuchAlgorithmException, InvalidKeyException, IOException {
  if(uploadedData == null) {
    log.info("Uploaded data was null.");
    return new ResponseEntity<String>("Data was null after upload.", HttpStatus.INTERNAL_SERVER_ERROR);
  }

  String receivedData = new String(uploadedData, StandardCharsets.UTF_8);
  log.info("" + receivedData);

但是虽然我发送的是这个...

-----BEGIN CERTIFICATE REQUEST-----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-----END CERTIFICATE REQUEST-----

我得到的是这样的:

-----BEGIN+CERTIFICATE+REQUEST-----%0D%0AMIIB0DCCATkCAQAwgY8xCzAJBgNVBAYTAkFVMSgwJgYDVQQKDB9UaGUgTGVnaW9u%0D%0AIG9mIHRoZSBCb3VuY3kgQ2FzdGxlMRIwEAYDVQQHDAlNZWxib3VybmUxETAPBgNV%0D%0ABAgMCFZpY3RvcmlhMS8wLQYJKoZIhvcNAQkBFiBmZWVkYmFjay1jcnlwdG9AYm91%0D%0AbmN5Y2FzdGxlLm9yZzCBnzANBgkqhkiG9w0BAQEFAAOBjQAwgYkCgYEAqE+pEyKS%0D%0A21JXgeldS83E+SSmQ4s2xYPqV7Yw2ebUKsxnbBz1KcXHdHS1MJ9PYzBJgogdigZW%0D%0A6hEQ83edP%2FAy%2FEQzGqeKUzFqNEsQh3PSdbF9N5k7b81tQHUbfIbNu1ofSBNa%2FEit%0D%0AMkOj1NAmwivpW0AA8aPZhGzYLYWcp0lsC78CAwEAAaAAMA0GCSqGSIb3DQEBBQUA%0D%0AA4GBAB0p0ySmfMkm3z8H4P8WwWJ8bMO2RNXEx0i9fU2ncJfdY0zEPYvM6zpUhwJP%0D%0AT9DsQBPdSy+VLbJ%2FPtYoiKIcupd+vriGYn3mqckXy7RBLqpiVsnw1rGE28oG4I9N%0D%0Au0p2AwDuC+KNuHgtrGxYrRnFTRKZpj2AoGuW1a6eSaNOhPeq%0D%0A-----END+CERTIFICATE+REQUEST-----%0D%0A=

这不好，因为它会杀死 PEM 解析器。

但是，我可以通过使用以下转换来克服这个问题，并且字符串变得可解析......

  receivedData = receivedData.replace("-----BEGIN+CERTIFICATE+REQUEST-----", "-----BEGIN CERTIFICATE REQUEST-----");
  receivedData = receivedData.replace("-----END+CERTIFICATE+REQUEST-----", "-----END CERTIFICATE REQUEST-----");
  receivedData = receivedData.replace("%0D%0A", "\r\n");
  receivedData = receivedData.replace("%2F", "/");
  receivedData = receivedData.substring(0, receivedData.length()-1);

在这种情况下解析成功...

2014-10-18 16:26:43.864  INFO 2432 --- [nio-8080-exec-1] demo.HelloController                     : PemParser returned: org.spongycastle.pkcs.PKCS10CertificationRequest@f5c81fb9
2014-10-18 16:26:43.870  INFO 2432 --- [nio-8080-exec-1] demo.HelloController                     : SUCCESS

...但我不认为这是一种真正可行的做事方式，因为它看起来不太稳定。一定有更好的方法。我尝试了 URLEncoder.decode(receivedData, "UTF-8"); 但它从实际数据中删除了 + 字符，这也不好，因为它使数据无法解析。

当我通过网络发送字符串时，是否有人对如何防止此URL转义有任何有用的想法？

Answer 1

嗯，我仍然不知道为什么会发生这种情况，但我确实做了一个解决方法。

我不是仅仅发送 String byte[ ] 内容，而是使用 https://jsonp.java.net/index.html 将数据包装在 JSON 对象中。根据http://docs.oracle.com/javaee/7/tutorial/doc/jsonp003.htm .

JsonObject model = Json.createObjectBuilder().add("data", sw.toString()).build();

httpsUrlConnection = (HttpsURLConnection)new URL("http://127.0.0.1:8443/upload").openConnection();

httpsUrlConnection.setDoInput(true);
httpsUrlConnection.setDoOutput(true);
httpsUrlConnection.setRequestProperty("Content-Type", "application/json");
httpsUrlConnection.setRequestMethod("POST");
httpsUrlConnection.connect();
try(JsonWriter jsonWriter = Json.createWriter(httpsUrlConnection.getOutputStream()))
{
    jsonWriter.write(model);  
}
httpsUrlConnection.getOutputStream().flush();
System.out.println("Response code: " + httpsUrlConnection.getResponseCode());

然后，在 Spring Boot 端，我进行了设置，以便按照 https://stackoverflow.com/a/8946142/2413303 接收 JSON 对象。 :

public class DataObject {
  private String data;

  public String getData() {
    return data;
  }

  public void setData(String data) {
    this.data = data;
  }
}

在 Controller 中:

@RequestMapping(value="/upload", method = RequestMethod.POST, consumes="application/json")
  public ResponseEntity<String> upload(@RequestBody DataObject uploadedData)
    throws NoSuchAlgorithmException, InvalidKeyException, IOException {
    if(uploadedData == null) {
      log.info("Uploaded data was null.");
      return new ResponseEntity<String>("Data was null after upload.", HttpStatus.INTERNAL_SERVER_ERROR);
    }
    String receivedData = uploadedData.getData();
    log.info("" + receivedData);

这样字符就不会被转义:

2014-10-20 13:03:15.550  INFO 6924 --- [nio-8443-exec-3] demo.HelloController                     : -----BEGIN CERTIFICATE REQUEST-----

编辑:我被告知问题是在没有设置内容类型的情况下，Spring默认为application/x-www-form-urlencoded，因为它应该已application/octet-stream。

编辑2:使用octet-stream，将其通过转义的\r和\n发送到\\r 和 \\n，并出于某种原因在字符串前后添加了 "。不知道为什么。