[HTTPCLIENT-1260] Javadoc incorrectly states StringBody(String)'s encoding uses the system default.

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 4.2.2
• Fix Version/s: 4.3 Alpha1
• Component/s: HttpClient (classic)
• Labels: documentation newbie
• Environment: Android
• Assignee: Unassigned
• Reporter: Tim
• Created: 15/Nov/12 14:03
• Updated: 05/Oct/13 19:42
• Resolved: 17/Nov/12 16:47

Description

(Sorry if this is the wrong component.)

In the class `org.apache.http.entity.mime.content.StringBody`, there is the following code:

/**

• Create a StringBody from the specified text.
• The mime type is set to "text/plain".
• The hosts default charset is used.
  *
• @param text to be used for the body, not {@code null}
  * @throws UnsupportedEncodingException
  * @throws IllegalArgumentException if the {@code text} parameter is null
  */
  public StringBody(final String text) throws UnsupportedEncodingException { this(text, "text/plain", null); }
  
  Note that it says "The hosts [sic] default charset is used.". This is incorrect. The code always uses US-ASCII no matter what the host's default charset is (e.g. on Android it is UTF-8).
  
  Quite annoying! It should read:
  
  /**
  * Create a StringBody from the specified text.
  * The mime type is set to "text/plain".
  * The charset is set to US-ASCII.
  *
  * @param text to be used for the body, not {@code null}
• @throws UnsupportedEncodingException
• @throws IllegalArgumentException if the {@code text}

  parameter is null
  */
  public StringBody(final String text) throws UnsupportedEncodingException

  { this(text, "text/plain", null); }

I'd appreciate it if you could fix this, so future developers won't have the frustration of working out why their non-ascii characters are all being changed to question marks.