Many Internet protocols and file types specify byte lengths for certain sections of data just prior to the section of data. Size Limit Input Stream provides a mechanism for the entire data to be opened as a stream, and then section of the data be parceled out to other handlers. The handler is easier to write because its mission is no longer "read from this stream and process the data, but be careful not to read more than 398 bytes", but rather "read all the data from this stream and process the data." Handlers that are easier to write have fewer bugs. This also allows handlers that know how to process a stream, but may not have been written with sub-streams in mind, to be re-purposed without a change to the handler code.

An example of a protocol for a Size Limit Input Stream would be appropriate is the HTTP 1.1 protocol. In the HTTP 1.1 protocol a header for the web page is sent that contains (among other things) the size of the document. The document is then sent, and the stream is left open for subsequent requests. An implementation of an HTTP 1.1 client might read the headers, determine the size, and then pass off the actual document handling to another class via a SizeLimitInputStream.

[Download /w Source | Browse Source | Documentation]