ZipInputStream brokenness

[Per Bothner]

Adam Megacz wrote:
> I can think of plenty; see the comment at the top of
> gnu.gcj.protocol.http.HttpURLConnection.getHttpHeaders() for a great
> example of how buffering can be bad.

(It's actually gnu.gcj.protocol.http.URLConnection.)

I don't think this code says very much about buffers pro or con.
The current code uses a BufferedInputStream, so it *is*
buffering.  The way it keeps doing String concatenation, instead
of using a StringBuffer is a beginner's mistake.  Using the
programs current character encoding is obviously incorrect.
And of course using a BufferedReader to extra headers from a
file where you'll use something else to read the rest is
clearly going to be difficult.

In any case, if you are reading a jar file sequentially, you will
definitely want to buffer it, at some level.
-- 
	--Per Bothner
per@bothner.com   http://www.bothner.com/per/