Read JAX-RS body InputStream twice

Question

I have a JAX-RS logging filter to log request and response details, something like this:

public class LoggingFilter implements ContainerRequestFilter, ContainerResponseFilter {
    @Override
    public void filter(final ContainerRequestContext requestContext) throws IOException {
        ...
        String body = getBody(request);           
        ...
        if (LOGGER.isDebugEnabled()) {
            LOGGER.debug("request: {}", httpRequest);
        }
    }
}

The getBody() method reads the body content from the InputStream but I need to do some trick because I can not reset this stream. Without this little trick my rest methods always receive empty request body content:

private String getBody(final ContainerRequestContext requestContext) {
    try {
        byte[] body = IOUtils.toByteArray(requestContext.getEntityStream());

        InputStream stream = new ByteArrayInputStream(body);
        requestContext.setEntityStream(stream);

        return new String(body);
    } catch (IOException e) {
        return null;
    }
}

Is there any better way to read the body content?

Answer 1

EDIT Here's an improved version that seem much more robust and use JDK classes. Just invoke close() before reusing.

    public class CachingInputStream extends BufferedInputStream {    
    public CachingInputStream(InputStream source) {
        super(new PostCloseProtection(source));
        super.mark(Integer.MAX_VALUE);
    }

    @Override
    public synchronized void close() throws IOException {
        if (!((PostCloseProtection) in).decoratedClosed) {
            in.close();
        }
        super.reset();
    }

    private static class PostCloseProtection extends InputStream {
        private volatile boolean decoratedClosed = false;
        private final InputStream source;

        public PostCloseProtection(InputStream source) {
            this.source = source;
        }

        @Override
        public int read() throws IOException {
            return decoratedClosed ? -1 : source.read();
        }

        @Override
        public int read(byte[] b) throws IOException {
            return decoratedClosed ? -1 : source.read(b);
        }

        @Override
        public int read(byte[] b, int off, int len) throws IOException {
            return decoratedClosed ? -1 : source.read(b, off, len);
        }

        @Override
        public long skip(long n) throws IOException {
            return decoratedClosed ? 0 : source.skip(n);
        }

        @Override
        public int available() throws IOException {
            return source.available();
        }

        @Override
        public void close() throws IOException {
            decoratedClosed = true;
            source.close();
        }

        @Override
        public void mark(int readLimit) {
            source.mark(readLimit);
        }

        @Override
        public void reset() throws IOException {
            source.reset();
        }

        @Override
        public boolean markSupported() {
            return source.markSupported();
        }
    }
}

This allows to read the whole stream in the buffer, by tweaking the mark to Integer.MAXVALUE. This also makes sure that the source is properly closed on the first close to free up OS resource.

---

Old Answer

As you can't be sure the actual implementation of the InputStream support mark (markSupported()). You're better of caching the input stream itself in a first apprach.

For exemple in a ContainerRequestFilter :

@Component
@Provider
@PreMatching
@Priority(1)
public class ReadSomethingInPayloadFilter implements ContainerRequestFilter {

    @Override
    public void filter(ContainerRequestContext request) throws IOException {
        CachingInputStream entityStream = new CachingInputStream(request.getEntityStream());

        readPayload(entityStream);

        request.setEntityStream(entityStream.getCachedInputStream());
    }
}

The caching input stream is a naive approach to input stream caching, it is way similar to your approach :

class CachingInputStream extends InputStream {
    public static final int END_STREAM = -1;
    private final InputStream is;
    private final ByteArrayOutputStream baos = new ByteArrayOutputStream();

    public CachingInputStream(InputStream is) {
        this.is = is;
    }

    public InputStream getCachedInputStream() {
        return new ByteArrayInputStream(baos.toByteArray());
    }

    @Override
    public int read() throws IOException {
        int result = is.read();
        // Avoid rewriting the end char (-1) otherwise it will be considered as a real char.
        if (result != END_STREAM)
            baos.write(result);
        return result;
    }

    @Override
    public int available() throws IOException {
        return is.available();
    }

    @Override
    public void close() throws IOException {
        is.close();
    }

}

This implementation is naive in various ways, it can be improved in the following area and probably more :

• Check markSupported on the original stream
• Don't use the heap to store the cached input stream, this would avoid pressure on the GC
• Cache is unbounded currently this may be a good improvement, at the very least use the same bounds as your http server.