Support for deflate compression

[rymis]

As long as httplib has gzip support it is easy to add deflate. I did it with automatic encoding detection because some clients could compress content with GZip and to set deflate compression instead of correct gzip.

As a discussion starting point, I'd like to add ContentTranscoder abstraction to encode and decode custom streams, like LZ4, Snappy, etc.

[yhirose]

@rymis, thank you for the pull request. Could you explain more about "because some clients could compress content with GZip and to set deflate compression instead of correct gzip"? I would like to stick to the HTML spec unless the behavior is a widely accepted common practice. Thanks!

[rymis]

Sorry, I thought a little and realized, I was wrong. I read that some client library was sending incorrect Content-Encoding somewhere on stackoverflow, but I think it is not wise to support this behavior. I'll fix the MR to be standard complient. The actual RFC looks like:

4.2.2.  Deflate Coding

   The "deflate" coding is a "zlib" data format [RFC1950] containing a
   "deflate" compressed data stream [RFC1951] that uses a combination of
   the Lempel-Ziv (LZ77) compression algorithm and Huffman coding.

      Note: Some non-conformant implementations send the "deflate"
      compressed data without the zlib wrapper.


4.2.3.  Gzip Coding

   The "gzip" coding is an LZ77 coding with a 32-bit Cyclic Redundancy
   Check (CRC) that is commonly produced by the gzip file compression
   program [RFC1952].  A recipient SHOULD consider "x-gzip" to be
   equivalent to "gzip".

[yhirose]

Since '32' is now used for the automatic stream type detection, this comment ("The offset of 16 specifies...") is no longer valid. Could you update it?

[rymis]

Yes, for sure

[yhirose]

@rymis, I checked the revised code and looks good to me except the comment for inflateInit2(&strm, 32 + 15) . Could you update it? Thanks!

[yhirose]

Also one more request. Could you squash all the commits into a commit, so that the commit history could be simpler?

[yhirose]

I believe the comment for inflateInit2 should be like this:

    // 15 is the value of wbits, which should be at the maximum possible value
    // to ensure that any gzip stream can be decoded. The offset of 32 specifies
    // that the stream type should be automatically detected either gzip or deflate.
    is_valid_ = inflateInit2(&strm, 32 + 15) == Z_OK;	 

Also I just wonder why the unit test case GetMethod200withPercentEncoding sneaked in...

[rymis]

It's because of squashing of commits. It came from master :( I'll change the comment.

[yhirose]

Actually I can squash the commits on my side. Thanks for the adjustments.

[yhirose]

I have just merged it into the master. Thanks for your contribution!

[rymis]

Thanks!