Adjust codebase to work with plain or gzipped request body payloads

[vkuznet]

This PR contains changes to DBSWriterModel to enable gzip payload reads. I adjusted the codebase to work with both types of payloads, the current one (plain) and gzipped payloads.

The gzipped payload can be supplied in HTTP request by using "Content-Encoding: gzip" header, e.g.

curl -H "Content-Encoding: gzip" -H "Content-type: application/json" --data-binary @/Users/vk/Downloads/bb.json.gz
http://localhost:8080

Once this code will be in place the DBS clients, like WMAgent, etc. can start supplying gzipped payloads to DBS POST APIs, like bulkblocks, blocks, files, datasets, etc.

The proposed changes are compatible with both 2.X and 3.X python versions, and fully support current mode of operations (without gzip) for payloads.

[vkuznet]

@yuyiguo , @amaltaro , @klannon , @KatyEllis I suggest that you take this PR into consideration as it can significantly improve our usage of DBS APIs and reduce latency related to large payloads. The changes are for DBS Writer, but once it is deployed, the other changes will be required to the clients, like WMAgent which should start adopting gzipped payloads and make adjustment to DBS POST API usage. I provided an example of how client should use HTTP request and it should be trivial to implement this in DMWM since it will only require to add extra HTTP header and start using gzip for payloads.

[yuyiguo]

@vkuznet This is on my to-do list. I'll let you know when I get into this.