Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on Nov 7, 2011
  1. Removing old variables.

  2. If we reset the current database to the same name and specify createI…

    then the creation logic should still run regardless of whether we are setting
    to the same database or not.
    Closes #33.
  3. Merge remote branch 'cygal/master'

Commits on Nov 6, 2011
Commits on Nov 5, 2011
Commits on Nov 4, 2011
  1. Sorry for the huge commit: should have merged master into httpStrateg…

    …y before
    merging it into master.
    1. Back ported fixes that were in master's procPacket() code into
    SagNativeHTTPAdapter's procPacket().
    2. Changed how the user selects which HTTP adapter to use - you now specify a
    string instead of instantiating your own adapter class (required redundant
    passing of host/port info).
    3. Wrote the cURL HTTP adapter: SagCURLHTTPAdapter.php. Really needs a
    different name.
    4. Moved the HTTP adapters into their own sub-directory: ./src/httpAdapters.
  2. Merge branch 'httpStrategy'

Commits on Oct 27, 2011
  1. @pquentin
Commits on Oct 21, 2011
Commits on Sep 16, 2011
Commits on Sep 11, 2011
  1. Fixing a typo bug (closes #26)

    Apparently I was trying to throw a concept of a class instead of an
    instantiation. Nice.
Commits on Sep 5, 2011
  1. @timjuravich

    Added getSession method, needed to grab the _session for finding curr…

    timjuravich authored
    …ent user when authenticating with cookie auth
Commits on Sep 3, 2011
  1. Cleaning up some comments.

Commits on Aug 29, 2011
  1. Merge branch 'HTTP/1.1'

Commits on Aug 26, 2011
  1. Now checking to see if the database exists when told to with a HEAD i…

    …nstead of a GET, as is proper. Not sure why I made this a GET before - whoops.
  2. Fixing a bug where HTTP response codes >=400 for a HEAD reqeust were not

    throwing a SagCouchException - they would just return a Sag response with error
    codes in the header and an empty body. Now instead of checking the body for
    errors, we first check to see if the request was a HEAD and make sure the HTTP
    status <400. Then we just return the response, thereby circumventing the body
    parsing logic.
    Side note, apparently json_decode(null) takes a few miliseconds longer than
    json_decode('{"a": "hi"}')
Commits on Aug 25, 2011
Commits on Aug 11, 2011
  1. First pass at supporting chunked transfer encoding. I basically hacke…

    …d around until I hit the target and now need to go back and refactor some pieces.
  2. Telling the server we want HTTP/1.1, no longer sending Connection: Ke…

    …ep-Alive, not connection pooling on Connection: close, and better !$line checking.
Commits on Aug 10, 2011
  1. Correcting the capitalization of stdClass from StdClass. Credit to ja…

    …nl for pointing it out - now I have to go back and correct years of muscle memory.
Something went wrong with that request. Please try again.