SSCollectionJSON.js

read
----
{
  "action": "read",
  "table": "user/shift/trail/comment",
  "constraints": {
     "user_id": 67 || [option1, option2, option3] || {range: [range-begin, range-end]} 
// if range-begin is null, this means all less than range-add. 
// if range-add is null, this means all greater than range-begin. 
// "NOW()" works for dates.
  },
  "properties": "*" || ["username", "summary", "modified"],
  "orderby": ["<" || ">", "modified"],
  "startIndex": 2,
  "range": {
    "startIndex": 2, // array base is 0
    "count": 20
  }
}
- returns array of results

delete
------
{
  "action": "delete",
  "table": "user/shift/trail/comment",
  "constraints": {
     "user_id": 67
  },
}
- returns number of rows deleted

update
------
{
  "action": "update",
  "table": "user/shift/trail/comment",
  "constraints": {
     "user_id": 67
  },
  "values": {
    "user_id": 68,
    "name": "Avital"
  }
}
- returns number of rows updated

create
------
{
  "action": "create",
  "table": "user/shift/trail/comment",
  "values": {
    "name": "Avital"
  }
}
- returns new id

bulk operations
++++++++++++++
just pass an array of those json's
the result will be an array of the results of each operation with respective indices

OH MY GOT HOW CRAZY IS THIS
###########################

[
  {
    "action": "read",
    "table": "workset",
    "properties": "*",
    "constraints": {
      "userid": 1
    }
  },
  {
    "action": "read",
    "table": "worksinset",
    "properties": "COUNT(*)",
    "constraints": {
      "setid": "%id%"
    }
  }
]

- "COUNT(*)" in the properties is just a simple way to get the count by (ab)using SQL syntax
- "%id%" in constraints means (CHECK THIS OUT): go over all the values for column "id" from the previous action in the bulk operation and substitute each one and then run it and put it all in an array. WOW HOW CRAZY


stream/collectiond bridge
+++++++++++++++++++++++++

Any stream can be acted upon by using pseudo-table !<stream-id>, for example "!3"


superstreams
++++++++++++

A stream that has "superstream"=1 means it is an aggregation of several other streams. Any event posted on that stream will use its object_ref to directly refer to another stream id. If that other stream is also a superstream, it will recursively read them all.

If you want to view which streams are part of a superstream rather than viewing the whole expanded aggregation, set "bare": 1 in the json description for your object, like

  {
    "action": "read",
    "table": "!2",
    "properties": "*",
    "bare": 1
  }


crazy url retrieval 
+++++++++++++++++++

Any table that has a url column (right now just events have this), if you set anything in the url column, when reading that table it will automatically retrieve the contents of that url and return them as if there is a column called url_reponse. It's like magic but wayyyyy insecure so we should later on restrict it only to certain URLs or something.