dln

; This is ~/.clojure/repl-init.clj
;
; I start a clojure repl something like:
;
; rlwrap --remember --complete-filenames \
;        --history-filename ~/.clojure/history \
;        --break-chars "\"\\'(){}[],^%$#@;:|" \
;     java -server -cp 'src/:lib/*' clojure.main \
;          -i ~/.clojure/repl-init.clj

dln

TimeUUIDs in Python are slow. Clojure (Java) is fast.

Python

85 >>> def test():
    for _ in xrange(1000000):
        uuid.uuid1()

...:

dln

data = {
 1 : {1.1: "data",
      1.2: "data",
      1.3: "data",
      1.4: "data",
      1.5: "data"},
 2 : {2.1: "data",
      2.2: "data",
      2.3: "data",
      2.4: "data",

dln

MyObjs(Standard)
----------------
 row_key => {ts:    1234,
             host:  "kremvax",
             attr1: data,
             attr2: data,
             attr2: data}

MyObjsIndices(Super)
--------------------

dln

Simple string
-------------

>>> s='foo bar baz'


>>> %timeit str(s)
1000000 loops, best of 3: 214 ns per loop

dln

import time
 
class TtlDict(dict):
    """A dict where values expire after a set time.
       Expiration is done on access, or using `prune` (linear)"""
 
    ttl = 5
 
    def __init__(self, *args, **kwargs):
        dict.__init__(self, *args, **kwargs)

dln

import sys
import termios
import time
import tty

def wait_space():
    if sys.stdin.read(1) == ' ':
        return time.time()
    return wait_space()

dln

First revision (Secondary Index)
==================================
Store objects in one CF, and index in another.

Note: In Cassandra 0.7, using a secondary index for columns id:google, id:fb, id:myspace would eliminate the need for UserIndex.

Users
------

"e0ee3406d83c11dfa47e002219990fd7": {

dln

"""
$ cat testar.thrift

struct Testar {
        1: i32 heltal = 42,
        2: optional string texten,
        3: optional string mertext = 'foo',
        4: optional map<string,string> properties
}

dln

ZOO_LOG4J_PROP="INFO,ROLLINGFILE"
ZOO_LOG_DIR="/var/log/zookeeper/"