# tuplespace

**Repository Path**: junphine/tuplespace

## Basic Information

- **Project Name**: tuplespace
- **Description**: light weigh mapreduce
- **Primary Language**: Unknown
- **License**: LGPL-2.1
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2014-05-29
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README


==========

Distributed TupleSpace Computing Systems，Have hightweight map reduce system. 


Feather:
Easy of use, unlike hadoop
High performence, like spark
Scale up,
Build with Python,can run in Cpython,Jython,IronPython,PyPy
JsonRPC interface build on bottle



define map-reduce job example:

```
        
<code>
class GroupCount:
    def __init__(self,group_field='_id'):
        self.group_field=group_field
        pass
        #print 'start map reduce.',init_value
        
    #yeild tupid,v1,v2,...
    def map(self,i,x):
        tup=x
        if(self.group_field=='_id'):
            yield 0,1
            return
        if self.group_field in tup:    
            k=x[self.group_field]    
            yield (int(k)/10,dict(count=1,value=k))

    def reduce(self,k,ys):
        sumv=0
        count=0
        for y in ys:
            sumv+=y['value']*y['count'];
            count+=y['count']
        yield k,{'sum':sumv,'count':count}
        
</code>      
      
on client:

<code>
space = TSpaceProxy(_config.ServerConfig)
print space.login.__name__
space.login('u', 'p') 

rv,s,job= space.map_reduce('table', 'example', 'GroupCount',['score']); 
   
rv,s,result= space.wait_for(job)    
if not rv:
    raise TSpaceError(-1,str(result))
for tup in result:
    print 'example.GroupCount.reduce=', tup



</code>
```