Welcome to RaQueL

RaQueL is a query language, through which the robots use RoboBrain for various robotic applica- tions. The RQL provides a rich set of retrieval functions and programming constructs to perform complex traversals on the RoboBrain graph.

Commands

fetch(Pattern) - For querying a user specified pattern in RoboBrain graph. For more on Pattern
SortBy(list, property) - To sort the given list of nodes or relationships according to a given property(currently supported on for 'belief' property).
map - Evaluates a lazy map on a list of nodes or relationships specified.
filter - Evaluates a lazy map on a list of nodes or relationships specified.
user_defined_functions =lambda n: funcName(n) - RaQueL supports user defined functions.

#this sample code defines a function to get affordances of objects in the list n. It can be called as affordances('knife')
affordances=lambda n:fetch("({handle :'\" + n + \"'})-[:`HAS_AFFORDANCE`]->(v)")

Sample program

This program finds path which are 1 to 3 edges away from standing_human node to the phone node and then ranks these paths in ascending order by their beliefs

paths =fetch("({handle:'standing_human'})-[e*1..3]->({handle:'phone'})")
SortBy(paths,'Belief')

Example 1

This RaQueL query returns all the objects which a 'standing human' can use.

fetch("({handle:'standing_human'})-[:`CAN_USE`]->(v)")

This RaQueL query returns all the materials which a 'wall' has.

fetch("({handle:'wall'})-[:`HAS_MATERIAL`]->(v)")

Example 2

This program finds path which are 1 to 5 edges away from standing_human node to the phone node and then ranks these paths in ascending order by their beliefs.

paths =fetch("({handle:'standing_human'})-[e*1..5]->({handle:'phone'})")
SortBy(paths,'Belief')

Example 3

This program finds objects which a 'sitting human' can use. Then maps them to their affordances using lazy evaluations. This program iterates overs 2 elements in the objects list.

global objects
objects =fetch("({handle:'sitting_human'})-[:`CAN_USE`]->(V)")
affordances=lambda n:fetch("({handle :'\" + n + \"'})-[:`HAS_AFFORDANCE`]->(v)")
iter = imap( lambda u: (u) ,objects[1])
print iter.next()
print iter.next()

Example 4

This program finds a unique media represenation of 'tv'.

global entities, media
entities = lambda n: fetch("(v)-[:`HAS_MEDIA`]->({handle :'" + n + "'})")
media = lambda n:fetch("({handle :'" + n + "'})-[:`HAS_MEDIA`]->(v)")
ind_media = lambda a: ifilter(lambda u: len(entities(u)[1])==1,(media(a))[1])
iter1= ind_media('tv')	
print iter1.next()