def get_strengthening_constraint(self, input_graph: Graph) -> Graph:
strengthened_input_graph = Graph()
for constraint in self.constraints.values():
strengthened_input_graph.merge(
constraint.get_strengthening_constraint(input_graph))
return strengthened_input_graph
def _get_canonical_query_plans(self,
sequence: List[str],
transformation: Transformation) -> Dict[Skeleton, Set[QueryPlan]]:
meta_plan = self._meta_plans[transformation]
blueprint_item_lists = self._get_blueprint_item_lists(sequence,
meta_plan,
_d=len(sequence))
canonical_transformation = meta_plan.canonical_transformations[len(sequence)]
mapping = next(canonical_transformation.get_subgraph_mappings(transformation))
skeletons_to_plans: Dict[Skeleton, Set[QueryPlan]] = collections.defaultdict(set)
for blueprint_item_list in blueprint_item_lists:
# Breakdown the overall transformation in terms of the unit plans contained in the blueprint items.
# Store the connections between them as a graph mapping.
connections = GraphMapping()
connections.update(mapping)
graph = Graph()
for item in blueprint_item_list:
graph.merge(item.unit.transformation)
connections = connections.apply_mapping(item.canonical_mapping, only_keys=True)
if item.border_mapping:
connections.update(item.border_mapping)
connections = connections.apply_mapping(connections, only_values=True)
# Assemble the query plan
query_plan = QueryPlan(transformation,
units=[item.unit.transformation for item in blueprint_item_list],
all_connections=connections,
strengthenings=[item.unit.strengthenings[component_name]
for component_name, item in zip(sequence, blueprint_item_list)])
# Obtain the skeletons for which this query plan would work.
# External inputs are negative integers. See gauss.synthesis.skeleton for details.
ent_to_idx = {ent: -idx for idx, ent in enumerate(transformation.get_input_entities(), 1)}
possible_arg_ints_lists = []
for component_name, (idx, item) in zip(sequence, enumerate(blueprint_item_list, 1)):
# Get the mapped entities to the inputs of this unit's transformation, and look up their idx values.
arg_ints = [ent_to_idx[connections.m_ent[ent]] for ent in item.unit.transformation.get_input_entities()]
# Get all the permutations as well.
arg_ints_list = [arg_num_mapping.apply_list(arg_ints)
for arg_num_mapping in item.unit.component_entries[component_name].argument_mappings]
possible_arg_ints_lists.append(arg_ints_list)
ent_to_idx[item.unit.transformation.get_output_entity()] = idx
# The skeletons are then simply the all the combinations
for arg_ints_list in itertools.product(*possible_arg_ints_lists):
skeleton = Skeleton(list(zip(sequence, arg_ints_list)))
skeletons_to_plans[skeleton].add(query_plan)
return skeletons_to_plans
def prepare_solution(self, output: Any, output_graph: Graph) -> Solution:
if self.problem.input_names is not None:
int_to_names: Dict[int, str] = {
-idx: name
for idx, name in enumerate(self.problem.input_names, 1)
}
else:
int_to_names: Dict[int, str] = {
-idx: f"inp{idx}"
for idx in range(1,
len(self.problem.inputs) + 1)
}
int_to_names[self.skeleton.length] = self.problem.output_name
graph = Graph()
for g in self.graphs:
graph.merge(g)
# Perform transitive closure w.r.t the nodes corresponding to the intermediate outputs
# and take the induced subgraph containing all nodes except those
if self.skeleton.length > 1:
join_nodes = set.union(*(set(self.int_to_graph[i].iter_nodes())
for i in range(1, self.skeleton.length)))
self.domain.perform_transitive_closure(graph,
join_nodes=join_nodes)
graph = graph.induced_subgraph(keep_nodes=set(graph.iter_nodes()) -
join_nodes)
return self.domain.prepare_solution(
self.problem.inputs,
output,
graph,
self.problem.graph_inputs,
output_graph,
self.enumeration_items,
arguments=[arg_ints for (comp_name, arg_ints) in self.skeleton],
int_to_names=int_to_names,
int_to_obj=self.int_to_val)
def _solve_for_skeleton_recursive(
self,
problem: SynthesisProblem,
skeleton: Skeleton,
query_plans: QueryPlans,
context: SolverContext,
_depth: int = 0) -> Iterator[Tuple[Any, Graph]]:
domain = self._domain
component_name, arg_ints = skeleton[_depth]
inputs, g_inputs = context.get_arguments(depth=_depth)
inp_entities = [
next(iter(g_inp.iter_entities())) for g_inp in g_inputs
]
inp_graph = Graph()
for g_inp in g_inputs:
inp_graph.merge(g_inp)
# Get the strengthening constraint for this depth.
# Specifically, for every query, get the intersection of the strengthenings of all the query plans for that
# query at this particular depth. Then take the union of all of these.
# In other words, this strengthening constraint is a graph containing the nodes, edges, tags and tagged edges
# that must be satisfied by the graph containing the inputs, that is `inp_graph` in this context.
# This constraint can then be used by the `enumerate` procedure to speed up the search.
strengthening_constraint: Graph = context.waypoints[
_depth].get_strengthening_constraint(inp_graph)
enumeration_item: EnumerationItem
for enumeration_item in domain.enumerate(
component_name=component_name,
inputs=inputs,
g_inputs=g_inputs,
constants=problem.constants,
strengthening_constraint=strengthening_constraint):
output = enumeration_item.output
c_graph = enumeration_item.graph
o_graph = enumeration_item.o_graph
# for g in g_inputs:
# assert set(g.iter_nodes()).issubset(set(c_graph.iter_nodes()))
if problem.timeout is not None and time.time(
) - self._time_start > problem.timeout:
raise TimeoutError("Exceeded time limit.")
out_entity = next(iter(o_graph.iter_entities()))
c_graph.add_node(PlaceholderNode(entity=out_entity))
c_graph = Transformation.build_from_graph(
c_graph, input_entities=inp_entities, output_entity=out_entity)
# Check if the returned graph is consistent with the query plans.
if not context.check_validity(c_graph, depth=_depth):
continue
# Prepare for the next round.
context.step(output=output,
graph=c_graph,
output_graph=o_graph,
enumeration_item=enumeration_item,
depth=_depth)
if _depth == skeleton.length - 1:
# This was the last component, prepare the program and return it along with the final output and graph.
yield output, o_graph
else:
# Move on to the next component.
yield from self._solve_for_skeleton_recursive(problem,
skeleton,
query_plans,
context,
_depth=_depth +
1)
def init(self):
domain = PandasLiteSynthesisDomain()
replay = {k: iter(v) for k, v in self.replay_map.items()}
graph = Graph()
g_inputs = self._g_inputs = [
self._convert_inp_to_graph(inp) for inp in self.inputs
]
int_to_val = {-idx: inp for idx, inp in enumerate(self.inputs, 1)}
int_to_graph = {-idx: g_inp for idx, g_inp in enumerate(g_inputs, 1)}
# Run the generators to extract the programs and graphs for each component call.
# Merge the individual graphs into the master graph.
call_strs: List[str] = []
for idx, (component_name, arg_ints) in enumerate(self.skeleton, 1):
c_inputs = [int_to_val[i] for i in arg_ints]
g_c_inputs = [int_to_graph[i] for i in arg_ints]
output, program, c_graph, output_graph = next(
domain.enumerate(component_name,
c_inputs,
g_c_inputs,
replay=replay))
int_to_val[idx] = output
int_to_graph[idx] = output_graph
call_strs.append(program)
graph.merge(c_graph)
# Check that the final output is equivalent to the original output specified in the benchmark.
assert domain.check_equivalent(self.output, int_to_val[self.skeleton.length]), \
f"Generated output inconsistent with specified output in Pandas benchmark {self.b_id}"
# Retrofit the value of the output entity to the original output
cur_out_entity = next(ent for ent in graph.iter_entities()
if ent.value is int_to_val[self.skeleton.length])
cur_out_entity.value = self.output
# Perform transitive closure w.r.t the nodes corresponding to the intermediate outputs
# and take the induced subgraph containing all nodes except those
if self.skeleton.length > 1:
join_nodes = set.union(*(set(int_to_graph[i].iter_nodes())
for i in range(1, self.skeleton.length)))
domain.perform_transitive_closure(graph, join_nodes=join_nodes)
intent_graph = graph.induced_subgraph(
keep_nodes=set(graph.iter_nodes()) - join_nodes)
else:
intent_graph = graph
self._graph = intent_graph
# Also construct the string representation of the ground-truth program.
program_list: List[str] = []
for depth, (call_str,
(component_name,
arg_ints)) in enumerate(zip(call_strs, self.skeleton), 1):
arg_strs = [f"inp{-i}" if i < 0 else f"v{i}" for i in arg_ints]
call_str = call_str.format(**{
f"inp{idx}": arg_str
for idx, arg_str in enumerate(arg_strs, 1)
})
if depth == self.skeleton.length:
program_list.append(call_str)
else:
program_list.append(f"v{depth} = {call_str}")
self.program = "\n".join(program_list)
