/*
* Redberry: symbolic tensor computations.
*
* Copyright (c) 2010-2013:
* Stanislav Poslavsky <stvlpos@mail.ru>
* Bolotin Dmitriy <bolotin.dmitriy@gmail.com>
*
* This file is part of Redberry.
*
* Redberry is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Redberry is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with Redberry. If not, see <http://www.gnu.org/licenses/>.
*/
package cc.redberry.core.transformations.collect;
import cc.redberry.concurrent.OutputPortUnsafe;
import cc.redberry.core.combinatorics.Combinatorics;
import cc.redberry.core.indexgenerator.IndexGenerator;
import cc.redberry.core.indexmapping.IndexMapping;
import cc.redberry.core.indexmapping.IndexMappings;
import cc.redberry.core.indexmapping.Mapping;
import cc.redberry.core.indices.Indices;
import cc.redberry.core.indices.IndicesBuilder;
import cc.redberry.core.indices.IndicesUtils;
import cc.redberry.core.indices.SimpleIndices;
import cc.redberry.core.number.Complex;
import cc.redberry.core.tensor.*;
import cc.redberry.core.transformations.EliminateMetricsTransformation;
import cc.redberry.core.transformations.Transformation;
import cc.redberry.core.transformations.expand.ExpandPort;
import cc.redberry.core.utils.ArraysUtils;
import cc.redberry.core.utils.IntArrayList;
import cc.redberry.core.utils.TensorUtils;
import gnu.trove.map.hash.TIntObjectHashMap;
import gnu.trove.set.hash.TIntHashSet;
import java.util.ArrayList;
import java.util.Arrays;
import static cc.redberry.core.indices.IndicesUtils.*;
import static cc.redberry.core.tensor.Tensors.multiply;
import static cc.redberry.core.tensor.Tensors.sum;
/**
* @author Dmitry Bolotin
* @author Stanislav Poslavsky
*/
public class CollectTransformation implements Transformation {
private final TIntHashSet patternsNames;
private final Transformation[] transformations;
public CollectTransformation(SimpleTensor[] patterns, Transformation[] transformations) {
patternsNames = new TIntHashSet();
for (SimpleTensor t : patterns)
patternsNames.add(t.getName());
this.transformations = transformations;
}
public CollectTransformation(SimpleTensor... patterns) {
this(patterns, new Transformation[0]);
}
@Override
public Tensor transform(Tensor t) {
SumBuilder notMatched = new SumBuilder();
TIntObjectHashMap<ArrayList<Split>> map = new TIntObjectHashMap<>();
OutputPortUnsafe<Tensor> port = ExpandPort.createPort(t);
Tensor current;
Split toAdd;
ArrayList<Split> nodes;
out:
while ((current = port.take()) != null) {
toAdd = split(current);
if (toAdd.factors.length == 0) {
notMatched.put(current);
continue;
}
nodes = map.get(toAdd.hashCode);
if (nodes == null) {
nodes = new ArrayList<>();
nodes.add(toAdd);
map.put(toAdd.hashCode, nodes);
continue;
}
int[] match;
for (Split base : nodes) {
if ((match = matchFactors(base.factors, toAdd.factors)) != null) {
SimpleTensor[] toAddFactors = Combinatorics.reorder(toAdd.factors, match);
Mapping mapping =
IndexMappings.createBijectiveProductPort(toAddFactors, base.factors).take();
// mapping = mapping.inverseStates();
// for (Map.Entry<Integer, IndexMappingBufferRecord> entry : mapping.getMap().entrySet())
// entry.getValue().invertStates();
base.summands.add(ApplyIndexMapping.applyIndexMappingAutomatically(toAdd.summands.get(0), mapping, base.forbidden));
continue out;
}
}
nodes.add(toAdd);
}
for (ArrayList<Split> splits : map.valueCollection())
for (Split split : splits)
notMatched.put(split.toTensor(transformations));
return notMatched.build();
}
private Split split(Tensor tensor) {
if (tensor instanceof SimpleTensor || tensor instanceof Product) {
SimpleTensor[] factors;
Tensor summand;
if (tensor instanceof SimpleTensor)
if (patternsNames.contains(tensor.hashCode())) {
factors = new SimpleTensor[1];
factors[0] = (SimpleTensor) tensor;
summand = Complex.ONE;
} else
return new Split(new SimpleTensor[0], tensor);
else {//Product
//early check
boolean containsMatch = false;
for (Tensor t : tensor)
if (t instanceof SimpleTensor && patternsNames.contains(((SimpleTensor) t).getName())) {
containsMatch = true;
break;
}
if (!containsMatch) return new Split(new SimpleTensor[0], tensor);
ArrayList<SimpleTensor> factorsList = new ArrayList<>();
summand = tensor;
for (Tensor t : tensor) {
if (t instanceof SimpleTensor && patternsNames.contains(((SimpleTensor) t).getName())) {
factorsList.add((SimpleTensor) t);
assert summand != Complex.ONE;
if (summand instanceof Product)
summand = ((Product) summand).remove(t);
else summand = Complex.ONE;
}
}
factors = factorsList.toArray(new SimpleTensor[factorsList.size()]);
}
TIntHashSet freeIndices = new TIntHashSet(IndicesUtils.getIndicesNames(tensor.getIndices().getFree()));
//now we need to uncontract dummies and free
Indices factorIndices = new IndicesBuilder().append(factors).getIndices();
TIntHashSet dummies = new TIntHashSet(IndicesUtils.getIntersections(
factorIndices.getUpper().copy(), factorIndices.getLower().copy()));
SimpleIndices currentFactorIndices;
IntArrayList from = new IntArrayList(), to = new IntArrayList();
ArrayList<Tensor> kroneckers = new ArrayList<>();
int j, index, newIndex;
IndexGenerator generator = new IndexGenerator(TensorUtils.getAllIndicesNamesT(tensor).toArray());
for (int i = 0; i < factors.length; ++i) {
from.clear();
to.clear();
currentFactorIndices = factors[i].getIndices();
for (j = currentFactorIndices.size() - 1; j >= 0; --j) {
index = currentFactorIndices.get(j);
if (freeIndices.contains(getNameWithType(index))) {
newIndex = setRawState(getRawStateInt(index), generator.generate(getType(index)));
from.add(index);
to.add(newIndex);
kroneckers.add(Tensors.createKronecker(index, inverseIndexState(newIndex)));
} else if (IndicesUtils.getState(index) && dummies.contains(getNameWithType(index))) {
newIndex = setRawState(getRawStateInt(index), generator.generate(getType(index)));
from.add(index);
to.add(newIndex);
kroneckers.add(Tensors.createKronecker(index, inverseIndexState(newIndex)));
}
}
factors[i] = applyDirectMapping(factors[i],
new StateSensitiveMapping(from.toArray(), to.toArray()));
}
//temp check
// factorIndices = new IndicesBuilder().append(factors).getIndices();
// assert factorIndices.size() == factorIndices.getFree().size();
kroneckers.add(summand);
summand = Tensors.multiply(kroneckers.toArray(new Tensor[kroneckers.size()]));
summand = EliminateMetricsTransformation.eliminate(summand);
return new Split(factors, summand);
}
return new Split(new SimpleTensor[0], tensor);
}
private static final class Split {
final SimpleTensor[] factors;
final ArrayList<Tensor> summands = new ArrayList<>();
final int hashCode;//real hash code (with fields args)
final int[] forbidden;
private Split(SimpleTensor[] factors, Tensor summand) {
this.factors = factors;
this.summands.add(summand);
Arrays.sort(factors);
int hash = 17;
for (SimpleTensor f : factors)
hash = hash * 17 + f.hashCode();
this.hashCode = hash;
this.forbidden = IndicesUtils.getIndicesNames(new IndicesBuilder().append(factors).getIndices());
}
@Override
public int hashCode() {
return hashCode;
}
Tensor toTensor(Transformation[] transformations) {
Tensor sum = Transformation.Util.applySequentially(
Tensors.sum(summands.toArray(new Tensor[summands.size()])),
transformations);
Tensor[] ms = new Tensor[factors.length + 1];
ms[ms.length - 1] = sum;
System.arraycopy(factors, 0, ms, 0, factors.length);
return Tensors.multiply(ms);
}
@Override
public String toString() {
return multiply(factors) + " : " + sum(summands.toArray(new Tensor[summands.size()]));
}
}
static int[] matchFactors(SimpleTensor[] a, SimpleTensor[] b) {
if (a.length != b.length) return null;
int begin = 0, j, n, length = a.length;
int[] permutation = new int[length];
Arrays.fill(permutation, -1);
for (int i = 1; i <= length; ++i) {
if (i == length || a[i].hashCode() != b[i - 1].hashCode()) {
if (i - 1 != begin) {
OUT:
for (n = begin; n < i; ++n) {
for (j = begin; j < i; ++j)
if (permutation[j] == -1 && matchSimpleTensors(a[n], b[j])) {
permutation[j] = n;
continue OUT;
}
return null;
}
} else {
if (!matchSimpleTensors(a[i - 1], b[i - 1])) return null;
else permutation[i - 1] = i - 1;
}
begin = i;
}
}
return Combinatorics.inverse(permutation);
}
private static boolean matchSimpleTensors(SimpleTensor a, SimpleTensor b) {
if (a.hashCode() != b.hashCode()) return false;
if (a.getClass() != b.getClass()) return false;
if (a instanceof TensorField)
for (int i = a.size() - 1; i >= 0; --i)
if (!IndexMappings.positiveMappingExists(a.get(i), b.get(i))) return false;
return true;
}
private static SimpleTensor applyDirectMapping(SimpleTensor st, DirectIndexMapping mapping) {
SimpleIndices newIndices = st.getIndices().applyIndexMapping(mapping);
if (st instanceof TensorField)
return Tensors.field(st.getName(), newIndices, ((TensorField) st).getArgIndices(), ((TensorField) st).getArguments());
else
return Tensors.simpleTensor(st.getName(), newIndices);
}
private static abstract class DirectIndexMapping implements IndexMapping {
final int[] from, to;
private DirectIndexMapping(int[] from, int[] to) {
ArraysUtils.quickSort(from, to);
this.from = from;
this.to = to;
}
}
private static final class StateSensitiveMapping extends DirectIndexMapping {
private StateSensitiveMapping(int[] from, int[] to) {
super(from, to);
}
@Override
public int map(int from) {
int index;
if ((index = Arrays.binarySearch(this.from, from)) >= 0)
return to[index];
return from;
}
}
}