com.actelion.research.chem.StructureSearch Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of openchemlib Show documentation
Open Source Chemistry Library
There is a newer version: 2024.12.1
package com.actelion.research.chem;

import com.actelion.research.calc.ProgressController;
import com.actelion.research.chem.descriptor.*;
import com.actelion.research.util.ByteArrayComparator;

import java.util.concurrent.ConcurrentLinkedQueue;
import java.util.concurrent.atomic.AtomicInteger;

public class StructureSearch {
	private volatile StructureSearchSpecification mSpecification;
	private volatile StructureSearchDataSource mDataSource;
	private volatile StructureSearchController mSearchController;
	private volatile ProgressController mProgressController;
	private volatile StereoMolecule[] mQueryFragment;
	private volatile ByteArrayComparator mIDCodeComparator;
	private volatile DescriptorHandler mDescriptorHandler;
	private volatile Object[] mQueryDescriptor;
	private volatile long[] mQueryHashCode;
	private volatile byte[][] mQueryIDCode;
	private volatile int mDescriptorColumn;
	private volatile int mMaxSSSMatches,mMaxNonSSSMatches;
	private ConcurrentLinkedQueue mResultQueue;
	private AtomicInteger mSMPIndex,mMatchCount;

	/**
	 * This contructs a new structure search, which upon calling start()
	 * runs a multithreaded structure search on the structure rows provided by dataSource.
	 * If a searchController is given, this is asked for every row, whether the row
	 * meets all preconditions and qualifies for the search. 
	 * @param specification
	 * @param dataSource
	 * @param searchController may be null, if all rows need to be searched
	 * @param progressController may be null
	 * @param dhFactory if null then the default DescriptorHandlerStandard2DFactory is used
	 */
	public StructureSearch(StructureSearchSpecification specification,
						   StructureSearchDataSource dataSource,
						   StructureSearchController searchController,
						   ProgressController progressController,
						   DescriptorHandlerFactory dhFactory) {
		mSpecification = specification;
		mDataSource = dataSource;
		mSearchController = searchController;
		mProgressController = progressController;

		if (mSpecification != null) {
			// define needed descriptor handlers
			if (mSpecification.isSimilaritySearch()) {
				DescriptorHandlerFactory factory = (dhFactory != null) ? dhFactory : DescriptorHandlerStandard2DFactory.getFactory();
				mDescriptorHandler = factory.getDefaultDescriptorHandler(specification.getDescriptorShortName());
				}
			else if (mSpecification.isSubstructureSearch()) {
				mDescriptorHandler = DescriptorHandlerLongFFP512.getDefaultInstance();
				}
			}
		}

	/**
	 * If the search shall be aborted once it exceed a given number of matches,
	 * then define the maximum number of matches with this method before starting the search.
	 * Calling start with then return the first maximum count valid matches.
	 * @param maxSSSMatches maximum number of allowed sub-structure search matches (0: no limit)
	 * @param maxNonSSSMatches maximum number of allowed matches for other search types (0: no limit)
	 */
	public void setMatchLimit(int maxSSSMatches, int maxNonSSSMatches) {
		mMaxSSSMatches = maxSSSMatches;
		mMaxNonSSSMatches = maxNonSSSMatches;
		}

	public int[] start() {
		if (!mDataSource.isSupportedSearchType(mSpecification))
			return null;

		mMatchCount = new AtomicInteger(0);

		if (!mSpecification.isNoStructureSearch()) {
			final int queryStructureCount = mSpecification.getStructureCount();
			if (queryStructureCount == 0)
				return null;

			mDescriptorColumn = -1;
	        boolean largestFragmentOnly = mSpecification.isLargestFragmentOnly();

			if (mSpecification.isSubstructureSearch() || mSpecification.isSimilaritySearch()) {
				if (mSpecification.isSubstructureSearch()) {
					mDescriptorColumn = mDataSource.getDescriptorColumn(DescriptorConstants.DESCRIPTOR_FFP512.shortName);
					mQueryFragment = new StereoMolecule[queryStructureCount];
					for (int i=0; i();

		if (mProgressController != null && mSpecification.getStructureCount() > 1023)
			mProgressController.startProgress("Searching structures", 0, mSpecification.getStructureCount());

		int threadCount = Runtime.getRuntime().availableProcessors();
    	SearchThread[] t = new SearchThread[threadCount];
    	for (int i=0; i= 0 && (mProgressController == null || !mProgressController.threadMustDie())) {
				if (mProgressController != null && row%1024==1023)
					mProgressController.updateProgress(mSpecification.getStructureCount()-row);

				if (mSearchController == null || mSearchController.rowQualifies(row)) {
					boolean isMatch = false;

					if (mSpecification.isSubstructureSearch()) {
						if (mMaxSSSMatches != 0 && mMatchCount.get() > mMaxSSSMatches)
							break;

						mSSSearcher.setMolecule(mDataSource.getIDCode(row, false), (long[])mDataSource.getDescriptor(mDescriptorColumn, row, false));
						for (int i=0; i mMaxNonSSSMatches)
							break;

						if (mSpecification.isNoStructureSearch()) {
							isMatch = true;
							}
						else if (mSpecification.isSimilaritySearch()) {
							for (int i=0; i= mSpecification.getSimilarityThreshold()) {
									isMatch = true;
									break;
									}
								}
							}
						else if (mSpecification.isExactSearch()) {
							for (int i=0; i