package com.keyware.composeanalysis.task; import cn.hutool.core.collection.CollectionUtil; import cn.hutool.core.lang.Pair; import cn.hutool.core.util.StrUtil; import com.keyware.common.constant.enums.AnalysisStatusEnum; import com.keyware.composeanalysis.constant.FixedValue; import com.keyware.composeanalysis.constant.FunctionAndAnalysisAssemblyConst; import com.keyware.composeanalysis.constant.RedisConst; import com.keyware.composeanalysis.constant.SolrDBConst; import com.keyware.composeanalysis.constant.enums.AnalysisLevelEnum; import com.keyware.composeanalysis.constant.enums.FileAnalysisStatusEnum; import com.keyware.composeanalysis.entity.AnalysisTask; import com.keyware.composeanalysis.mongo.FileDataMongoDto; import com.keyware.composeanalysis.mongo.LineDataMongoDto; import com.keyware.composeanalysis.mongo.MatchOpenFile; import com.keyware.composeanalysis.mongo.MatchOpenFileMongoDto; import com.keyware.composeanalysis.solr.VersionTree; import com.keyware.composeanalysis.util.*; import com.keyware.keyswan.anaysis.Analysis; import com.keyware.keyswan.anaysis.AnalysisFactory; import com.keyware.keyswan.common.CodeFile; import com.keyware.utils.IdGenerator; import lombok.extern.log4j.Log4j2; import org.apache.commons.lang3.StringUtils; import org.apache.solr.common.SolrDocument; import org.apache.solr.common.SolrDocumentList; import org.springframework.data.mongodb.core.MongoTemplate; import org.springframework.data.mongodb.core.query.Update; import java.math.BigDecimal; import java.math.RoundingMode; import java.util.*; import java.util.concurrent.CountDownLatch; import java.util.stream.Collectors; import static org.springframework.data.mongodb.core.query.Criteria.where; /** * @author liuzongren * @ClassName LineAnalysisTask * @description: 行级别 特征提取定时任务 * @datetime 2024年 07月 25日 16:19 * @version: 1.0 */ @Log4j2 public class LineAnalysisTask extends IAnalysisTask { private MongoTemplate mongoTemplate; private AnalysisTask analysisTask; //被测件的文件信息 private FileDataMongoDto analysisFile; private SolrUtils solrUtils; private RedisUtil redisUtil; private CountDownLatch countDownLatch; public LineAnalysisTask(AnalysisTask analysisTask, FileDataMongoDto analysisFile, MongoTemplate mongoTemplate, CountDownLatch countDownLatch) { this.mongoTemplate = mongoTemplate; this.analysisTask = analysisTask; this.analysisFile = analysisFile; this.countDownLatch = countDownLatch; this.solrUtils = SpringContextUtils.getBean(SolrUtils.class); this.redisUtil = SpringContextUtils.getBean(RedisUtil.class); } /** * 行级别 源代码溯源 * 当前任务 需要在 文件级分析完成后 进行 */ @Override public void run() { //执行任务前,判断一下任务执行的状态 Object status = redisUtil.get(String.format(RedisConst.TASK_RUNNING_STATUS_KEY_PREFIX, analysisTask.getId())); if (status != null && (status.equals(AnalysisStatusEnum.STOP_ANALYSIS.getCode()) || status.equals(AnalysisStatusEnum.PAUSE_ANALYSIS.getCode()))) { log.info("任务已取消,fileName:{}", analysisFile.getName()); countDownLatch.countDown(); return; } //获取文件地址 String filePath = analysisFile.getFileUrl(); //获取文件名称 String fileName = analysisFile.getName(); AnalysisLogUtil.insert(mongoTemplate, "【行级特征提取】正在提取" + fileName); try { Analysis analysis = AnalysisFactory.getAnalysis(filePath); //获取文件行级特征md5 CodeFile codeFile = analysis.analysisFile(filePath, FunctionAndAnalysisAssemblyConst.LINE_EXTRACT, FunctionAndAnalysisAssemblyConst.LINE_EXTRACT); //从solr中获取特征相似的 文件 SolrDocumentList featureSimilarityFromSolr = getFeatureSimilarityFromSolr(codeFile); //计算文件的开源率 doAnalysis(featureSimilarityFromSolr, codeFile); //更新文件表的分析状态为3 行级特征以分析完毕 analysisFile.setFileAnalysisStatus(FileAnalysisStatusEnum.ANALYSIS_DONE.getCode()); mongoTemplate.update(FileDataMongoDto.class) .matching(where("_id").is(analysisFile.getId())) .replaceWith(analysisFile) .findAndReplace(); //插入日志 AnalysisLogUtil.insert(mongoTemplate, "【行级分析】完成" + fileName); log.info("文件" + fileName + ":行级分析完成"); } catch (Exception e) { AnalysisLogUtil.insertErrorInfo(mongoTemplate, "【行级分析】失败" + fileName, e); log.error("文件:{}行级别分析失败!", fileName,e); //修改当前文件分析状态未失败 mongoTemplate.update(FileDataMongoDto.class) .matching(where("_id").is(analysisFile.getId())) .apply(new Update().set("fileAnalysisStatus", FileAnalysisStatusEnum.FAILED_ANALYSIS.getCode())) .first(); } finally { countDownLatch.countDown(); } } /** * 计算开源率 被测件的开源率 * * @param matcheOpenSourceFiles * @param codeFile 文件解析结果 */ private void doAnalysis(SolrDocumentList matcheOpenSourceFiles, CodeFile codeFile) { //根据文件后缀判断需要查询的文件版本库名称 String versionIdCoreName = FixedValue.SUFFIX_SOLR_VERSION.get(analysisFile.getSuffix()); if (CollectionUtil.isEmpty(matcheOpenSourceFiles)) { //因为行的特征库较少，这里补充一个对比逻辑,如果当前文件解析失败，或者没有通过代码块匹配到数据，则直接通过文件的md5 再次查询一次solr库 checkByOriginalFileMd5(versionIdCoreName, analysisFile.getMd5()); return; } //保存所有匹配的行数信息,方便统计总的匹配行数 Set matchedFeatureMd5 = new HashSet<>(); //保存所有匹配的行数信息,方便统计总的匹配行数 Set matchLineRowsNum = new HashSet<>(); //获取文件总特征行数 String traitFileLineMd5 = codeFile.getTraitFileLineMd5(); List lineFeatureList = Arrays.asList(traitFileLineMd5.split(",")); //统计每个文件的开源率 List matchOpenFilesRes = calculateSimilarityAndOpenRate(matcheOpenSourceFiles, codeFile, versionIdCoreName, matchLineRowsNum, matchedFeatureMd5); //计算文件的总体特征相似度 BigDecimal featureSimilarity = new BigDecimal(matchedFeatureMd5.size()).divide(new BigDecimal(lineFeatureList.size()), 4, RoundingMode.HALF_UP).multiply(new BigDecimal(100)).setScale(2); //计算文件的总体开源率 BigDecimal openRate = new BigDecimal(matchLineRowsNum.size()).divide(new BigDecimal(analysisFile.getCodeRowNum()), 4, RoundingMode.HALF_UP).multiply(new BigDecimal(100)).setScale(2); //获取开源率的阈值 Integer openRateThreshold = analysisTask.getOpenRateThreshold(); //如果开源率大于阈值，则将当前文件设置成开源 if (openRate.compareTo(new BigDecimal(openRateThreshold)) >= 0) { analysisFile.setOpenType(true); } //保存当前文件的开源信息到mongo库中 MatchOpenFileMongoDto matchOpenFileMongo = new MatchOpenFileMongoDto(); matchOpenFileMongo.setId(IdGenerator.uuid32()) .setFilePath(analysisFile.getFileUrl()) .setFileName(analysisFile.getName()) .setFeatureSimilarity(featureSimilarity.floatValue()) .setOpenRate(openRate.floatValue()) .setOpenType(analysisFile.getOpenType()) .setMatchOpenFile(matchOpenFilesRes); log.info("文件" + analysisFile.getName() + ":开源率:" + openRate.floatValue() + ",特征相似度:" + featureSimilarity.floatValue()); mongoTemplate.save(matchOpenFileMongo); } /** * 计算当前文件的特征相似度 和 开源率 * * @param matchOpenFiles 通过MD5 匹配到的所有开源文件 * @param sourceFileBaseCoreName 当前文件特征文件的 solr coreName * @param matchLineRowsNum 所有开源文件匹配到的开源行号列表 * @param matchFeatureLineMd5s 所有开源文件匹配到的特征行MD5 * @return 匹配的开源文件解析后的结果集 */ private List calculateSimilarityAndOpenRate(SolrDocumentList matchOpenFiles, CodeFile fileAnalysisRes, String sourceFileBaseCoreName, Set matchLineRowsNum, Set matchFeatureLineMd5s) { //匹配的开源文件列表 List matchOpenFilesRes = new ArrayList<>(); //首先根据文件的MD5查询开源文件的版本ID,和路径信息 Set openSourceFileMd5s = matchOpenFiles.stream().map(doc -> (String) doc.get("sourceMd5")).collect(Collectors.toSet()); Map md5VersionInfoMap = solrUtils.batchQueryVersionIdFromSourceFileBaseBySourceMd5(sourceFileBaseCoreName, openSourceFileMd5s); //根据版本ID查询版本的详细信息 //todo 这里 查询一个版本的信息 需要检索 两个 solr 库 而且还需要检索 versioinTree 后面需要优化 Set openSourceFileVersionIds = md5VersionInfoMap.values().stream().map(doc -> (String) doc.get("versionId")).collect(Collectors.toSet()); List versionTrees = solrUtils.queryBatchVersionInfoByVersionIds(openSourceFileVersionIds); Map versionIdVersionInfoMap = versionTrees.stream().collect(Collectors.toMap(VersionTree::getVersionId, java.util.function.Function.identity())); String traitFileLineMd5 = fileAnalysisRes.getTraitFileLineMd5(); List lineFeatureList = Arrays.asList(traitFileLineMd5.split(",")); for (SolrDocument openSourceFile : matchOpenFiles) { //开源文件MD5 String openSourceFileMd5 = openSourceFile.getFieldValue("sourceMd5").toString(); //解析文件的特征行 String lineFeatureMd5s = (String) openSourceFile.get("tz_line_hay"); List openFileLineFeatures = Arrays.asList(lineFeatureMd5s.split(",")); //获取开源文件的文本信息 String openSourceContent = solrUtils.getOpenFileContentByMd5(openSourceFileMd5); //当前文件匹配特征行总行数 int currentFileMatchFeatureLineCount = 0; //遍历函数特征MD5 for (String lineFeatureMd5 : lineFeatureList) { //源文件的特征行列表 for (String openFileLineFeature : openFileLineFeatures) { if (lineFeatureMd5.equals(openFileLineFeature)) { matchFeatureLineMd5s.add(lineFeatureMd5); currentFileMatchFeatureLineCount++; } } } //当前文件的开源率 Pair> openRateAndSaveRowNum = SimilarityUtil.getOpenRateAndSaveRowNum(fileAnalysisRes.getSourceFileContent(), openSourceContent); //将当前文件匹配的行号，存储到缓存中，方便统计整体的开源率 matchLineRowsNum.addAll(openRateAndSaveRowNum.getValue()); //统计当前文件的特征相似度 BigDecimal featureSimilarity = new BigDecimal(currentFileMatchFeatureLineCount).divide(new BigDecimal(lineFeatureList.size()), 4, RoundingMode.HALF_UP).multiply(new BigDecimal(100)).setScale(2); SolrDocument openEntries = md5VersionInfoMap.get(openSourceFileMd5); VersionTree versionInfo = versionIdVersionInfoMap.get(openEntries.get("versionId")); if (versionInfo == null) { log.error("根据版本ID,未查询到相关的版本信息。versionId:{}", openEntries.get("versionId")); continue; } //组装当前开源文件的开源项目信息 MatchOpenFile matchOpenFileInfo = new MatchOpenFile(); matchOpenFileInfo.setPId(versionInfo.getProId()) .setPName(versionInfo.getProName()) .setSourceUrl((String) openEntries.get("fullPath")) .setFeatureSimilarity(featureSimilarity.floatValue()) .setOpenRate(openRateAndSaveRowNum.getKey()) .setVersion(versionInfo.getVersionName()) .setLicenseType(versionInfo.getLicenseType()) .setAnalyzeType(AnalysisLevelEnum.FUNCTION_LEVEL.getCode()); matchOpenFilesRes.add(matchOpenFileInfo); } return matchOpenFilesRes; } /** * 防止代码块特征库不全，再次根据文件MD5查询开源文件信息, 做二次校验 * * @param originalFileMd5 * @param versionIdCoreName */ private void checkByOriginalFileMd5(String versionIdCoreName, String originalFileMd5) { //根据文件的MD5,查询特征库，看当前文件是否在开源代码库中 SolrDocument versionIdAndPath = solrUtils.queryOne(versionIdCoreName, "sourceFileMd5:" + originalFileMd5, "versionId,fullPath,sourceFileMd5"); if (versionIdAndPath != null) { //根据版本ID查询版本的详细信息 VersionTree versionInfo = solrUtils.queryVersionInfoByVersionId((String) versionIdAndPath.get("versionId")); if (versionInfo != null) { //当前开源文件的开源项目信息 MatchOpenFile matchOpenFileInfo = new MatchOpenFile(); matchOpenFileInfo.setPId(versionInfo.getProId()) .setPName(versionInfo.getProName()) .setSourceUrl(versionInfo.getDownUrl()) .setFeatureSimilarity(100.00f) .setOpenRate(100.00f) .setAnalyzeType(AnalysisLevelEnum.FILE_LEVEL.getCode()); //保存当前文件的开源信息到mongo库中 MatchOpenFileMongoDto matchOpenFileMongo = new MatchOpenFileMongoDto(); matchOpenFileMongo.setId(IdGenerator.uuid32()) .setFilePath(analysisFile.getFileUrl()) .setFileName(analysisFile.getName()) .setOpenRate(100.00f) .setOpenType(analysisFile.getOpenType()) .setMatchOpenFile(Arrays.asList(matchOpenFileInfo)); mongoTemplate.save(matchOpenFileMongo); } } } /** * 将特征值插入到mongo库中 * @param features 特征集合 * @param lineDataMongoDto 当前分析任务 ，特征信息存储 * todo 后期 看看有没有插入的必要 * @param */ @Deprecated private void insertFeatureValue(String features, LineDataMongoDto lineDataMongoDto) { String[] featureMd5Arr = {}; if (StringUtils.isNotBlank(features)) { featureMd5Arr = features.split(","); } List lineFeatures = Arrays.asList(featureMd5Arr); List batchInsertList = new ArrayList<>(); if (CollectionUtil.isNotEmpty(lineFeatures)) { //这里的批量插入逻辑可以进行校验 //每10条存一次,解析的数据量如果过大，可能会超过MongoDB数据限制 int batchInsertStpe = 5000; int total = 0; for (int i = 0; i < lineFeatures.size(); i++) { if (total != batchInsertStpe) { batchInsertList.add(lineFeatures.get(i)); total++; } if (i == lineFeatures.size() - 1 && total != batchInsertStpe) { total = 0; lineDataMongoDto.setId(IdGenerator.uuid32()) .setLineFeatueMd5s(batchInsertList); mongoTemplate.insert(lineDataMongoDto); } if (total == batchInsertStpe) { total = 0; lineDataMongoDto.setId(IdGenerator.uuid32()) .setLineFeatueMd5s(batchInsertList); mongoTemplate.insert(lineDataMongoDto); batchInsertList.clear(); } } } else { lineDataMongoDto.setId(IdGenerator.uuid32()); mongoTemplate.insert(lineDataMongoDto); } } /** * 根据 特征值 从特征库中检索 具有特征相似的 * * @param codeFile 行特征信息 * @return */ private SolrDocumentList getFeatureSimilarityFromSolr(CodeFile codeFile) { Set queryMd5Set = new HashSet<>(); //每一行原内容MD5值集合 String cutFileLineMd5 = codeFile.getCutFileLineMd5(); if (StrUtil.isNotBlank(cutFileLineMd5)) { List lineCutList = Arrays.asList(cutFileLineMd5.split(",")); queryMd5Set.addAll(lineCutList); } //每一行特征内容MD5值集合 String traitFileLineMd5 = codeFile.getTraitFileLineMd5(); if (StrUtil.isNotBlank(traitFileLineMd5)) { List lineFeatureList = Arrays.asList(traitFileLineMd5.split(",")); queryMd5Set.addAll(lineFeatureList); } if (CollectionUtil.isEmpty(queryMd5Set)) { log.error("特征相似度检索失败，特征为空:{}", analysisFile.getName()); return new SolrDocumentList(); } //拼接行特征查询条件 String queryStr = "tz_line_hay:(" + StringUtils.join(queryMd5Set, " OR ") + ")"; log.info("查询条件: solrCoreName:{},queryStr:{}", SolrDBConst.CORE_NAME_SOURCE_FILE_INFO_TEMP, queryStr); SolrDocumentList result = solrUtils.query(SolrDBConst.CORE_NAME_SOURCE_FILE_INFO_TEMP, queryStr, "sourceMd5,tz_line_hay"); return result; } }