java 连接操作elasticsearch(四)

  • 与关系查询
/**
     * &-与关系查询
     * 
     * @param parentheses  合并查询关键词 使用&&连接的
     * @param except_keywords 排除关键词
     * @param field1,field2
     */
    @Autowired
    public static BoolQueryBuilder getDataByMultiCondition(String parentheses, String except_keywords, String field1,
            String field2) {
        BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();
        BoolQueryBuilder areaQueryBuilder = QueryBuilders.boolQuery();
        BoolQueryBuilder personQueryBuilder = QueryBuilders.boolQuery();
        BoolQueryBuilder eventQueryBuilder = QueryBuilders.boolQuery();
        String[] strArray = parentheses.split("\\)");
        for (int i = 0; i < strArray.length; i++) {
            String subStr = strArray[i].replaceAll("\\(|\\)", "");
            String[] fieldArray = new String[] { field1, field2 };

            if (subStr != null && !"".equals(subStr)) {
                String regEx = "&";   //切割关键词
                Pattern p = Pattern.compile(regEx);
                Matcher m = p.matcher(subStr);

                String regEx2 = "\\|\\|";
                Pattern p2 = Pattern.compile(regEx2);
                Matcher m2 = p2.matcher(subStr);

                String[] subStrArray = {};

                if (m.find(1)) {
                    subStrArray = subStr.split(regEx);
                    for (int k = 0; k < fieldArray.length; k++) {
                        for (int j = 0; j < subStrArray.length; j++) {
                            if (i == 0) {
                                areaQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                
                            } else if (i == 1) {
                                personQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                            
                            } else {
                                eventQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                            }
                        }
                    }
                } else if (m2.find(1)) {
                    subStrArray = subStr.split(regEx2);
                    for (int k = 0; k < fieldArray.length; k++) {
                        for (int j = 0; j < subStrArray.length; j++) {
                            if (i == 0) {
                                areaQueryBuilder.should(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("area-should--subStr--" +
                                // subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            } else if (i == 1) {
                                personQueryBuilder
                                        .should(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("person-should--subStr--"
                                // + subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            } else {
                                eventQueryBuilder.should(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("event-should--subStr--" +
                                // subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            }
                        }
                    }
                } else {
                    subStrArray = subStr.split(regEx);
                    for (int k = 0; k < fieldArray.length; k++) {
                        for (int j = 0; j < subStrArray.length; j++) {
                            if (i == 0) {
                                areaQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("area-single--subStr--" +
                                // subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            } else if (i == 1) {
                                personQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("person-single--subStr--"
                                // + subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            } else {
                                eventQueryBuilder.must(QueryBuilders.matchPhraseQuery(fieldArray[k], subStrArray[j]));
                                // System.out.println("event-single--subStr--" +
                                // subStr + "--" + fieldArray[k] + ": "+
                                // subStrArray[j]);
                            }
                        }
                    }
                }
            }
        }

        if (areaQueryBuilder.hasClauses() == true && personQueryBuilder.hasClauses() == true
                && eventQueryBuilder.hasClauses() == true) {
            boolQueryBuilder.must(areaQueryBuilder);
            boolQueryBuilder.must(personQueryBuilder);
            boolQueryBuilder.must(eventQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == true && personQueryBuilder.hasClauses() == true
                && eventQueryBuilder.hasClauses() == false) {
            boolQueryBuilder.must(areaQueryBuilder);
            boolQueryBuilder.must(personQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == true && personQueryBuilder.hasClauses() == false
                && eventQueryBuilder.hasClauses() == true) {
            boolQueryBuilder.must(areaQueryBuilder);
            boolQueryBuilder.must(eventQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == false && personQueryBuilder.hasClauses() == true
                && eventQueryBuilder.hasClauses() == true) {
            boolQueryBuilder.must(personQueryBuilder).must(eventQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == true && personQueryBuilder.hasClauses() == false
                && eventQueryBuilder.hasClauses() == false) {
            boolQueryBuilder.must(areaQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == false && personQueryBuilder.hasClauses() == true
                && eventQueryBuilder.hasClauses() == false) {
            boolQueryBuilder.must(personQueryBuilder);
        } else if (areaQueryBuilder.hasClauses() == false && personQueryBuilder.hasClauses() == false
                && eventQueryBuilder.hasClauses() == true) {
            boolQueryBuilder.must(eventQueryBuilder);
        }

        if (field1 != null && field2 != null && field1 != field2) {
            if (except_keywords != null && !"".equals(except_keywords)) {
                String regEx3 = "\\|\\|";
                Pattern p3 = Pattern.compile(regEx3);
                Matcher m3 = p3.matcher(except_keywords);
                if (m3.find(1)) {
                    String[] exceptKwsArray = {};
                    exceptKwsArray = except_keywords.split(regEx3);
                    for (int k = 0; k < exceptKwsArray.length; k++) {
                        boolQueryBuilder.mustNot(QueryBuilders.matchPhraseQuery(field1, exceptKwsArray[k]));
                        boolQueryBuilder.mustNot(QueryBuilders.matchPhraseQuery(field2, exceptKwsArray[k]));
                    }
                }
            }
        }

        return boolQueryBuilder;
    }
  • 多个索引index表查询返回结果集SearchHits关键词高亮显示
public static SearchHits getResponseMultiResult(String[] indexsName, String type, TransportClient client,
            BoolQueryBuilder boolQueryBuilder, String sortStr, int pageNum, int recordNum, String field1,
            String field2) {
        SearchHits hits = null;
        if (indexsName.length > 0) {
            try {
                // 设置高亮属性
                HighlightBuilder highlightBuilder = new HighlightBuilder();
                highlightBuilder.preTags("<span style=\"color:red\">").postTags("</span>").field(field1).field(field2);
                // 设置ES查询属性
                SearchResponse response = null;
                if (sortStr != null && !"".equals(sortStr)) {
                    String[] sortArray = sortStr.split("###");
                    // 判断排序方式
                    SortOrder sortFlag = null;
                    if ("+".equals(sortArray[1])) {
                        sortFlag = SortOrder.ASC;
                    } else {
                        sortFlag = SortOrder.DESC;
                    }
                    // 执行ES查询
                    if ("articlesSort".equals(sortArray[0])) {
                        // 按ES库中的_score得分排序
                        response = client.prepareSearch(indexsName).setTypes(type).setQuery(boolQueryBuilder)
                                .highlighter(highlightBuilder).addSort(SortBuilders.scoreSort().order(sortFlag))
                                .setFrom(pageNum).setSize(recordNum).setExplain(true).execute().actionGet();
                    } else {
                        // 按排序字段进行排序
                        response = client.prepareSearch(indexsName).setTypes(type).setQuery(boolQueryBuilder)
                                .highlighter(highlightBuilder).addSort(sortArray[0], sortFlag).setFrom(pageNum)
                                .setSize(recordNum).setExplain(true).execute().actionGet();
                    }
                } else {
                    // 无排序
                    response = client.prepareSearch(indexsName).setTypes(type).setQuery(boolQueryBuilder)
                            .highlighter(highlightBuilder).setFrom(pageNum).setSize(recordNum).setExplain(true)
                            .execute().actionGet();
                }
                hits = response.getHits();
            } catch (Exception e) {
                e.printStackTrace();
            }
        }
        return hits;
    }
  • java处理hits集合
    StringBuffer sb=new StringBuffer();
            Findindex find = new Findindex();
            try {
                find = mapper.readValue(hit.getSourceAsString(), Findindex.class);
                find.setId(hit.getId());
                find.setIndex(hit.getIndex());    //添加索引名
                if(!"微博".equals(find.getSource_type())){
                    //标题涉及词高亮显示
                    HighlightField hTitle = hit.getHighlightFields().get("title");
                    if (hTitle != null) {
                        Text[] fragments = hTitle.fragments();
                        String hTitleStr = "";
                        for (Text text : fragments) {
                            hTitleStr += text;
                            String[] asr=text.toString().replaceAll("</span><span style=\"color:red\">","").split("</span>");
                            for (String string : asr) {
                                if(string.indexOf("<span")!=-1){
                                    try {
                                        /**style="color:red">的长度是是18,所以+18,
                                         * 为了避免其他地方出现>导致关键词挖取混乱,尽量使用长一点的字符串挖取
                                         */
                                        String s = string.substring(string.indexOf("style=\"color:red\">") + 18,string.length());
                                        if(sb.toString().indexOf(s)==-1){
                                            sb.append(s+" ");
                                        }
                                    } catch (Exception e) {
                                    }
                                    
                                }
                            }
                        }
                        find.setTitle(hTitleStr);
                    }
                }
                //内容涉及词高亮显示
                HighlightField hContent = hit.getHighlightFields().get("main_body");
                if (hContent != null) {
                    Text[] fragments = hContent.fragments();
                    String hContentStr = "";
                    for (Text text : fragments) {
                        hContentStr += text;
                        String[] asr=text.toString().replaceAll("</span><span style=\"color:red\">","").split("</span>");
                        for (String string : asr) {
                            if(string.indexOf("<span")!=-1){
                                try {
                                    /**style="color:red">的长度是是18,所以+18,
                                     * 为了避免其他地方出现>导致关键词挖取混乱,尽量使用长一点的字符串挖取
                                     */
                                    String s = string.substring(string.indexOf("style=\"color:red\">") + 18,string.length());
                                    if(sb.toString().indexOf(s)==-1){
                                        sb.append(s+" ");
                                    }
                                } catch (Exception e) {
                                }
                            }
                        }
                    }
                    find.setMain_body(hContentStr);
                }
                //是否已读
                EsStatus stsatus=select(find.getId(),userId);
                if(stsatus==null||stsatus.getRead()==null){
                    find.setRead(false);
                }else{
                    find.setRead(true);
                }
                find.setKeyword(sb.toString());   //添加涉及词
                String content_fingerprint = find.getContent_fingerprint();
                if(article==null||article.get(content_fingerprint)==null){    //article==null为不合并,后者为只有一条数据
                    find.setSimilarity(1);
                }else{
                    find.setSimilarity(article.get(content_fingerprint)); //添加相似文章数
                }
                list.add(find);
            } catch (Exception e) {
                e.printStackTrace();
            }
        }

     

posted @ 2020-11-04 10:47  过氧化氢  阅读(346)  评论(0编辑  收藏  举报