ES系列：search处理过程实现1框架
Posted 等你归去来
tags:
篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了ES系列：search处理过程实现1框架相关的知识，希望对你有一定的参考价值。
上一篇文章中，我们看了get在es的实现过程，虽只是一个简单的单条查询，但看起来实现却非常之复杂。纠其原因，是我们围绕了太多外围的东西讲了，而其核心则无外乎三点：1. 定义id对应的机器节点； 2. 查找真正的docId； 3. 查找docId对应的field信息；
本篇，我们再看另一个es的重要功能：search. 可以说，整个es就是立足于search的，所以，单就这事，足够我们啃上许久许久了。但，我们可以分步来，今日只需聊个大概框架，细节留待日后再说。实际上，平时我们聊事物时，又何偿不是在聊框架类的东西呢。只是有时候钻到牛角尖去，反倒能体现一个人的水平问题了。而针对这一点，则往往会牵出一个人的两个能力点：抓取问题核心的能力；深度理解和思考的能力。
闲话休绪，进入正题：search的处理框架。
1. searchAction框架
我们就以如下请求作为研究来源，即如何发起一个普通的search请求：
# 查找 test 索引中字段 name=ali 的记录 curl -X GET -H \'content-type:application/json\' -d \'{"query":{"match":{"name":"ali"}}}\' http://localhost:9200/test/job/_search
如上备注所说，过滤条件其实就一个 name=ali, 我们可以很容易类比到sql中的表达：
select * from test where name=\'ali\';
看起来问题并不复杂，那么es中又是如何处理该事务的呢？ es中的包划分得比较清晰，比如http请求，会先交给rest包下实现处理，内部处理交由action模块处理，启动模块由bootstrap处理等等。总之，这是一个优秀应用必备一个特性：代码清晰易懂。
整个http请求的search入口包，其存放位置如下：
client实例负责许多的请求转发入口，负责与远程或者本机的es节点进行通讯，调度等重要工作。它是在es启动时初始化的一个重要实例，其存放位置如下：
action的内部请求定义包，用于在启动时注册处理器，以及在接收到http请求后，将其统一转发到内部节点处理，其存放位置如下：
最后，来看下 search 的语法包含哪些？这可以在每个具体Action的 routes() 方法中找到：
// org.elasticsearch.rest.action.search.RestSearchAction#routes @Override public List<Route> routes() { return unmodifiableList(asList( new Route(GET, "/_search"), new Route(POST, "/_search"), new Route(GET, "/{index}/_search"), new Route(POST, "/{index}/_search"), // Deprecated typed endpoints. new Route(GET, "/{index}/{type}/_search"), new Route(POST, "/{index}/{type}/_search"))); }
即 search支持不带索引、带索引、带索引带type、GET/POST搜索, 可谓是语法宽松得很呐。
2. search的框架实现
本节我们就来看看实现search功能，es都是如何做的呢？
接到外部请求后，会交给nettyHandler, 然后交给RestController, 然后再找到具体的handler, 然后进行prepare, accept. 具体实现细节可以参考前几篇文章。这里只想说明，最终会交到 RestSearchAction 进行处理。 而 RestSearchAction 继承了BaseRestHandler, 会统一走处理流程: prepareAction() -> 参数检查 -> 具体实现调用 ;
2.1. 搜索请求的接入
本部分主要讲解，当接到外部请求后，如果内部的searcher，因是有许多的线程池，以及许多的分布式节点，这前面许多工作并不会执行真正查询。故只会有一些分发起查询工作，参数解析等。我们可以先看个时序图，解其大致：
实现如下:
// org.elasticsearch.rest.BaseRestHandler#handleRequest @Override public final void handleRequest(RestRequest request, RestChannel channel, NodeClient client) throws Exception { // prepare the request for execution; has the side effect of touching the request parameters // 具体的处理实现上下文准备 final RestChannelConsumer action = prepareRequest(request, client); // validate unconsumed params, but we must exclude params used to format the response // use a sorted set so the unconsumed parameters appear in a reliable sorted order final SortedSet<String> unconsumedParams = request.unconsumedParams().stream().filter(p -> !responseParams().contains(p)).collect(Collectors.toCollection(TreeSet::new)); // validate the non-response params if (!unconsumedParams.isEmpty()) { final Set<String> candidateParams = new HashSet<>(); candidateParams.addAll(request.consumedParams()); candidateParams.addAll(responseParams()); throw new IllegalArgumentException(unrecognized(request, unconsumedParams, candidateParams, "parameter")); } if (request.hasContent() && request.isContentConsumed() == false) { throw new IllegalArgumentException("request [" + request.method() + " " + request.path() + "] does not support having a body"); } usageCount.increment(); // execute the action action.accept(channel); } // org.elasticsearch.rest.action.search.RestSearchAction#prepareRequest @Override public RestChannelConsumer prepareRequest(final RestRequest request, final NodeClient client) throws IOException { SearchRequest searchRequest = new SearchRequest(); /* * We have to pull out the call to `source().size(size)` because * _update_by_query and _delete_by_query uses this same parsing * path but sets a different variable when it sees the `size` * url parameter. * * Note that we can\'t use `searchRequest.source()::size` because * `searchRequest.source()` is null right now. We don\'t have to * guard against it being null in the IntConsumer because it can\'t * be null later. If that is confusing to you then you are in good * company. */ IntConsumer setSize = size -> searchRequest.source().size(size); // 解析参数 request.withContentOrSourceParamParserOrNull(parser -> parseSearchRequest(searchRequest, request, parser, client.getNamedWriteableRegistry(), setSize)); // 具体的search 业务处理入口 return channel -> { RestCancellableNodeClient cancelClient = new RestCancellableNodeClient(client, request.getHttpChannel()); cancelClient.execute(SearchAction.INSTANCE, searchRequest, new RestStatusToXContentListener<>(channel)); }; }
这前置工作，两个重点：1. 解析参数； 2. 构建业务处理的consumer逻辑； 其中，es中大量使用了lamda表达式，大大简化了java编程的繁文缛节，算是为java扳回点颜面。
search作为es中重要且复杂功能，其参数也是超级复杂，要想完全理解各参数，倒真是可以花上几篇的文章好好讲上几讲。不过想稍微多了解点，也可以展开下面的实现，看个大概。
// org.elasticsearch.rest.action.search.RestSearchAction#parseSearchRequest /** * Parses the rest request on top of the SearchRequest, preserving values that are not overridden by the rest request. * * @param requestContentParser body of the request to read. This method does not attempt to read the body from the {@code request} * parameter * @param setSize how the size url parameter is handled. {@code udpate_by_query} and regular search differ here. */ public static void parseSearchRequest(SearchRequest searchRequest, RestRequest request, XContentParser requestContentParser, NamedWriteableRegistry namedWriteableRegistry, IntConsumer setSize) throws IOException { if (searchRequest.source() == null) { searchRequest.source(new SearchSourceBuilder()); } searchRequest.indices(Strings.splitStringByCommaToArray(request.param("index"))); if (requestContentParser != null) { // 将外部请求转换为可读的格式，比如解析出 {"query":{"match":{"xx":"1"}}} // 此处相当于词法语法解析，有些难度呢 searchRequest.source().parseXContent(requestContentParser, true); } final int batchedReduceSize = request.paramAsInt("batched_reduce_size", searchRequest.getBatchedReduceSize()); searchRequest.setBatchedReduceSize(batchedReduceSize); if (request.hasParam("pre_filter_shard_size")) { searchRequest.setPreFilterShardSize(request.paramAsInt("pre_filter_shard_size", SearchRequest.DEFAULT_PRE_FILTER_SHARD_SIZE)); } if (request.hasParam("max_concurrent_shard_requests")) { // only set if we have the parameter since we auto adjust the max concurrency on the coordinator // based on the number of nodes in the cluster final int maxConcurrentShardRequests = request.paramAsInt("max_concurrent_shard_requests", searchRequest.getMaxConcurrentShardRequests()); searchRequest.setMaxConcurrentShardRequests(maxConcurrentShardRequests); } if (request.hasParam("allow_partial_search_results")) { // only set if we have the parameter passed to override the cluster-level default searchRequest.allowPartialSearchResults(request.paramAsBoolean("allow_partial_search_results", null)); } // do not allow \'query_and_fetch\' or \'dfs_query_and_fetch\' search types // from the REST layer. these modes are an internal optimization and should // not be specified explicitly by the user. String searchType = request.param("search_type"); if ("query_and_fetch".equals(searchType) || "dfs_query_and_fetch".equals(searchType)) { throw new IllegalArgumentException("Unsupported search type [" + searchType + "]"); } else { searchRequest.searchType(searchType); } // 解析顶层参数备用 parseSearchSource(searchRequest.source(), request, setSize); searchRequest.requestCache(request.paramAsBoolean("request_cache", searchRequest.requestCache())); String scroll = request.param("scroll"); if (scroll != null) { // scroll 解析，与时间相关 searchRequest.scroll(new Scroll(parseTimeValue(scroll, null, "scroll"))); } if (request.hasParam("type")) { deprecationLogger.deprecate("search_with_types", TYPES_DEPRECATION_MESSAGE); searchRequest.types(Strings.splitStringByCommaToArray(request.param("type"))); } searchRequest.routing(request.param("routing")); searchRequest.preference(request.param("preference")); searchRequest.indicesOptions(IndicesOptions.fromRequest(request, searchRequest.indicesOptions())); checkRestTotalHits(request, searchRequest); if (searchRequest.pointInTimeBuilder() != null) { preparePointInTime(searchRequest, request, namedWriteableRegistry); } else { searchRequest.setCcsMinimizeRoundtrips( request.paramAsBoolean("ccs_minimize_roundtrips", searchRequest.isCcsMinimizeRoundtrips())); } } // org.elasticsearch.search.builder.SearchSourceBuilder#parseXContent /** * Parse some xContent into this SearchSourceBuilder, overwriting any values specified in the xContent. Use this if you need to set up * different defaults than a regular SearchSourceBuilder would have and use {@link #fromXContent(XContentParser, boolean)} if you have * normal defaults. * * @param parser The xContent parser. * @param checkTrailingTokens If true throws a parsing exception when extra tokens are found after the main object. */ public void parseXContent(XContentParser parser, boolean checkTrailingTokens) throws IOException { XContentParser.Token token = parser.currentToken(); String currentFieldName = null; if (token != XContentParser.Token.START_OBJECT && (token = parser.nextToken()) != XContentParser.Token.START_OBJECT) { throw new ParsingException(parser.getTokenLocation(), "Expected [" + XContentParser.Token.START_OBJECT + "] but found [" + token + "]", parser.getTokenLocation()); } // 循环解析直到结束 while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) { if (token == XContentParser.Token.FIELD_NAME) { currentFieldName = parser.currentName(); } else if (token.isValue()) { if (FROM_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { from = parser.intValue(); } else if (SIZE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { size = parser.intValue(); } else if (TIMEOUT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { timeout = TimeValue.parseTimeValue(parser.text(), null, TIMEOUT_FIELD.getPreferredName()); } else if (TERMINATE_AFTER_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { terminateAfter = parser.intValue(); } else if (MIN_SCORE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { minScore = parser.floatValue(); } else if (VERSION_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { version = parser.booleanValue(); } else if (SEQ_NO_PRIMARY_TERM_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { seqNoAndPrimaryTerm = parser.booleanValue(); } else if (EXPLAIN_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { explain = parser.booleanValue(); } else if (TRACK_SCORES_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { trackScores = parser.booleanValue(); } else if (TRACK_TOTAL_HITS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { if (token == XContentParser.Token.VALUE_BOOLEAN || (token == XContentParser.Token.VALUE_STRING && Booleans.isBoolean(parser.text()))) { trackTotalHitsUpTo = parser.booleanValue() ? TRACK_TOTAL_HITS_ACCURATE : TRACK_TOTAL_HITS_DISABLED; } else { trackTotalHitsUpTo = parser.intValue(); } } else if (_SOURCE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { fetchSourceContext = FetchSourceContext.fromXContent(parser); } else if (STORED_FIELDS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { storedFieldsContext = StoredFieldsContext.fromXContent(SearchSourceBuilder.STORED_FIELDS_FIELD.getPreferredName(), parser); } else if (SORT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { sort(parser.text()); } else if (PROFILE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { profile = parser.booleanValue(); } else { throw new ParsingException(parser.getTokenLocation(), "Unknown key for a " + token + " in [" + currentFieldName + "].", parser.getTokenLocation()); } } else if (token == XContentParser.Token.START_OBJECT) { // 解析query 参数 if (QUERY_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { queryBuilder = parseInnerQueryBuilder(parser); } else if (POST_FILTER_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { postQueryBuilder = parseInnerQueryBuilder(parser); } else if (_SOURCE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { fetchSourceContext = FetchSourceContext.fromXContent(parser); } else if (SCRIPT_FIELDS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { scriptFields = new ArrayList<>(); while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) { scriptFields.add(new ScriptField(parser)); } } else if (INDICES_BOOST_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { deprecationLogger.deprecate("indices_boost_object_format", "Object format in indices_boost is deprecated, please use array format instead"); while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) { if (token == XContentParser.Token.FIELD_NAME) { currentFieldName = parser.currentName(); } else if (token.isValue()) { indexBoosts.add(new IndexBoost(currentFieldName, parser.floatValue())); } else { throw new ParsingException(parser.getTokenLocation(), "Unknown key for a " + token + " in [" + currentFieldName + "].", parser.getTokenLocation()); } } } else if (AGGREGATIONS_FIELD.match(currentFieldName, parser.getDeprecationHandler()) || AGGS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { aggregations = AggregatorFactories.parseAggregators(parser); } else if (HIGHLIGHT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { highlightBuilder = HighlightBuilder.fromXContent(parser); } else if (SUGGEST_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { suggestBuilder = SuggestBuilder.fromXContent(parser); } else if (SORT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { sorts = new ArrayList<>(SortBuilder.fromXContent(parser)); } else if (RESCORE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { rescoreBuilders = new ArrayList<>(); rescoreBuilders.add(RescorerBuilder.parseFromXContent(parser)); } else if (EXT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { extBuilders = new ArrayList<>(); String extSectionName = null; while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) { if (token == XContentParser.Token.FIELD_NAME) { extSectionName = parser.currentName(); } else { SearchExtBuilder searchExtBuilder = parser.namedObject(SearchExtBuilder.class, extSectionName, null); if (searchExtBuilder.getWriteableName().equals(extSectionName) == false) { throw new IllegalStateException("The parsed [" + searchExtBuilder.getClass().getName() + "] object has a " + "different writeable name compared to the name of the section that it was parsed from: found [" + searchExtBuilder.getWriteableName() + "] expected [" + extSectionName + "]"); } extBuilders.add(searchExtBuilder); } } } else if (SLICE.match(currentFieldName, parser.getDeprecationHandler())) { sliceBuilder = SliceBuilder.fromXContent(parser); } else if (COLLAPSE.match(currentFieldName, parser.getDeprecationHandler())) { collapse = CollapseBuilder.fromXContent(parser); } else if (POINT_IN_TIME.match(currentFieldName, parser.getDeprecationHandler())) { pointInTimeBuilder = PointInTimeBuilder.fromXContent(parser); } else if (RUNTIME_MAPPINGS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { runtimeMappings = parser.map(); } else { throw new ParsingException(parser.getTokenLocation(), "Unknown key for a " + token + " in [" + currentFieldName + "].", parser.getTokenLocation()); } } else if (token == XContentParser.Token.START_ARRAY) { if (STORED_FIELDS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { storedFieldsContext = StoredFieldsContext.fromXContent(STORED_FIELDS_FIELD.getPreferredName(), parser); } else if (DOCVALUE_FIELDS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { docValueFields = new ArrayList<>(); while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) { docValueFields.add(FieldAndFormat.fromXContent(parser)); } } else if (FETCH_FIELDS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { fetchFields = new ArrayList<>(); while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) { fetchFields.add(FieldAndFormat.fromXContent(parser)); } } else if (INDICES_BOOST_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) { indexBoosts.add(new IndexBoost(parser)); } } else if (SORT_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { sorts = new ArrayList<>(SortBuilder.fromXContent(parser)); } else if (RESCORE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { rescoreBuilders = new ArrayList<>(); while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) { rescoreBuilders.add(RescorerBuilder.parseFromXContent(parser)); } } else if (STATS_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { stats = new ArrayList<>(); while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) { if (token == XContentParser.Token.VALUE_STRING) { stats.add(parser.text()); } else { throw new ParsingException(parser.getTokenLocation(), "Expected [" + XContentParser.Token.VALUE_STRING + "] in [" + currentFieldName + "] but found [" + token + "]", parser.getTokenLocation()); } } } else if (_SOURCE_FIELD.match(currentFieldName, parser.getDeprecationHandler())) { fetchSourceContext = FetchSourceContext.fromXContent(parser); } else if (SEARCH_AFTER.match(currentFieldName, parser.getDeprecationHandler())) { searchAfterBuilder = SearchAfterBuilder.fromXContent(parser); } else { throw new ParsingException(parser.getTokenLocation(), "Unknown key for a " + token + " in [" + currentFieldName + "].", parser.getTokenLocation()); } } else { throw new ParsingException(parser.getTokenLocation(), "Unknown key for a " + token + " in [" + currentFieldName + "].", parser.getTokenLocation()); } } // 解析完成，token被使用完 if (checkTrailingTokens) { token = parser.nextToken(); if (token != null) { throw new ParsingException(parser.getTokenLocation(), "Unexpected token [" + token + "] found after the main object."); } } } // org.elasticsearch.index.query.AbstractQueryBuilder#parseInnerQueryBuilder /** * Parses a query excluding the query element that wraps it */ public static QueryBuilder parseInnerQueryBuilder(XContentParser parser) throws IOException { if (parser.currentToken() != XContentParser.Token.START_OBJECT) { if (parser.nextToken() != XContentParser.Token.START_OBJECT) { throw new ParsingException(parser.getTokenLocation(), "[_na] query malformed, must start with start_object"); } } if (parser.nextToken() == XContentParser.Token.END_OBJECT) { // we encountered \'{}\' for a query clause, it used to be supported, deprecated in 5.0 and removed in 6.0 throw new IllegalArgumentException("query malformed, empty clause found at [" + parser.getTokenLocation() +"]"); } if (parser.currentToken() != XContentParser.Token.FIELD_NAME) { throw new ParsingException(parser.getTokenLocation(), "[_na] query malformed, no field after start_object"); } String queryName = parser.currentName(); // move to the next START_OBJECT if (parser.nextToken() != XContentParser.Token.START_OBJECT) { throw new ParsingException(parser.getTokenLocation(), "[" + queryName + "] query malformed, no start_object after query name"); } QueryBuilder result; try { result = parser.namedObject(QueryBuilder.class, queryName, null); } catch (NamedObjectNotFoundException e) { String message = String.format(Locale.ROOT, "unknown query [%s]%s", queryName, SuggestingErrorOnUnknown.suggest(queryName, e.getCandidates())); throw new ParsingException(new XContentLocation(e.getLineNumber(), e.getColumnNumber()), message, e); } //end_object of the specific query (e.g. match, multi_match etc.) element if (parser.currentToken() != XContentParser.Token.END_OBJECT) { throw new ParsingException(parser.getTokenLocation(), "[" + queryName + "] malformed query, expected [END_OBJECT] but found [" + parser.currentToken() + "]"); } //end_object of the query object if (parser.nextToken() != XContentParser.Token.END_OBJECT) { throw new ParsingException(parser.getTokenLocation(), "[" + queryName + "] malformed query, expected [END_OBJECT] but found [" + parser.currentToken() + "]"); } return result; } // org.elasticsearch.rest.action.search.RestSearchAction#parseSearchSource /** * Parses the rest request on top of the SearchSourceBuilder, preserving * values that are not overridden by the rest request. */ private static void parseSearchSource(final SearchSourceBuilder searchSourceBuilder, RestRequest request, IntConsumer setSize) { // ?q=xx 格式的搜索 QueryBuilder queryBuilder = RestActions.urlParamsToQueryBuilder(request); if (queryBuilder != null) { searchSourceBuilder.query(queryBuilder); } int from = request.paramAsInt("from", -1); if (from != -1) { searchSourceBuilder.from(from); } int size = request.paramAsInt("size", -1); if (size != -1) { setSize.accept(size); } if (request.hasParam("explain")) { searchSourceBuilder.explain(request.paramAsBoolean("explain", null)); } if (request.hasParam("version")) { searchSourceBuilder.version(request.paramAsBoolean("version", null)); } if (request.hasParam("seq_no_primary_term")) { searchSourceBuilder.seqNoAndPrimaryTerm(request.paramAsBoolean("seq_no_primary_term", null)); } if (request.hasParam("timeout")) { searchSourceBuilder.timeout(request.paramAsTime("timeout", null)); } if (request.hasParam("terminate_after")) { int terminateAfter = request.paramAsInt("terminate_after", SearchContext.DEFAULT_TERMINATE_AFTER); if (terminateAfter < 0) { throw new IllegalArgumentException("terminateAfter must be > 0"); } else if (terminateAfter > 0) { searchSourceBuilder.terminateAfter(terminateAfter); } } StoredFieldsContext storedFieldsContext = StoredFieldsContext.fromRestRequest(SearchSourceBuilder.STORED_FIELDS_FIELD.getPreferredName(), request); if (storedFieldsContext != null) { searchSourceBuilder.storedFields(storedFieldsContext); } String sDocValueFields = request.param("docvalue_fields"); if (sDocValueFields != null) { if (Strings.hasText(sDocValueFields)) { String[] sFields = Strings.splitStringByCommaToArray(sDocValueFields); for (String field : sFields) { searchSourceBuilder.docValueField(field, null); } } } FetchSourceContext fetchSourceContext = FetchSourceContext.parseFromRestRequest(request); if (fetchSourceContext != null) { searchSourceBuilder.fetchSource(fetchSourceContext); } if (request.hasParam("track_scores")) { searchSourceBuilder.trackScores(request.paramAsBoolean("track_scores", false)); } if (request.hasParam("track_total_hits")) { if (Booleans.isBoolean(request.param("track_total_hits"))) { searchSourceBuilder.trackTotalHits( request.paramAsBoolean("track_total_hits", true) ); } else { searchSourceBuilder.trackTotalHitsUpTo( request.paramAsInt("track_total_hits", SearchContext.DEFAULT_TRACK_TOTAL_HITS_UP_TO) ); } } String sSorts = request.param("sort"); if (sSorts != null) { String[] sorts = Strings.splitStringByCommaToArray(sSorts); for (String sort : sorts) { int delimiter = sort.lastIndexOf(":"); if (delimiter != -1) { String sortField = sort.substring(0, delimiter); String reverse = sort.substring(delimiter + 1); if ("asc".equals(reverse)) { searchSourceBuilder.sort(sortField, SortOrder.ASC); } else if ("desc".equals(reverse)) { searchSourceBuilder.sort(sortField, SortOrder.DESC); } } else { searchSourceBuilder.sort(sort); } } } String sStats = request.param("stats"); if (sStats != null) { searchSourceBuilder.stats(Arrays.asList(Strings.splitStringByCommaToArray(sStats))); } String suggestField = request.param("suggest_field"); if (suggestField != null) { String suggestText = request.param("suggest_text", request.param("q")); int suggestSize = request.paramAsInt("suggest_size", 5); String suggestMode = request.param("suggest_mode"); searchSourceBuilder.suggest(new SuggestBuilder().addSuggestion(suggestField, termSuggestion(suggestField) .text(suggestText).size(suggestSize) .suggestMode(SuggestMode.resolve(suggestMode)))); } } // org.elasticsearch.rest.action.RestActions#urlParamsToQueryBuilder Spring框架系列(13) - SpringMVC实现原理之DispatcherServlet的初始化过程框架手写系列---apt注解处理器方式实现ButterKnife框架
Motan在服务provider端用于处理request的线程池
SpringCloud系列四：Eureka 服务发现框架（定义 Eureka 服务端Eureka 服务信息Eureka 发现管理Eureka 安全配置Eureka-HA(高可用) 机制Eur(代码片段