diff --git a/projects/fangjia-service/src/main/java/com/fangjia/service/address/processor/AbstractAddressProcessor.java b/projects/fangjia-service/src/main/java/com/fangjia/service/address/processor/AbstractAddressProcessor.java index ad0885831..f8e262112 100644 --- a/projects/fangjia-service/src/main/java/com/fangjia/service/address/processor/AbstractAddressProcessor.java +++ b/projects/fangjia-service/src/main/java/com/fangjia/service/address/processor/AbstractAddressProcessor.java @@ -133,10 +133,9 @@ public abstract class AbstractAddressProcessor { * @return */ protected String tokenDistrictName(String keyword){ - List tokens = CnTokenizer.tokenizer(keyword, true); - Token token = getTokenByType(tokens, KeywordEnum.district); - if(token != null && "district".equals(token.getKeywordTypes()) && keyword.contains(token.getName())){ - return token.getName(); + List tokens = tokenDistrictNames(keyword); + if(tokens.size() > 0){ + return tokens.get(0); } return null; } @@ -159,7 +158,35 @@ public abstract class AbstractAddressProcessor { protected List tokenDistrictNames(String keyword){ List names = new ArrayList(); List tokens = CnTokenizer.tokenizer(keyword, true); - for (Token token : tokens) { + List mergeTokens = new ArrayList(); + for(int i = 0; i < tokens.size(); i++){ + Token token = tokens.get(i); + mergeTokens.add(token); + if(token.getKeywordTypes().equals("none") && i != tokens.size() - 1){ + for(int j = i +1; j < tokens.size(); j++){ + Token nextToken = tokens.get(j); + if("none".equals(nextToken.getKeywordTypes())){ + token.setEnd(nextToken.getEnd()); + token.setName(token.getName() + nextToken.getName()); + i++; + }else{ + break; + } + } + }else if(token.getKeywordTypes().equals("district") && i != tokens.size() - 1){ + for(int j = i +1; j < tokens.size(); j++){ + Token nextToken = tokens.get(j); + if("partition".equals(nextToken.getKeywordTypes()) || "stage".equals(nextToken.getKeywordTypes())){ + token.setEnd(nextToken.getEnd()); + token.setName(token.getName() + nextToken.getName()); + i++; + }else{ + break; + } + } + } + } + for (Token token : mergeTokens) { if (token.containsType(KeywordEnum.district) && keyword.contains(token.getName())) { names.add(token.getName()); } diff --git a/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/KeywordEnum.java b/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/KeywordEnum.java index a86e1c897..fc6b7bf5f 100644 --- a/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/KeywordEnum.java +++ b/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/KeywordEnum.java @@ -19,7 +19,7 @@ public enum KeywordEnum { channel, // 房价网频道 house_number, // 室号 stage, // 分期 - partitions, // 分区 + partition, // 分区 town,//乡镇 village,//居委会 number_cn, total_price, room, area, hall, toilet, alpha; diff --git a/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/PatternTokenizer.java b/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/PatternTokenizer.java index 447ae9cc6..b9d18029c 100644 --- a/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/PatternTokenizer.java +++ b/projects/fangjia-service/src/main/java/com/fangjia/service/cnAnalyzer/PatternTokenizer.java @@ -14,6 +14,8 @@ public class PatternTokenizer { private static Pattern reg_house_number; // 匹配"5号301室" private static Pattern REG_RANGE_PATT; // 匹配"5号301室" private static Pattern REG_NUMBER_CN; // 匹配面积 + private static Pattern STAGE_PATTERN;// 分期 + private static Pattern PARTITION_PATTERN;// 分区 private static final String REG_NUM = "(((,|\\.|\\d| )*\\d)|(一|二|三|四|五|六|七|八|九|十|零|两|个|十|百|千|万|亿| )+)"; private static final String REG_RANGE_SPLITTER = "(-|~|~|-|到|至| )"; @@ -31,6 +33,8 @@ public class PatternTokenizer { REG_RANGE_PATT = Pattern.compile(REG_RANGE_SPLITTER); REG_NUMBER_CN = Pattern.compile(REG_NUM + "(弄|号|栋|厅|卫)"); + STAGE_PATTERN = Pattern.compile("(([零一二三四五六七八九十百千\\d]+)期?[,、至到]?)?([零一二三四五六七八九十百千\\d]+)期"); + PARTITION_PATTERN = Pattern.compile("(\\w|东|南|北|中|西|甲|乙|丙|丁|零|一|二|三|四|五|六|七|八|九|十)+(区|坊)"); } public static void main(String[] args) { @@ -52,6 +56,8 @@ public class PatternTokenizer { } public static String tokenizers(List tokens, String k) { + k = parsePartition(tokens, k); + k = parseStage(tokens, k); // 面积 k = parseArea(tokens, k); // 房型 @@ -77,6 +83,29 @@ public class PatternTokenizer { } return t; } + private static String parseStage(List ret, String t) { + String value = null; + Matcher m = STAGE_PATTERN.matcher(t); + while (m.find()) { + value = m.group(); + t = disableValue(t, value); + ret.add(new Token(value, m.start(), m.end(), KeywordEnum.stage.name())); + m = STAGE_PATTERN.matcher(t); + } + return t; + } + + private static String parsePartition(List ret, String t) { + String value = null; + Matcher m = PARTITION_PATTERN.matcher(t); + while (m.find()) { + value = m.group(); + t = disableValue(t, value); + ret.add(new Token(value, m.start(), m.end(), KeywordEnum.partition.name())); + m = PARTITION_PATTERN.matcher(t); + } + return t; + } private static String parseArea(List ret, String t) { String value = null;