网站建设专业的公司/宁波网站推广营销

目录

验证数据

加载城市数据

查询列车站点信息

查询列车余票信息

构建列车返回数据

---

12306 项目中列车数据检索接口路径  TicketController的pageListTicketQuery。

@GetMapping("/api/ticket-service/ticket/query")public Result<TicketPageQueryRespDTO> pageListTicketQuery(TicketPageQueryReqDTO requestParam) {return Results.success(ticketService.pageListTicketQueryV1(requestParam));}

验证数据

查询列车数据 Service 实现层接口第一行代码，就是通过责任链模式验证数据是否必填以及城市

数据是否存在等执行逻辑。

ticketPageQueryAbstractChainContext.handler(TicketChainMarkEnum.TRAIN_QUERY_FILTER.name(), requestParam);

handler具体代码如下：

1. 检查相关数据是否为空或空的字符串，这个是最先被执行的：

/*** 查询列车车票流程过滤器之验证数据是否为空或空的字符串** */
@Component
public class TrainTicketQueryParamNotNullChainFilter implements TrainTicketQueryChainFilter<TicketPageQueryReqDTO> {@Overridepublic void handler(TicketPageQueryReqDTO requestParam) {if (StrUtil.isBlank(requestParam.getFromStation())) {throw new ClientException("出发地不能为空");}if (StrUtil.isBlank(requestParam.getToStation())) {throw new ClientException("目的地不能为空");}if (requestParam.getDepartureDate() == null) {throw new ClientException("出发日期不能为空");}}@Overridepublic int getOrder() {return 0;}
}

2. 检查数据是否正确

/*** 查询列车车票流程过滤器之验证数据是否正确** */
@Component
@RequiredArgsConstructor
public class TrainTicketQueryParamVerifyChainFilter implements TrainTicketQueryChainFilter<TicketPageQueryReqDTO> {private final RegionMapper regionMapper;private final StationMapper stationMapper;private final DistributedCache distributedCache;private final RedissonClient redissonClient;/*** 缓存数据为空并且已经加载过标识*/private static boolean CACHE_DATA_ISNULL_AND_LOAD_FLAG = false;@Overridepublic void handler(TicketPageQueryReqDTO requestParam) {StringRedisTemplate stringRedisTemplate = (StringRedisTemplate) distributedCache.getInstance();HashOperations<String, Object, Object> hashOperations = stringRedisTemplate.opsForHash();List<Object> actualExistList = hashOperations.multiGet(QUERY_ALL_REGION_LIST,ListUtil.toList(requestParam.getFromStation(), requestParam.getToStation()));long emptyCount = actualExistList.stream().filter(Objects::isNull).count();if (emptyCount == 0L) {return;}if (emptyCount == 1L || (emptyCount == 2L && CACHE_DATA_ISNULL_AND_LOAD_FLAG && distributedCache.hasKey(QUERY_ALL_REGION_LIST))) {throw new ClientException("出发地或目的地不存在");}RLock lock = redissonClient.getLock(LOCK_QUERY_ALL_REGION_LIST);lock.lock();try {if (distributedCache.hasKey(QUERY_ALL_REGION_LIST)) {actualExistList = hashOperations.multiGet(QUERY_ALL_REGION_LIST,ListUtil.toList(requestParam.getFromStation(), requestParam.getToStation()));emptyCount = actualExistList.stream().filter(Objects::nonNull).count();if (emptyCount != 2L) {throw new ClientException("出发地或目的地不存在");}return;}List<RegionDO> regionDOList = regionMapper.selectList(Wrappers.emptyWrapper());List<StationDO> stationDOList = stationMapper.selectList(Wrappers.emptyWrapper());HashMap<Object, Object> regionValueMap = Maps.newHashMap();for (RegionDO each : regionDOList) {regionValueMap.put(each.getCode(), each.getName());}for (StationDO each : stationDOList) {regionValueMap.put(each.getCode(), each.getName());}hashOperations.putAll(QUERY_ALL_REGION_LIST, regionValueMap);CACHE_DATA_ISNULL_AND_LOAD_FLAG = true;emptyCount = regionValueMap.keySet().stream().filter(each -> StrUtil.equalsAny(each.toString(), requestParam.getFromStation(), requestParam.getToStation())).count();if (emptyCount != 2L) {throw new ClientException("出发地或目的地不存在");}} finally {lock.unlock();}}@Overridepublic int getOrder() {return 20;}
}

加载城市数据

12306 站点查询实际功能中，比如你搜索了北京南到杭州东的搜索条件，它会帮你列出北京到杭州所有的列车车次。这个很好实现，直接通过站点关联到城市，通过城市查询列车即可。我们在缓存中，有一个 Hash 结构数据，专门负责保存列车站点 Code 值与城市之间的关联关系。

// 列车查询逻辑较为复杂，详细解析文章查看 https://nageoffer.com/12306/question// v1 版本存在严重的性能深渊问题，v2 版本完美的解决了该问题。通过 Jmeter 压测聚合报告得知，性能提升在 300% - 500%+List<Object> stationDetails = stringRedisTemplate.opsForHash().multiGet(REGION_TRAIN_STATION_MAPPING, Lists.newArrayList(requestParam.getFromStation(), requestParam.getToStation()));long count = stationDetails.stream().filter(Objects::isNull).count();if (count > 0) {RLock lock = redissonClient.getLock(LOCK_REGION_TRAIN_STATION_MAPPING);lock.lock();try {stationDetails = stringRedisTemplate.opsForHash().multiGet(REGION_TRAIN_STATION_MAPPING, Lists.newArrayList(requestParam.getFromStation(), requestParam.getToStation()));count = stationDetails.stream().filter(Objects::isNull).count();if (count > 0) {List<StationDO> stationDOList = stationMapper.selectList(Wrappers.emptyWrapper());Map<String, String> regionTrainStationMap = new HashMap<>();stationDOList.forEach(each -> regionTrainStationMap.put(each.getCode(), each.getRegionName()));stringRedisTemplate.opsForHash().putAll(REGION_TRAIN_STATION_MAPPING, regionTrainStationMap);stationDetails = new ArrayList<>();stationDetails.add(regionTrainStationMap.get(requestParam.getFromStation()));stationDetails.add(regionTrainStationMap.get(requestParam.getToStation()));}} finally {lock.unlock();}}

查询列车站点信息

采用Redis而不是Elasticsearch，因为搜索只允许选择一天的出发日期。

同时在12306网站上尝试，虽然页面上有很多查询条件，但大多数条件都是由前端进行筛选，实际上并没有触发后端的请求，发现只有在点击“查询”按钮时才会真正触发后端的请求，而点击页面上的其他筛选条件并不会向后端发出请求。

列车站点数据存入 Redis 中，结构如下：

具体查询代码：

List<TicketListDTO> seatResults = new ArrayList<>();String buildRegionTrainStationHashKey = String.format(REGION_TRAIN_STATION, stationDetails.get(0), stationDetails.get(1));Map<Object, Object> regionTrainStationAllMap = stringRedisTemplate.opsForHash().entries(buildRegionTrainStationHashKey);if (MapUtil.isEmpty(regionTrainStationAllMap)) {RLock lock = redissonClient.getLock(LOCK_REGION_TRAIN_STATION);lock.lock();try {regionTrainStationAllMap = stringRedisTemplate.opsForHash().entries(buildRegionTrainStationHashKey);if (MapUtil.isEmpty(regionTrainStationAllMap)) {LambdaQueryWrapper<TrainStationRelationDO> queryWrapper = Wrappers.lambdaQuery(TrainStationRelationDO.class).eq(TrainStationRelationDO::getStartRegion, stationDetails.get(0)).eq(TrainStationRelationDO::getEndRegion, stationDetails.get(1));List<TrainStationRelationDO> trainStationRelationList = trainStationRelationMapper.selectList(queryWrapper);for (TrainStationRelationDO each : trainStationRelationList) {TrainDO trainDO = distributedCache.safeGet(TRAIN_INFO + each.getTrainId(),TrainDO.class,() -> trainMapper.selectById(each.getTrainId()),ADVANCE_TICKET_DAY,TimeUnit.DAYS);TicketListDTO result = new TicketListDTO();result.setTrainId(String.valueOf(trainDO.getId()));result.setTrainNumber(trainDO.getTrainNumber());result.setDepartureTime(convertDateToLocalTime(each.getDepartureTime(), "HH:mm"));result.setArrivalTime(convertDateToLocalTime(each.getArrivalTime(), "HH:mm"));result.setDuration(DateUtil.calculateHourDifference(each.getDepartureTime(), each.getArrivalTime()));result.setDeparture(each.getDeparture());result.setArrival(each.getArrival());result.setDepartureFlag(each.getDepartureFlag());result.setArrivalFlag(each.getArrivalFlag());result.setTrainType(trainDO.getTrainType());result.setTrainBrand(trainDO.getTrainBrand());if (StrUtil.isNotBlank(trainDO.getTrainTag())) {result.setTrainTags(StrUtil.split(trainDO.getTrainTag(), ","));}long betweenDay = cn.hutool.core.date.DateUtil.betweenDay(each.getDepartureTime(), each.getArrivalTime(), false);result.setDaysArrived((int) betweenDay);result.setSaleStatus(new Date().after(trainDO.getSaleTime()) ? 0 : 1);result.setSaleTime(convertDateToLocalTime(trainDO.getSaleTime(), "MM-dd HH:mm"));seatResults.add(result);regionTrainStationAllMap.put(CacheUtil.buildKey(String.valueOf(each.getTrainId()), each.getDeparture(), each.getArrival()), JSON.toJSONString(result));}stringRedisTemplate.opsForHash().putAll(buildRegionTrainStationHashKey, regionTrainStationAllMap);}} finally {lock.unlock();}}

查询出来列车基本信息后，开始对列车按照出发时间进行排序。

seatResults = CollUtil.isEmpty(seatResults)? regionTrainStationAllMap.values().stream().map(each -> JSON.parseObject(each.toString(), TicketListDTO.class)).toList(): seatResults;
seatResults = seatResults.stream().sorted(new TimeStringComparator()).toList();

查询列车余票信息

列车基本信息已经全部填充完毕了，接下来就是查询列车余票信息并填充到基本信息中。

列车余票数据是实时变更的，如果在存储到基本信息中，就没办法变更了，所以单独存储。

for (TicketListDTO each : seatResults) {String trainStationPriceStr = distributedCache.safeGet(String.format(TRAIN_STATION_PRICE, each.getTrainId(), each.getDeparture(), each.getArrival()),String.class,() -> {LambdaQueryWrapper<TrainStationPriceDO> trainStationPriceQueryWrapper = Wrappers.lambdaQuery(TrainStationPriceDO.class).eq(TrainStationPriceDO::getDeparture, each.getDeparture()).eq(TrainStationPriceDO::getArrival, each.getArrival()).eq(TrainStationPriceDO::getTrainId, each.getTrainId());return JSON.toJSONString(trainStationPriceMapper.selectList(trainStationPriceQueryWrapper));},ADVANCE_TICKET_DAY,TimeUnit.DAYS);List<TrainStationPriceDO> trainStationPriceDOList = JSON.parseArray(trainStationPriceStr, TrainStationPriceDO.class);List<SeatClassDTO> seatClassList = new ArrayList<>();trainStationPriceDOList.forEach(item -> {String seatType = String.valueOf(item.getSeatType());String keySuffix = StrUtil.join("_", each.getTrainId(), item.getDeparture(), item.getArrival());Object quantityObj = stringRedisTemplate.opsForHash().get(TRAIN_STATION_REMAINING_TICKET + keySuffix, seatType);int quantity = Optional.ofNullable(quantityObj).map(Object::toString).map(Integer::parseInt).orElseGet(() -> {Map<String, String> seatMarginMap = seatMarginCacheLoader.load(String.valueOf(each.getTrainId()), seatType, item.getDeparture(), item.getArrival());return Optional.ofNullable(seatMarginMap.get(String.valueOf(item.getSeatType()))).map(Integer::parseInt).orElse(0);});seatClassList.add(new SeatClassDTO(item.getSeatType(), quantity, new BigDecimal(item.getPrice()).divide(new BigDecimal("100"), 1, RoundingMode.HALF_UP), false));});each.setSeatClassList(seatClassList);}

构建列车返回数据

查看 12306 列车查询页可知，会存在不同的查询条件，这些查询条件都是通过本次查询所有列车数据构建出来的。不同地区的不同查询列车数据，车次类型、出发车站、到达车站以及车次席别都不同。

接下来就是通过构建者模式构建列车查询返回数据。

return TicketPageQueryRespDTO.builder().trainList(seatResults).departureStationList(buildDepartureStationList(seatResults)).arrivalStationList(buildArrivalStationList(seatResults)).trainBrandList(buildTrainBrandList(seatResults)).seatClassTypeList(buildSeatClassList(seatResults)).build();