How does Net Core perform a multi conditional query against MongoDB

MySQL database was basically used in previous projects. Recently, leader decided to migrate the log module to MongoDB with better insert/query performance. The writing method of multi-condition query really took some effort, and wrote a record.

1. Preparations

1. The installation process is not repeated

2. Add ReferencePackage

dotnet add package mongodb.bson
dotnet add package mongodb.driver

3. appsetting. json Add Connection Configuration

"MongodbHost": {
  "Connection": "mongodb://[username]:[password]@[ip]:[port]",
  "DataBase": "[database]",
  "Table": ""
 },

4. Method to get MongoDBConfig

public static MongodbHostOptions MongodbConfig()
{
  var builder = new ConfigurationBuilder()
        .SetBasePath(Directory.GetCurrentDirectory())
        .AddJsonFile("appsettings.json");

 　　 IConfiguration Configuration = builder.Build();
 　　 var option = Configuration.GetSection("MongodbHost");
 　　 return new MongodbHostOptions { Connection = option["Connection"], DataBase = option["DataBase"], Table = option["Table"] };
}

2. Query method

The query method here is found online, directly used. If it is a single 1 data source, here host can be extracted into helper class attributes.

#region FindListByPage  Paged query set 
  /// <summary>
  ///  Paged query set 
  /// </summary>
  /// <param name="filter"> Query criteria </param>
  /// <param name="pageIndex"> Current page </param>
  /// <param name="pageSize"> Page capacity </param>
  /// <param name="count"> Total number of articles </param>
  /// <param name="field"> Fields to query , Query all when not writing </param>
  /// <param name="sort"> Fields to sort </param>
  /// <returns></returns>
  public static List<T> FindListByPage(FilterDefinition<T> filter, int pageIndex, int pageSize, out int count, string[] field = null, SortDefinition<T> sort = null)
  {
    try
    {
      MongodbHostOptions host = Tools.AppSettingsTools.MongodbConfig();
      host.Table = "WSMessageLog";
      var client = MongodbClient<T>.MongodbInfoClient(host);
      count = Convert.ToInt32(client.CountDocuments(filter));
      // Do not specify query fields 
      if (field == null || field.Length == 0)
      {
        if (sort == null) return client.Find(filter).Skip((pageIndex - 1) * pageSize).Limit(pageSize).ToList();
        // Sort 
        return client.Find(filter).Sort(sort).Skip((pageIndex - 1) * pageSize).Limit(pageSize).ToList();
      }

      // Specify query fields 
      var fieldList = new List<ProjectionDefinition<T>>();
      for (int i = 0; i < field.Length; i++)
      {
        fieldList.Add(Builders<T>.Projection.Include(field[i].ToString()));
      }
      var projection = Builders<T>.Projection.Combine(fieldList);
      fieldList?.Clear();

      // Do not sort 
      if (sort == null) return client.Find(filter).Project<T>(projection).Skip((pageIndex - 1) * pageSize).Limit(pageSize).ToList();

      // Sort query 
      return client.Find(filter).Sort(sort).Project<T>(projection).Skip((pageIndex - 1) * pageSize).Limit(pageSize).ToList();

    }
    catch (Exception ex)
 　　 {
  　　  　　  throw ex;
 　　 }
}
#endregion

3. Call the query method

There is also a pit. The time stored in MongoDB is Greenwich Mean Time. Insert 8:00, and you will find that it becomes 0:00 when querying, so you need to add a label when defining the time attribute

[BsonDateTimeOptions(Kind = DateTimeKind.Local)]
public DateTime logtime { get; set; }

Here, OprLogModel is a class that defines query conditions.

public static LogPager<log_operate> Get_operate_log_mongo(OprLogModel qModel)
{
 　　 LogPager<log_operate> pager = new LogPager<log_operate>();
 　　 FilterDefinition<log_operate> filters;

 　　  var sortbuilder = Builders<log_operate>.Sort;
 　　 var sort = sortbuilder.Descending("operate_time");

 　　 #region  User rights filtering 
 　　 IEnumerable<string> IdList = dev_deviceRepository.GetBinding(qModel.user_id);
 　　  filters = Builders<log_operate>.Filter.In("device_id", IdList);
 　　 #endregion

 　　 if (!string.IsNullOrEmpty(qModel.device_id))
  {
   　　 var filters_did = Builders<log_operate>.Filter.Eq("device_id", qModel.device_id);
   　　 filters = Builders<log_operate>.Filter.And(filters, filters_did);
 　　 }
 　　 if (qModel.sDate != null)
 　　 {
  　　  　　  var filters_sdate = Builders<log_operate>.Filter.Gte<DateTime>("operate_time", Convert.ToDateTime(qModel.sDate));
  　　  　　  filters = Builders<log_operate>.Filter.And(filters, filters_sdate);
 　　 }
 　　 if (qModel.eDate != null)
 　　 {
  　　  　　  var filters_edate = Builders<log_operate>.Filter.Lte<DateTime>("operate_time", Convert.ToDateTime(qModel.eDate));
  　　  　　  filters = Builders<log_operate>.Filter.And(filters, filters_edate);
 　　 }
 　　 int total;
 　　 pager.data = MongoTools<log_operate>.FindListByPage(filters, qModel.pageindex, (qModel.pageindex - 1) * qModel.pagesize, out total, null, sort);
 　　 pager.total = total;
 　　 return pager;
}
#endregion

You can also define an empty filterdefinition first, and then aggregate it with each query condition through And:

FilterDefinition<log_operate> filters = FilterDefinition<log_operate>.Empty;
var filters_idlist = Builders<log_operate>.Filter.In("device_id", IdList);
filters = Builders<log_operate>.Filter.And(filters, filters_idlist);

Above is. Net Core how to MongoDB multi-condition query details, more about. Net Core to MongoDB multi-condition query information please pay attention to other related articles on this site!