NoSQL: MongoDB快速上手

一、introduction

1. 基本概念

二、 相关操作

1. 基本操作

1. 终端启动： start_mongo
2. 查看版本：db.version()
3. 查看数据库list： show dbs
4. 创建database< training >： use training
5. 创建collection： db.createCollection("mycollection")
6. 查看创建好的collection：show collections
7. 插入一条document：db.mycollection.insert({"color":"blue","example":"sky"})
8. 查看documents数量：db.mycollection.count()
9. 列出所有document：db.mycollection.find()
10. 退出服务器： exit

2. create、read、update、delete

本部分采用的collection名字为languages

首先插入一些数据：

db.languages.insert({"name":"java","type":"object oriented"})
db.languages.insert({"name":"python","type":"general purpose"})
db.languages.insert({"name":"scala","type":"functional"})
db.languages.insert({"name":"c","type":"procedural"})
db.languages.insert({"name":"c++","type":"object oriented"})

read

1. 列出第一条doc：db.languages.findOne()
2. 列出所有doc：db.languages.find()
3. 列出前三个doc：db.languages.find().limit(3)
4. 列出对应标签的doc：db.languages.find({"name":"python"})
5. 列出所有的doc，但是仅仅显示name：db.languages.find({},{"name":1})
6. 列出所有doc，但是不显示name：db.languages.find({},{"name":0})
7. 列出所有面向对象语言名字： db.languages.find({"type":"object oriented"},{"name":1})

update

通用语法：

db.collection.updateMany({what documents to find},{$set:{what fields to set}})

举例：

1. 给所有doc多插入一块相同的内容：
   db.languages.updateMany({},{$set:{"description":"programming language"}})
2. 为python对应的doc添加作者：
   db.languages.updateMany({"name":"python"},{$set:{"creator":"Guido van Rossum"}})

delete

1. 删除对应名字的doc：db.languages.remove({"name":"scala"})
2. 删除对应类型的doc：db.languages.remove({"type":"object oriented"})
3. 删除全部docs：db.languages.remove({})

总结：
CRUD四种操作，基本模式使用{key:value}来进行筛选要改动的doc
其中update稍微复杂一点，需要先找到，然后再更改，改的部分，需要使用$set标识出来。

3. indexes

bigdata是学习本部分创建的Collection，通过如下循环创建了20万个doc：

use training
for (i=1;i<=200000;i++){print(i);db.bigdata.insert({"account_no":i,"balance":Math.round(Math.random()*1000000)})}

1. 创建index： db.bigdata.createIndex({"account_no":1})
   {}中通常选取最常用的field
2. 查询所有的indexes: db.bigdata.getIndexes()
3. 删除index：db.bigdata.dropIndex({"account_no":1})

4. aggregation framework

marks是学习本部分创建的Collection，具体数据内容如下：

use training
db.marks.insert({"name":"Ramesh","subject":"maths","marks":87})
db.marks.insert({"name":"Ramesh","subject":"english","marks":59})
db.marks.insert({"name":"Ramesh","subject":"science","marks":77})
db.marks.insert({"name":"Rav","subject":"maths","marks":62})
db.marks.insert({"name":"Rav","subject":"english","marks":83})
db.marks.insert({"name":"Rav","subject":"science","marks":71})
db.marks.insert({"name":"Alison","subject":"maths","marks":84})
db.marks.insert({"name":"Alison","subject":"english","marks":82})
db.marks.insert({"name":"Alison","subject":"science","marks":86})
db.marks.insert({"name":"Steve","subject":"maths","marks":81})
db.marks.insert({"name":"Steve","subject":"english","marks":89})
db.marks.insert({"name":"Steve","subject":"science","marks":77})
db.marks.insert({"name":"Jan","subject":"english","marks":0,"reason":"absent"})

1. limit 限制输出行数：
   db.marks.aggregate([{"$limit":2}])
   仅仅输出前两行的内容
2. sort 基于column：
   db.marks.aggregate([{"$sort":{"marks":1}}])
   命令行按照mark分数升序输出，如果{"marks":-1}那么将会按照降序顺序输出。
3. sort和limit配合，输出成绩最好或者最差的：
   db.marks.aggregate([ {"$sort":{"marks":-1}}, {"$limit":2} ])
   像pipeline一样，在[]中按照顺序排开，两种operator中用逗号隔开。
   注意： 每个operator都要上引号
4. 输出各个学科的平均分数：
   group 经常和 sum,avg,min,max等配合使用

   db.marks.aggregate([
   {
       "$group":{
           "_id":"$subject",
           "average":{"$avg":"$marks"}
           }
   }
   ])
   

   上式与如下sql相同含义：

   select subject, avg(marks) from marks
   group by subject;
   

   group中的tag，_id不能改

最后，所有如上的操作，都可以像pipeline一样，在[]中按照逗号排开。

5. replication $ Sharding

redundancy
availibility
fault

sharding:

5. 使用python连接mongoDB

连接数据库

from pymongo import MongoClient
user = 'root'
password = 'MjQwOTgtcnNhbm5h'  # 改为自己的password
host='localhost'
#create the connection url
connecturl = "mongodb://{}:{}@{}:27017/?authSource=admin".format(user,password,host)


# connect to mongodb server
print("Connecting to mongodb server")
connection = MongoClient(connecturl)


# get database list
print("Getting list of databases")
dbs = connection.list_database_names()

# print the database names

for db in dbs:
    print(db)
print("Closing the connection to the mongodb server")


# close the server connecton
connection.close()

操作文件

在连接好数据库之后，可以与数据库交互：

connection = MongoClient(connecturl)


# select the 'training' database 

db = connection.training

# select the 'python' collection 

collection = db.python

# create a sample document

doc = {"lab":"Accessing mongodb using python", "Subject":"No SQL Databases"}

# insert a sample document

print("Inserting a document into collection.")
db.collection.insert(doc)

# query for all documents in 'training' database and 'python' collection

docs = db.collection.find()

print("Printing the documents in the collection.")

for document in docs:
    print(document)