pymongo操作mongodb

1. 连接数据库

# 方式一
client = pymongo.MongoClient(host='localhost', port=27017)

# 方式2: URI = "mongodb://root:123456@localhost:27017"
URI = "mongodb://localhost:27017"
client = pymongo.MongoClient(URI)

2. 数据库操作

all_dbs = client.list_database_names()		# 查询所有数据库
db = client.testdb		# 使用testdb库
client.drop_database("testdb")		# 删除数据库

3. 集合操作

db = client.test
colls = db.list_collections()		# 查询test库里所有的集合
coll = db.user		# 选取user集合
db.drop_collection("user")		# 删除user集合

4. 索引操作

client = pymongo.MongoClient(host='localhost', port=27017)
db = client.test
user_coll = db.user		

# 创建单列升序索引
sing_asc_idx = user_coll.create_index([("name", pymongo.ASCENDING)])
print(sing_asc_idx)     # name_1

# 创建单列降序索引
sing_desc_idx = user_coll.create_index([("age", pymongo.DESCENDING)])
print(sing_desc_idx)    # age_-1

# 创建联合索引
union_idx = user_coll.create_index([("name", pymongo.ASCENDING), ("age", pymongo.DESCENDING)])
print(union_idx)        # name_1_age_-1

# 创建唯一索引
unique_idx = user_coll.create_index([("name", pymongo.ASCENDING)], unique=True)
print(unique_idx)       # name_1

# 删除某个索引
user_coll.drop_index("name_1")
# 删除全部索引
user_coll.drop_indexes()

# 为一个集合同时创建多个索引
uname = pymongo.IndexModel([("name", pymongo.DESCENDING)])
uage = pymongo.IndexModel([("age", pymongo.ASCENDING)])
idx_res = user_coll.create_indexes([uname, uage])
print(idx_res)      # ['name_-1', 'age_1']

5. 文档操作

• 增

client = pymongo.MongoClient(host='localhost', port=27017)
user_coll = client.test.user

# 单值插入
doc = {"name": "lihua", "age": 25, "gender": "male"}
res = user_coll.insert_one(doc)
insert_status = res.acknowledged        # 插入成功与否：True
insert_id = res.inserted_id             # 插入之后的id：64858e451beea93d5224e44e

# 批量插入
docs = [{"name": "zhangli", "age": 28, "gender": "female"}, 
		{"name": "lizhu", "age": 30, "gender": "female"}]
res = user_coll.insert_many(docs)
insert_status = res.acknowledged         # True
insert_id = res.inserted_ids             # [ObjectId('64858e451beea93d5224e44e'), ObjectId('64858e451beea93d5224e44f')]

• 删

client = pymongo.MongoClient(host='localhost', port=27017)
user_coll = client.test.user

# 单条删除，只删除第一条
del_res = user_coll.delete_one({"name": "zhangsan"})
print(del_res.acknowledged)     # True
print(del_res.deleted_count)    # 1

# 批量删除
del_res = user_coll.delete_many({"name": "zhangsan"})
print(del_res.acknowledged)     # True
print(del_res.deleted_count)   	# 2

• 改

# 修改单条
update_res = user_coll.update_one(filter={"name": "lisi"}, update={"$set": {"age": 20, "gender": "male"}})
print(update_res.acknowledged)      # True
print(update_res.matched_count)     # 1
print(update_res.modified_count)    # 1
print(update_res.raw_result)        # {'n': 1, 'nModified': 1, 'ok': 1.0, 'updatedExisting': True}
print(update_res.upserted_id)       # None

# upsert=True: 不存在该条记录的话就插入
update_res = user_coll.update_one(filter={"name": "lisi"}, update={"$set": {"age": 20, "gender": "male"}}, upsert=True)

# 修改多条
update_res = user_coll.update_many(filter={"name": "lisi"}, update={"$set": {"age": 20, "gender": "male"}}, upsert=True)

• 查

  client = pymongo.MongoClient(host='localhost', port=27017)
  book_coll = client.test.books
  

  • and

  # 书籍类型为technology，并且作者名字为 xx009 的书籍
  res = book_coll.find({"type": "technology", "author.name": "xx009"})	
  res = book_coll.find({"$and": [{"type": "technology"}, {"author.name": "xx009"}]})
  两种方式是等价的
  

  • or

  res = book_coll.find({"$or": [{"type": "technology"}, {"type": "travel"}]})
  

  • and or

  # 作者xx009和作者xx005的旅游类书籍
  res = book_coll.find({"type": "travel",
                        "$or": [{"author.name": "xx005"}, {"author.name": "xx007"}]
                        })
  

  • 运算符

  运算符	含义
  $gt	>
  $gte	>=
  $lt	<
  $lte	<=
  $eq	==
  $ne	<>
  $in	in
  $nin	not in
  $not	not
  $exists	是否存在

  book_coll.find({"author.age": {"$lte": 30}})		# 作者年龄小于30岁的
  book_coll.find({"type": {"$nin": ["travel", "literature", "technology"]}})	# 查询不在这几种类型里的book
  book_coll.find({"type": {"$exists": True}})		# 查询有type字段的数据
  

• 复杂查询

  • 匹配数组字段中的单个元素

  res = book_coll.find({"tag": "mongodb"})    # 只要数组中有该元素，就输出
  

  • 匹配数组字段中的第N个元素

  res = book_coll.find({"tag.0": "mongodb"})	# tag中第一个元素为mongodb的book
  

  • 数组字段中至少有一个元素满足要求即可

  res = book_coll.find({"score": {"$elemMatch": {"$gt": 15, "$lt": 20}}})		# 各平台得分中有一个大于15或者小于20即可
  

  • 正则筛选

  res = book_coll.find({"type": {"$regex": "chno"}})		# type字段包含'chno'的数据
  res = book_coll.find({"type": {"$regex": "^tech"}})		# type字段以'tech'开头的数据
  res = book_coll.find({"type": {"$regex": "logy$"}})		# type字段'logy'结尾的数据
  

  • 投影

  res = book_coll.find({}, {"title": 1, "type": 1, "tag": 1, "author.name": 1})	# 只显示这几列
  

• 聚合操作

  # 每种类别的作者年龄平均值
  pipelines = [
      {"$group": {"_id": "$type", "avg_age": {"$avg": "$author.age"}}}
  ]
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': 'novel', 'avg_age': 25.615384615384617}
  {'_id': 'technology', 'avg_age': 25.9}
  {'_id': 'literature', 'avg_age': 25.625}
  {'_id': 'travel', 'avg_age': 30.333333333333332}
  {'_id': 'sociality', 'avg_age': 25.7}
  

  # 每种类别的数的受欢迎程度总和
  pipelines = [
      {"$group": {"_id": "$type", "sum_pop": {"$sum": "$favCount"}}}
  ]
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': 'novel', 'sum_pop': 603.0}
  {'_id': 'technology', 'sum_pop': 594.0}
  {'_id': 'literature', 'sum_pop': 329.0}
  {'_id': 'travel', 'sum_pop': 495.0}
  {'_id': 'sociality', 'sum_pop': 454.0}
  

  # 每种类型，每种标签的book总和
  pipelines = [
      {"$group": {"_id": {"outType": "$type", "outTag": "$tag"}, "bookCount": {"$sum": 1}}}
  ]
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': {'outType': 'sociality', 'outTag': ['developer', 'popular']}, 'bookCount': 1}
  {'_id': {'outType': 'travel', 'outTag': ['document', 'mongodb']}, 'bookCount': 1}
  {'_id': {'outType': 'travel', 'outTag': ['mongodb', 'popular']}, 'bookCount': 1}
  {'_id': {'outType': 'travel', 'outTag': ['document', 'developer']}, 'bookCount': 1}
  {'_id': {'outType': 'sociality', 'outTag': ['document', 'popular']}, 'bookCount': 1}
  {'_id': {'outType': 'technology', 'outTag': ['developer', 'developer']}, 'bookCount': 1}
  {'_id': {'outType': 'novel', 'outTag': ['popular', 'nosql']}, 'bookCount': 2}
  {'_id': {'outType': 'technology', 'outTag': ['mongodb', 'developer']}, 'bookCount': 1}
  {'_id': {'outType': 'literature', 'outTag': ['mongodb', 'document']}, 'bookCount': 1}
  {'_id': {'outType': 'novel', 'outTag': ['popular', 'document']}, 'bookCount': 2}
  

  # 查询每个作者的book的平均受欢迎程度，并按照降序排序
  pipelines = [
      {"$group": {"_id": "$author.name", "avg_pop": {"$avg": "$favCount"}}},
      {"$sort": {"avg_pop": -1}}
  ]
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': 'xx001', 'avg_pop': 71.66666666666667}
  {'_id': 'xx004', 'avg_pop': 70.4}
  {'_id': 'xx008', 'avg_pop': 60.857142857142854}
  {'_id': 'xx000', 'avg_pop': 58.0}
  {'_id': 'xx002', 'avg_pop': 48.0}
  {'_id': 'xx007', 'avg_pop': 44.5}
  

  # 按照不同类型book的平均受欢迎程度、作者的平均年龄以及book数量分组并排序，最后输出指定字段。
  pipelines = [
      {"$group": {
          "_id": "$type",
          "avg_pop": {"$avg": "$favCount"},
          "avg_age": {"$avg": "$author.age"},
          "sum_book": {"$sum": 1}
      }},
      {"$project": {"_id": 1, "avg_pop": 1}},
      {"$sort": {"avg_pop": -1}}
  ]
  
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': 'technology', 'avg_pop': 59.4}
  {'_id': 'travel', 'avg_pop': 55.0}
  {'_id': 'novel', 'avg_pop': 46.38461538461539}
  {'_id': 'sociality', 'avg_pop': 45.4}
  {'_id': 'literature', 'avg_pop': 41.125}
  

  # 不同tag的book，其受欢迎程度排序
  pipelines = [
      {"$unwind": "$tag"},	# 数组里的N个元素拆分成N行
      {"$group": {"_id": "$tag", "avg_pop": {"$avg": "$favCount"}, "book_count": {"$sum": 1}}},
      {"$sort": OrderedDict(**{"avg_pop": -1, "book_count": 1})}
  ]
  
  res = book_coll.aggregate(pipelines)
  
  output:
  {'_id': 'nosql', 'avg_pop': 58.476190476190474, 'book_count': 21}
  {'_id': 'mongodb', 'avg_pop': 51.5, 'book_count': 20}
  {'_id': 'document', 'avg_pop': 49.142857142857146, 'book_count': 21}
  {'_id': 'popular', 'avg_pop': 47.18181818181818, 'book_count': 22}
  {'_id': 'developer', 'avg_pop': 38.875, 'book_count': 16}