（轉）python 全棧開發，Day74(基於雙下劃線的跨表查詢,聚合查詢,分組查詢,F查詢,Q查詢)

阿新 • • 發佈：2018-12-22

昨日內容回顧

# 一對多的新增方式1（推薦）
# book=Book.objects.create(title="水滸傳",price=100,pub_date="1643-4-12",publish_id=1)
# print(book.title)

# 一對多的新增方式2  # publish必須接受一個物件
# xigua=Publish.objects.filter(name="西瓜出版社").first()
# book=Book.objects.create(title="三國演義",price=300,pub_date="1643-4-12",publish=xigua) 

# print(book.title)

# ###
# book=Book(title="三國演義",price=300,pub_date="1643-4-12",publish=xigua)
# book.save()
#
#
# ###
# book=Book()
# book.title="三國演義"
# book.price=300
# book.pub_date="1643-4-12"
# book.publish=xigua

# book.publish   # 與這本書籍關聯的出版社物件
# print(book.publish)
# print(book.publish_id)

# 
###############################################################################
# 新增多對多關係

# book=Book.objects.create(title="python葵花寶典",price=122,pub_date="2012-12-12",publish_id=1)
# alex=Author.objects.filter(name="alex").first()
# egon=Author.objects.filter(name="egon").first()
# jinxin=Author.objects.filter(name="jinxin").first() 

# book.publish=xigua
# book.authors.add(alex,egon)
# book.authors.add(*[alex,egon])

# author_list=Author.objects.all() #  [obj,.....]
# book.authors.add(*author_list)

# 解除繫結的關係

# book=Book.objects.filter(id=4).first()
# book.authors.clear()
# book.authors.add(jinxin)

# book.authors.set([jinxin,])

# book.authors.remove(alex,egon)
# book.authors.remove(*[alex,egon])
# book.authors.clear()

# book.authors.add(1,2)
# book.authors.add(*[1,2])

#############################################
# print(book.publish)          #  與這本書籍關聯的出版社物件
# print(book.authors.all())    # 與這本書關聯的作者物件queryset物件集合 <QuerySet [<Author: alex>, <Author: egon>, <Author: jinxin>]>
# #############################################
# print(book.authors) # app01.Author.None
# print(type(book.authors))

# print(book.pub_date) # 2012-12-12     2012-12-12 00:00:00+00:00
# print(type(book.pub_date)) # <class 'datetime.date'>  <class 'datetime.datetime'>

View Code

一、基於雙下劃線的跨表查詢

Django 還提供了一種直觀而高效的方式在查詢(lookups)中表示關聯關係，它能自動確認 SQL JOIN 聯絡。要做跨關係查詢，就使用兩個下劃線來連結模型(model)間關聯欄位的名稱，直到最終連結到你想要的model 為止。

核心得學會通知ORM引擎什麼時候，join哪張表

join看似複雜，實則最簡單。因為把欄位列出來之後，就相當於單表操作了！想怎麼取值都可以！

正向查詢按欄位,反向查詢按表名小寫用來告訴ORM引擎join哪張表

返回值是QuerySet

一對多：

正向查詢

返回結構是queryset()

正向查詢：關聯屬性在book表中，所以book物件找關聯出版社物件，正向查詢
反向查詢：關聯屬性在book表中，所以publish物件找關聯書籍，反向查詢

       按欄位：xx
book  ------------------ >  publish
      <--------------------
      按表名小寫__欄位名。比如publish__name

舉例：查詢西遊記這本書的出版社名字

先使用原生sql查詢

SELECT app01_publish.name from app01_book
INNER JOIN app01_publish on 
app01_book.publish_id = app01_publish.id
WHERE app01_book.title = '西遊記'

View Code

執行結果為：榴蓮出版社

它的步驟為：先連表，再過濾

使用orm引擎查詢

修改settings.py，最後一行新增。表示開啟日誌

LOGGING = {
    'version': 1,
    'disable_existing_loggers': False,
    'handlers': {
        'console':{
            'level':'DEBUG',
            'class':'logging.StreamHandler',
        },
    },
    'loggers': {
        'django.db.backends': {
            'handlers': ['console'],
            'propagate': True,
            'level':'DEBUG',
        },
    }
}

View Code

修改urls.py，增加路徑query

urlpatterns = [
    path('admin/', admin.site.urls),
    path('add/', views.add),
    path('query/', views.query),
]

View Code

修改views.py，增加query檢視函式

def query(request):
    ret = Book.objects.filter(title='西遊記').values("publish__name")
    print(ret)

View Code

解釋：

Book.objects 表示基礎表，它是鏈式程式設計的開始。

publish__name 表示publish表下的name欄位。因為要的name欄位，它不在Book表中。那麼指定外部表示，需要加雙下劃線。注意：此表必須要和Book表有關聯！

訪問url：http://127.0.0.1:8000/query/

檢視Pycharm控制檯，輸出：

<QuerySet [{'publish__name': '榴蓮出版社'}]>
(0.000) SELECT "app01_publish"."name" FROM "app01_book" INNER JOIN "app01_publish" ON ("app01_book"."publish_id" = "app01_publish"."id") WHERE "app01_book"."title" = '西遊記' LIMIT 21; args=('西遊記',)

View Code

可以看出，ORM執行的sql和手寫的sql，大致是一樣的！

反向查詢

舉例：查詢出版過西遊記的出版社

def query(request):
    ret = Publish.objects.filter(book__title="西遊記").values("name")
    print(ret)

    return HttpResponse('查詢成功')

View Code

解釋：

book__title 表示book表中的title欄位，它不需要加引號

values("name") 表示publish表的中name欄位，為什麼呢？因為基礎表示publish，它可以直接取name欄位！

重新整理頁面，檢視控制檯輸出：

(0.001) SELECT "app01_publish"."name" FROM "app01_publish" INNER JOIN "app01_book" ON ("app01_publish"."id" = "app01_book"."publish_id") WHERE "app01_book"."title" = '西遊記' LIMIT 21; args=('西遊記',)
<QuerySet [{'name': '榴蓮出版社'}]>

View Code

查詢結果上面的例子是一樣的。

多對多查詢

正向查詢：關聯屬性在book表中，所以book物件找關聯作者集合，正向查詢
反向查詢：關聯屬性在book表中，所以author物件找關聯書籍，反向查詢

      正 按欄位：xx
book  ------------------------- >  author
      <-------------------------
      反 按表名小寫__欄位名

正向查詢

舉例：查詢西遊記這本書籍的所有作者的姓名和年齡

先用原生sql查詢

SELECT app01_author.name,app01_author.age from app01_book
INNER JOIN app01_book_authors on 
app01_book_authors.book_id = app01_book.id INNER JOIN app01_author on
app01_book_authors.author_id = app01_author.id
WHERE app01_book.title = '西遊記'

View Code

涉及到3表查詢，查詢結果為：

使用orm引擎查詢

def query(request):
    ret = Book.objects.filter(title="西遊記").values("authors__name","authors__age")
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'authors__name': 'xiao', 'authors__age': 25}]>
(0.001) SELECT "app01_author"."name", "app01_author"."age" FROM "app01_book" LEFT OUTER JOIN "app01_book_authors" ON ("app01_book"."id" = "app01_book_authors"."book_id") LEFT OUTER JOIN "app01_author" ON ("app01_book_authors"."author_id" = "app01_author"."id") WHERE "app01_book"."title" = '西遊記' LIMIT 21; args=('西遊記',)

View Code

解釋：

由於book表和author表是多對多關係，所以使用ORM查詢時，它會自動對應關係

authors__name 表示author表的name欄位

那麼使用ORM處理多表，就顯得很簡單了！

反向查詢

還是上面的需求，以author為基礎表查詢

def query(request):
    ret = Author.objects.filter(book__title="西遊記").values("name","age")
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'name': 'xiao', 'age': 25}]>
(0.002) SELECT "app01_author"."name", "app01_author"."age" FROM "app01_author" INNER JOIN "app01_book_authors" ON ("app01_author"."id" = "app01_book_authors"."author_id") INNER JOIN "app01_book" ON ("app01_book_authors"."book_id" = "app01_book"."id") WHERE "app01_book"."title" = '西遊記' LIMIT 21; args=('西遊記',)

View Code

執行結果同上！

一對一

正向查詢：關聯屬性在authordetail表中，所以author物件找關聯作者詳情，正向查詢
反向查詢：關聯屬性在author表中，所以authordetail物件找關聯作者資訊，反向查詢

        正向: 按欄位：.ad
author  ------------------------- >  authordetail
      <-------------------------
        反向: 按表名小寫  authordetail_obj.author

正向查詢

舉例：查詢xiao的女朋友名字

def query(request):
    ret = Author.objects.filter(name="xiao").values("ad__gf")
    print(ret)

    return HttpResponse('查詢成功')

View Code

解釋：ORM查詢時，會自動對應關係。ad__gf表示authordetail表的gf欄位

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'ad__gf': '趙麗穎'}]>
(0.001) SELECT "app01_authordetail"."gf" FROM "app01_author" INNER JOIN "app01_authordetail" ON ("app01_author"."ad_id" = "app01_authordetail"."id") WHERE "app01_author"."name" = 'xiao' LIMIT 21; args=('xiao',)

View Code

反向查詢

還是上面的需求，以authordetail表為基礎表查詢

def query(request):
    ret = AuthorDetail.objects.filter(author__name="xiao").values("gf")
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'gf': '趙麗穎'}]>
(0.002) SELECT "app01_authordetail"."gf" FROM "app01_authordetail" INNER JOIN "app01_author" ON ("app01_authordetail"."id" = "app01_author"."ad_id") WHERE "app01_author"."name" = 'xiao' LIMIT 21; args=('xiao',)

View Code

進階練習(連續跨表)

舉例1

查詢榴蓮出版社出版過的所有書籍的名字以及作者的姓名

正向查詢

def query(request):
    ret = Book.objects.filter(publish__name="榴蓮出版社").values_list("title","authors__name")
    print(ret)

    return HttpResponse('查詢成功')

View Code

解釋：

book表示連接出版社和作者的核心表。以它為基礎表查詢，比較好處理！

publish__name 表示publish表的name欄位。

authors__name表示book_authors表(book和author的關係表)的name欄位。

重新整理頁面，檢視控制檯輸出：

<QuerySet [('西遊記', 'xiao')]>
(0.000) SELECT "app01_book"."title", "app01_author"."name" FROM "app01_book" INNER JOIN "app01_publish" ON ("app01_book"."publish_id" = "app01_publish"."id") LEFT OUTER JOIN "app01_book_authors" ON ("app01_book"."id" = "app01_book_authors"."book_id") LEFT OUTER JOIN "app01_author" ON ("app01_book_authors"."author_id" = "app01_author"."id") WHERE "app01_publish"."name" = '榴蓮出版社' LIMIT 21; args=('榴蓮出版社',)

View Code

反向查詢

還是上面的需求，以publish表為基礎表查詢

def query(request):
    ret = Publish.objects.filter(name="榴蓮出版社").values_list("book__title","book__authors__age","book__authors__name")
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

<QuerySet [('西遊記', 25, 'xiao')]>
(0.001) SELECT "app01_book"."title", "app01_author"."age", "app01_author"."name" FROM "app01_publish" LEFT OUTER JOIN "app01_book" ON ("app01_publish"."id" = "app01_book"."publish_id") LEFT OUTER JOIN "app01_book_authors" ON ("app01_book"."id" = "app01_book_authors"."book_id") LEFT OUTER JOIN "app01_author" ON ("app01_book_authors"."author_id" = "app01_author"."id") WHERE "app01_publish"."name" = '榴蓮出版社' LIMIT 21; args=('榴蓮出版社',)

View Code

舉例2

手機號以11開頭的作者出版過的所有書籍名稱以及出版社名稱

提示：涉及到5表查詢！

先使用原生sql查詢

SELECT app01_book.title,a01p.name from app01_book 
INNER JOIN app01_book_authors as a
on app01_book.id = a.book_id
INNER JOIN app01_author as a3
on a.author_id = a3.id
INNER JOIN app01_authordetail as a2
on a3.ad_id = a2.id
INNER JOIN app01_publish as a01p 
on app01_book.publish_id = a01p.id
WHERE a2.tel like '11%'

View Code

執行結果：

使用orm引擎查詢

正向查詢：

def query(request):
    ret = Book.objects.filter(authors__ad__tel__startswith="11").values("title","publish__name")
    print(ret)

    return HttpResponse('查詢成功')

View Code

解釋：

authors__ad__tel 表示book_authors表，author表的ad欄位，authordetail表的tel欄位，做關聯查詢。

__startswith 表示以什麼開頭，它會使用like查詢，比如'11%'

"title","publish__name" 分別表示book表的title欄位，publish表的name欄位

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'title': 'python', 'publish__name': '西瓜出版社'}, {'title': 'python', 'publish__name': '西瓜出版社'}, {'title': 'python', 'publish__name': '西瓜出版社'}, {'title': '西遊記', 'publish__name': '榴蓮出版社'}, {'title': '三國演義', 'publish__name': '西瓜出版社'}]>
(0.001) SELECT "app01_book"."title", "app01_publish"."name" FROM "app01_book" INNER JOIN "app01_book_authors" ON ("app01_book"."id" = "app01_book_authors"."book_id") INNER JOIN "app01_author" ON ("app01_book_authors"."author_id" = "app01_author"."id") INNER JOIN "app01_authordetail" ON ("app01_author"."ad_id" = "app01_authordetail"."id") INNER JOIN "app01_publish" ON ("app01_book"."publish_id" = "app01_publish"."id") WHERE "app01_authordetail"."tel" LIKE '11%' ESCAPE '\' LIMIT 21; args=('11%',)

View Code

反向查詢：

def query(request):
    ret = Author.objects.filter(ad__tel__startswith="11").values("book__title","book__publish__name")
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

<QuerySet [{'book__publish__name': '西瓜出版社', 'book__title': 'python'}, {'book__publish__name': '榴蓮出版社', 'book__title': '西遊記'}, {'book__publish__name': '西瓜出版社', 'book__title': 'python'}, {'book__publish__name': '西瓜出版社', 'book__title': '三國演義'}, {'book__publish__name': '西瓜出版社', 'book__title': 'python'}]>
(0.001) SELECT "app01_book"."title", "app01_publish"."name" FROM "app01_author" INNER JOIN "app01_authordetail" ON ("app01_author"."ad_id" = "app01_authordetail"."id") LEFT OUTER JOIN "app01_book_authors" ON ("app01_author"."id" = "app01_book_authors"."author_id") LEFT OUTER JOIN "app01_book" ON ("app01_book_authors"."book_id" = "app01_book"."id") LEFT OUTER JOIN "app01_publish" ON ("app01_book"."publish_id" = "app01_publish"."id") WHERE "app01_authordetail"."tel" LIKE '11%' ESCAPE '\' LIMIT 21; args=('11%',)

View Code

二、聚合查詢

聚合是aggreate(*args,**kwargs),通過QuerySet 進行計算。做求值運算的時候使用

主要有Sum,Avg,Max,Min。使用前，需要匯入模組

from django.db.models import Sum,Avg,Max,Min,Count

Sum (總和)

返回值是一個字典

舉例：查詢所有書籍的總價格

def query(request):
    ret = Book.objects.all().aggregate(Sum("price"))
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

{'price__sum': Decimal('522.00')}
(0.001) SELECT CAST(SUM("app01_book"."price") AS NUMERIC) AS "price__sum" FROM "app01_book"; args=()
[02/Jul/2018 21:45:26] "GET /query/ HTTP/1.1" 200 12

View Code

Avg (平均值)

返回值是一個字典

舉例：查詢所有書籍的平均價格

使用原生sql查詢

select avg(price) from app01_book

執行輸出：174

使用orm引擎查詢

def query(request):
    ret = Book.objects.all().aggregate(Avg("price"))
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

(0.001) SELECT AVG("app01_book"."price") AS "price__avg" FROM "app01_book"; args=()
[02/Jul/2018 21:38:42] "GET /query/ HTTP/1.1" 200 12
{'price__avg': 174.0}

View Code

aggregate()是QuerySet 的一個終止子句，意思是說，它返回一個包含一些鍵值對的字典。鍵的名稱是聚合值的識別符號，值是計算出來的聚合值。鍵的名稱是按照欄位和聚合函式的名稱自動生成出來的。如果你想要為聚合值指定一個名稱，可以向聚合子句提供它。

對返回的key起別名

def query(request):
    ret = Book.objects.all().aggregate(avg = Avg("price"))
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

{'avg': 174.0}
(0.001) SELECT AVG("app01_book"."price") AS "avg" FROM "app01_book"; args=()

View Code

Max 和Min

返回值是一個字典

如果你希望生成不止一個聚合，你可以向aggregate()子句中新增另一個引數。

舉例：所有圖書價格的最大值和最小值

def query(request):
    ret = Book.objects.aggregate(Avg('price'), Max('price'), Min('price'))
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

{'price__avg': 174.0, 'price__min': Decimal('100.00'), 'price__max': Decimal('300.00')}
(0.001) SELECT AVG("app01_book"."price") AS "price__avg", CAST(MIN("app01_book"."price") AS NUMERIC) AS "price__min", CAST(MAX("app01_book"."price") AS NUMERIC) AS "price__max" FROM "app01_book"; args=()

View Code

Count (統計結果行數)

舉例：查詢西瓜出版社總共出版過多少本書籍

def query(request):
    ret = Book.objects.all().aggregate(count = Count("id"))
    print(ret)

    return HttpResponse('查詢成功')

View Code

重新整理頁面，檢視控制檯輸出：

{'count': 3}
(0.001) SELECT COUNT("app01_book"."id") AS "count" FROM "app01_book"; args=()

View Code

三、分組查詢

分組：將查詢結果按照某個欄位或多個欄位進行分組。欄

（轉）python 全棧開發，Day74(基於雙下劃線的跨表查詢,聚合查詢,分組查詢,F查詢,Q查詢)

昨日內容回顧

一、基於雙下劃線的跨表查詢

一對多：

正向查詢

舉例：查詢西遊記這本書的出版社名字

反向查詢

舉例：查詢出版過西遊記的出版社

多對多查詢

正向查詢

舉例：查詢西遊記這本書籍的所有作者的姓名和年齡

反向查詢

一對一

正向查詢

舉例：查詢xiao的女朋友名字

反向查詢

進階練習(連續跨表)

舉例1

查詢榴蓮出版社出版過的所有書籍的名字以及作者的姓名

正向查詢

反向查詢

舉例2

手機號以11開頭的作者出版過的所有書籍名稱以及出版社名稱

正向查詢：

反向查詢：

二、聚合查詢

Sum (總和)

舉例：查詢所有書籍的總價格

Avg (平均值)

舉例：查詢所有書籍的平均價格

對返回的key起別名

Max 和Min

舉例：所有圖書價格的最大值和最小值

Count (統計結果行數)

舉例：查詢西瓜出版社總共出版過多少本書籍

三、分組查詢

相關推薦