Django ORM实现按天获取数据去重求和例子

我就废话不多说了，大家还是直接看代码吧！

def total_data(request):
  data = request_body(request, 'POST')
  if not data:
    return http_return(400, '参数错误')
  # 前端传入毫秒为单位的时间戳
  startTimestamp = data.get('startTime', '')
  endTimestamp = data.get('endTime', '')

  if startTimestamp and endTimestamp:
    startTimestamp = int(startTimestamp/1000)
    endTimestamp = int(endTimestamp/1000)
  else:
    return http_return(400, '参数有误')
  # 小于2019-05-30 00:00:00的时间不合法
  if endTimestamp < startTimestamp or endTimestamp <= 1559145600 or startTimestamp <= 1559145600:
    return http_return(400, '无效时间')
  if startTimestamp and endTimestamp:
    # 给定时间查询
    startTime = datetime.fromtimestamp(startTimestamp)
    endTime = datetime.fromtimestamp(endTimestamp)
    t1 = datetime(startTime.year, startTime.month, startTime.day)
    t2 = datetime(endTime.year, endTime.month, endTime.day, 23, 59, 59, 999999)
    # 用户总人数
    totalUsers = User.objects.exclude(status='destroy').count()
    # 音频总数
    totalAudioStory = AudioStory.objects.filter(isDelete=False).count()
    # 专辑总数
    totalAlbums = Album.objects.filter(isDelete=False).count()
    # 新增用户人数
    newUsers = User.objects.filter(createTime__range=(t1, t2)).exclude(status='destroy').count()
    # 活跃用户人数
    activityUsers = LoginLog.objects.filter(createTime__range=(t1, t2), isManager=False).values('userUuid_id').
      annotate(Count('userUuid_id')).count()
    # 新增音频数
    newAudioStory = AudioStory.objects.filter(createTime__range=(t1, t2)).count()

    # 男性
    male = User.objects.filter(gender=1).exclude(status='destroy').count()

    # 女性
    female = User.objects.filter(gender=2).exclude(status='destroy').count()

    # 未知
    unkonwGender = User.objects.filter(gender=0).exclude(status='destroy').count()


    # 模板音频
    aduioStoryCount = AudioStory.objects.filter(
      isDelete=False, audioStoryType=1, isUpload=1, createTime__range=(t1, t2)).count()

    # 自由录制
    freedomStoryCount = AudioStory.objects.filter(
      isDelete=False, audioStoryType=0, isUpload=1, createTime__range=(t1, t2)).count()

    # 儿歌
    tags1 = Tag.objects.filter(code="RECORDTYPE", name='儿歌').first()
    tags1Count = tags1.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).count()   # 儿歌作品数
    user1Count = tags1.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).
      values('userUuid_id').annotate(Count('userUuid_id')).count()                # 录音类型人数，去重

    # result = Tag.objects.filter(code="RECORDTYPE").annotate(Count('tagsAudioStory'))

    # 父母学堂
    tags2 = Tag.objects.filter(code="RECORDTYPE", name='父母学堂').first()
    tags2Count = tags2.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).count()
    user2Count = tags2.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).
      values('userUuid_id').annotate(Count('userUuid_id')).count()

    # 国学
    tags3 = Tag.objects.filter(code="RECORDTYPE", name='国学').first()
    tags3Count = tags3.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).count()
    user3Count = tags3.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).
      values('userUuid_id').annotate(Count('userUuid_id')).count()

    # 英文
    tags4 = Tag.objects.filter(code="RECORDTYPE", name='英文').first()
    tags4Count = tags4.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).count()
    user4Count = tags4.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)). 
      values('userUuid_id').annotate(Count('userUuid_id')).count()

    # 其他
    tags5 = Tag.objects.filter(code="RECORDTYPE", name='其他').first()
    tags5Count = tags5.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).count()
    user5Count = tags5.tagsAudioStory.filter(isDelete=False, createTime__range=(t1, t2)).
      values('userUuid_id').annotate(Count('userUuid_id')).count()

    recordTypePercentage = [
      {'name': '儿歌', 'tagsNum': tags1Count, 'userNum': user1Count},
      {'name': '儿歌', 'tagsNum': tags2Count, 'userNum': user2Count},
      {'name': '国学', 'tagsNum': tags3Count, 'userNum': user3Count},
      {'name': '英文', 'tagsNum': tags4Count, 'userNum': user4Count},
      {'name': '其他', 'tagsNum': tags5Count, 'userNum': user5Count}
    ]

    # 活跃用户排行
    data1_list = []
    # result = AudioStory.objects.filter(isDelete=False, createTime__range=(t1, t2)).values('userUuid_id').annotate(Count('userUuid_id'))[:1]
    res = User.objects.annotate(audioStory_count_by_user = Count("useAudioUuid")).order_by('-audioStory_count_by_user')[:5]
    for index,item in enumerate(res.values()):
      data = {
        'orderNum': index+1,
        'name': item['nickName'],
        'recordCount': item['audioStory_count_by_user']
      }
      data1_list.append(data)
    # 热门录制排行
    data2_list = []
    res = Story.objects.filter(status="normal", createTime__range=(t1, t2)).order_by('-recordNum')[:5]
    for index,item in enumerate(res.values()):
      data = {
        'orderNum': index + 1 or -1,
        'name': item['name'] or '',
        'recordNum': item['recordNum'] or 0
      }
      data2_list.append(data)

    # 热门播放排行
    data3_list = []
    audioStory = AudioStory.objects.filter(isDelete=False, createTime__range=(t1, t2)).order_by('-playTimes')[:5]
    for index,item in enumerate(audioStory):
      data = {
        'orderNum': index + 1,
        'name': item.storyUuid.name if item.audioStoryType else item.name,
        'playTimes': item.playTimes
      }
      data3_list.append(data)

    # 图表数据--新增用户
    graph1 = User.objects.filter(createTime__range=(t1, t2)).
      extra(select={"time": "DATE_FORMAT(createTime,'%%Y-%%m-%%e')"}).
      order_by('time').values('time')
      .annotate(userNum=Count('createTime')).values('time', 'userNum')
    if graph1:
      graph1 = list(graph1)
    else:
      graph1 = []

    # 活跃用户
    graph2 = LoginLog.objects.filter(createTime__range=(t1, t2), isManager=False). 
      extra(select={"time": "DATE_FORMAT(createTime,'%%Y-%%m-%%e')"}). 
      values('time').annotate(userNum=Count('createTime', distinct=True)).values('time', 'userNum')
    if graph2:
      graph2 = list(graph2)
    else:
      graph2 = []

    return http_return(200, 'OK',
              {
                'totalUsers': totalUsers,      # 总用户人数
                'totalAudioStory': totalAudioStory, # 音频总数
                'totalAlbums': totalAlbums,     # 总的专辑数
                'newUsers': newUsers,        # 新增用户人数
                'activityUsers': activityUsers,   # 活跃用户人数
                'newAudioStory': newAudioStory,   # 新增音频数
                'activityUsersRank': data1_list,   # 活跃用户排行
                'male': male,             # 男性
                'female': female,           # 女性
                'unkonwGender': unkonwGender,    # 未知性别
                'aduioStoryCount': aduioStoryCount, # 模板音频数量
                'freedomStoryCount': freedomStoryCount, # 自由录制音频数量
                'recordTypePercentage': recordTypePercentage,
                'hotRecordRank': data2_list,     # 热门录制排行
                'hotPlayAudioStoryRank': data3_list,   # 热门播放排行
                'newUserGraph': graph1,       # 新增用户折线图
                'activityUserGraph': graph2,     # 活跃用户折线图
              })

补充知识：Django 对符合条件的某个字段进行求和,聚合函数annotate()

开发环境：Ubuntu16.04+Django 1.11.9+Python2.7

对符合条件的某个字段求和　

之前在开发的时候,有同事问Django是否存在着这样的方法,可以直接将符合条件的某个字段直接求和.

当时不知道这样的方法是否存在,但是想了想自己解决这类似问题的方法,先用filter将符合条件的取出来,然后进行for循环,取出需要的字段,进行求和.感觉是挺low的,于是一起BAIdu,写代码测试最后找到了可以求值的方法,聚合函数annotate().

from django.db.models import Sum
from models import Book
all_price = Book.objects.values('price').annotate(num_books=Sum('price')).filter(author='Yu')
print all_price[0]['num_books']

输出结果:650

上面的参数换个顺序,不会出错但不符合预期结果.

all_price = Book.objects.annotate(num_books=Sum('price')).filter(author='Yu').values('price')
print all_youxibi[0]['num_books']

输出结果:'nums_book'