我有一个在开发中运行良好的rails应用程序(SQLite),但是当我通过Heroku部署它时会抛出很多错误,Heroku使用我收集的PostgreSQL。
返回错误消息:
ActionView::Template::Error (PGError: ERROR:
column "practices.id" must appear in the GROUP BY clause or be used in an aggregate function:
SELECT "practices".*
FROM "practices" WHERE ("practices".activity_id = 1)
AND ("practices"."created_at" BETWEEN '2011-01-01' AND '2011-01-31')
GROUP BY DATE(created_at) ORDER BY created_at DESC):
当我打电话给以下内容时会抛出这个:
def month_days_not_practiced(date = Date.today)
p = practices.where(:created_at => date.at_beginning_of_month..date.at_end_of_month).group("DATE(created_at)").to_a.count
days_in_month(date.year, date.month) - p
end
我真的希望保持代码清洁,以便它适用于开发和生产数据库......任何人都可以解决这些问题吗?
我试过这个:
def month_days_not_practiced(date = Date.today)
p = practices.where(:created_at => date.at_beginning_of_month..date.at_end_of_month).group("practices.id, DATE(created_at)").to_a.count
days_in_month(date.year, date.month) - p
end
无济于事......
TIA。
答案 0 :(得分:5)
Practice.group(Practice.col_list)
def self.col_list
Practice.column_names.collect {|c| "practices.#{c}"}.join(",")
end
答案 1 :(得分:4)
您需要按所有分组选择列表中的列,所以
.group("practices.id, practices.foo, practices.bar, ..., DATE(created_at)")
您只想按id
分组的想法是合理的(假设id
是主键),但PostgreSQL只支持9.1版本。
答案 2 :(得分:0)
Practice.joins(:foobar).group_all
def self.group_all
group(group_all_columns)
end
def self.group_all_columns
@group_all_columns ||= column_names.collect {|c| "#{table_name}.#{c}"}.join(",")
end