SQL计数每天保持查询不按预期工作

Question

Let's assume I manage medical patient stays information system.

I want to get the patient count per day with the following minimal structure :

• stay table has begin and end datetime columns
• PHP gives me $first_day and $last_day limits

The following snippet is NOT what I want, since it only counts entries per day, and not present stays per day:

SELECT 
  DATE_FORMAT(`stay`.`begin`, '%Y-%m-%d') AS `date`, 
  COUNT(`stay`.`stay_id`) AS `total`
FROM `stay`
WHERE `stay`.`begin` <= '$first_day'
  AND `stay`.`end`   >= '$last_day'
GROUP BY `date` 
ORDER BY `date`

Last but not least, I'm looking for a full SQL query.

It goes without saying that making one SQL query for each day would be totally trivial.

Use of temporary (dates ?) table is clearly an option.

Answer 1

A quick Google around leads me to this page: What is the most straightforward way to pad empty dates in sql results (on either mysql or perl end)?

What I would suggest is that you either follow the advice in that question, or construct your own loop in PHP.

Answer 2

As you mentioned using a temporary table of all dates in the range you want is one way to handle this. If you created a table of date called foo with all dates between $first_day and $last_day inclusive (see here ).

Then you can write your query like:

SELECT f.date, count(s.stay_id)
FROM foo f
JOIN stay s ON s.begin <= f.date AND s.end >= date
GROUP BY f.date
ORDER BY f.date