Mysql - 选择一天内不使用的小时

Question

I have two tables in my database that have a relation. I use MySQL. Basically, I created an app to manage 'Futsal's Field Order' So, here we go :

The first table is named Lapangan means "Field in Indonesian" :

mysql> SELECT id,nama_lapangan FROM lapangan;
+----+---------------+
| id | nama_lapangan |
+----+---------------+
|  1 | Lap 01        |
|  2 | Lap 02        |
|  3 | Lap 03        |
+----+---------------+
3 rows in set (0.00 sec)

And The Second Table is Booking : ,

mysql> SELECT id, nomor_booking, date_booking, date_end_booking, lapangan_id FROM `yfutsal`.`booking` LIMIT 1000;
+----+---------------+---------------------+---------------------+-------------+
| id | nomor_booking | date_booking        | date_end_booking    | lapangan_id |
+----+---------------+---------------------+---------------------+-------------+
|  1 |             1 | 2017-07-16 10:00:00 | 2017-07-16 12:00:00 |           1 |
|  2 |             2 | 2017-07-16 15:00:00 | 2017-07-16 16:00:00 |           3 |
+----+---------------+---------------------+---------------------+-------------+

Here is the problem.

For example, The Owner start at 08.00 AND end in 23.00.

It means,

1. lap 1 is not available on 10.00 - 12.00
2. lap 3 is not available on 15.00 - 16.00.

The goal is, I want to display the Lapangan (field) that available with hour field, based hour that not include from case above.

So, the cashier can choice it. Something like this :

+----+---------------+----------------------+-----------------------+
| id | nama_lapangan |   Available Start    |    Available End      |
+----+---------------+----------------------+-----------------------+
|  1 | Lap 01        |  2017-07-16 08:00:00 |   2017-07-16 09:59:00 |
|  1 | Lap 01        |  2017-07-16 12:01:00 |   2017-07-16 23:00:00 |
|  2 | Lap 02        |  2017-07-16 08:00:00 |   2017-07-16 23:00:00 |
|  3 | Lap 03        |  2017-07-16 08:00:00 |   2017-07-16 14:59:00 |
|  3 | Lap 03        |  2017-07-16 16:01:00 |   2017-07-16 23:00:00 |
+----+---------------+----------------------+-----------------------+

From bottom of my heart, please advise.

Answer 1

This is a rather tricky goal without window functions (which MySQL 5.7 doesn't have). It is feasible, even if it isn't easy to understand.

For the sake of ease, I've used an extra table I've called parameters :

CREATE TABLE parameters
(
    start_date_time DATETIME,
    end_date_time DATETIME
) ;
INSERT INTO parameters 
VALUES ('2017-07-16 08:00', '2017-07-16 23:00') ;

The idea is, for a given lapangan (field), if one day there's just one booking (start_booking_1,end_booking_1), we will get two available periods:

start_date_time .. start_booking_1      <- first period
end_booking_1   .. end_date_time        <- 2

If one day there are two bookings (start_booking_1, end_booking_1) and (start_booking_2, end_booking_2), sorted, we'll have:

start_date_time .. start_booking_1      <- first period
end_booking_1   .. start_booking_2      <- 2
end_booking_2   .. end_date_time        <- 3

So, we need to distinguish the first period from the rest:

First free segment can be computed with:

-- first free period
SELECT
    lapangan.id, parameters.start_date_time AS available_start, 
    coalesce( (SELECT date_start_booking
                 FROM booking b
                WHERE b.lapangan_id = lapangan.id
              ORDER BY b.date_start_booking ASC
                LIMIT 1
               ), parameters.end_date_time) AS available_end
FROM
    lapangan, parameters

NOTE: the way to search for start_booking_1 is the awful subquery. If it doesn't return a value, we'll go to the end_date_time .

The intermediate (and last) periods, are computed with:

-- 2..n free periods
SELECT
    lapangan.id, 
    b1.date_end_booking AS available_start, 
    coalesce ( (SELECT date_start_booking
                  FROM booking b2
                 WHERE b2.lapangan_id = b1.lapangan_id 
                     AND b2.date_start_booking >= b1.date_end_booking
              ORDER BY b2.date_start_booking ASC
                 LIMIT 1), 
               (SELECT parameters.end_date_time 
                  FROM parameters)
    ) AS avilable_end
FROM
    lapangan
    JOIN booking b1 ON b1.lapangan_id = lapangan.id

You'll have to put everything together, and take care of possible empty periods. Then you'll get

SELECT DISTINCT
    lapangan.id, lapangan.nama_lapangan, av.available_start, av.available_end AS available_end
FROM
    (    
    -- first free period
    SELECT
        lapangan.id, parameters.start_date_time AS available_start, 
        coalesce( (SELECT date_start_booking
                     FROM booking b
                    WHERE b.lapangan_id = lapangan.id
                  ORDER BY b.date_start_booking ASC
                    LIMIT 1
                   ), parameters.end_date_time) AS available_end
    FROM
        lapangan, parameters

    UNION  
    -- 2..n free periods
    SELECT
        lapangan.id, 
        b1.date_end_booking AS available_start, 
        coalesce ( (SELECT date_start_booking
                      FROM booking b2
                     WHERE b2.lapangan_id = b1.lapangan_id 
                         AND b2.date_start_booking >= b1.date_end_booking
                  ORDER BY b2.date_start_booking ASC
                     LIMIT 1), 
                   (SELECT parameters.end_date_time 
                      FROM parameters)
        ) AS avilable_end
    FROM
        lapangan
        JOIN booking b1 ON b1.lapangan_id = lapangan.id
    ) AS av
    JOIN lapangan ON lapangan.id = av.id
WHERE
    -- Ignore empty segments
    av.available_start < av.available_end
ORDER BY
    lapangan.id, available_start ;

Which will give you the intended result.

id | nama_lapangan | available_start     | available_end      
-: | :------------ | :------------------ | :------------------
 1 | Lap 01        | 2017-07-16 08:00:00 | 2017-07-16 10:00:00
 1 | Lap 01        | 2017-07-16 12:00:00 | 2017-07-16 23:00:00
 2 | Lap 02        | 2017-07-16 08:00:00 | 2017-07-16 23:00:00
 3 | Lap 03        | 2017-07-16 08:00:00 | 2017-07-16 15:00:00
 3 | Lap 03        | 2017-07-16 16:00:00 | 2017-07-16 23:00:00

Note that I have not subtracted 1 minute or 1 second from the start or end. Assume that your periods are considered:

start <= available_time < end 

or, in range terminology

[start, end)

If you really need to subtract 1 second, do it in the first line of the select.

You can check everything (with some step-by-step approach to reaching the solution) at dbfiddle here

---

Side Note: this is much easier if you use a DB that knows how to use the LEAD or LAG window functions.

Answer 2

The essence of this problem is to generate rows of data where there isn't anything stored. In other words you have the hours that are booked stored, but not the hours which remain un-booked. So you need a way to generate these. While there are several alternatives to this a simple option is to use a Cartesian product to do this.

Here I have opted to store the hours of operation in a table as a set of integers 8 to 23 (maybe 8 to 22 depending on your intention). This can then be used in a CROSS JOIN (to produce a Cartesian product) on any given date of all hours for all fields (nama_lapangan). Once this is generated we can LEFT JOIN all the available hours to these which are already booked, and then where there is no current related booking (IS NULL) we determine the available hours for that day.

DATA

CREATE TABLE OpenHours
    (`StartAt` int)
;

INSERT INTO OpenHours     (`StartAt`)
VALUES
    (8),    (9),    (10),    (11),    (12),    (13),    (14),    (15),
    (16),    (17),    (18),    (19),    (20),    (21),    (22),    (23)
;

CREATE TABLE lapangan
    (`id` int, `nama_lapangan` varchar(6))
;

INSERT INTO lapangan     (`id`, `nama_lapangan`)
VALUES
    (1, 'Lap 01'), (2, 'Lap 02'), (3, 'Lap 03')
;


CREATE TABLE yfutsal
    (`id` int, `nomor_booking` int, `date_booking` datetime, `date_end_booking` datetime, `lapangan_id` int)
;

INSERT INTO yfutsal
    (`id`, `nomor_booking`, `date_booking`, `date_end_booking`, `lapangan_id`)
VALUES
    (1, 1, '2017-07-16 10:00:00', '2017-07-16 12:00:00', 1),
    (2, 2, '2017-07-16 15:00:00', '2017-07-16 16:00:00', 3)
;

QUERY

set @dt := str_to_date('2017-07-16','%Y-%m-%d');

select
    l.id
  , l.nama_lapangan
  , date_add(@dt,INTERVAL h.StartAt HOUR) AvailStartHr
  , date_add(@dt,INTERVAL h.StartAt+1 HOUR) AvailEndHr
from lapangan l
cross join OpenHours h
left join yfutsal y on l.id = y.lapangan_id
                   and date_add(@dt,INTERVAL h.StartAt HOUR) between date_booking and date_end_booking
where y.date_booking IS NULL
order by l.nama_lapangan, AvailStartHr
;

RESULT

| id | nama_lapangan |        AvailStartHr |          AvailEndHr |
|----|---------------|---------------------|---------------------|
|  1 |        Lap 01 | 2017-07-16 08:00:00 | 2017-07-16 09:00:00 |
|  1 |        Lap 01 | 2017-07-16 09:00:00 | 2017-07-16 10:00:00 |
|  1 |        Lap 01 | 2017-07-16 13:00:00 | 2017-07-16 14:00:00 |
|  1 |        Lap 01 | 2017-07-16 14:00:00 | 2017-07-16 15:00:00 |
|  1 |        Lap 01 | 2017-07-16 15:00:00 | 2017-07-16 16:00:00 |
|  1 |        Lap 01 | 2017-07-16 16:00:00 | 2017-07-16 17:00:00 |
|  1 |        Lap 01 | 2017-07-16 17:00:00 | 2017-07-16 18:00:00 |
|  1 |        Lap 01 | 2017-07-16 18:00:00 | 2017-07-16 19:00:00 |
|  1 |        Lap 01 | 2017-07-16 19:00:00 | 2017-07-16 20:00:00 |
|  1 |        Lap 01 | 2017-07-16 20:00:00 | 2017-07-16 21:00:00 |
|  1 |        Lap 01 | 2017-07-16 21:00:00 | 2017-07-16 22:00:00 |
|  1 |        Lap 01 | 2017-07-16 22:00:00 | 2017-07-16 23:00:00 |
|  1 |        Lap 01 | 2017-07-16 23:00:00 | 2017-07-17 00:00:00 |
|  2 |        Lap 02 | 2017-07-16 08:00:00 | 2017-07-16 09:00:00 |
|  2 |        Lap 02 | 2017-07-16 09:00:00 | 2017-07-16 10:00:00 |
|  2 |        Lap 02 | 2017-07-16 10:00:00 | 2017-07-16 11:00:00 |
|  2 |        Lap 02 | 2017-07-16 11:00:00 | 2017-07-16 12:00:00 |
|  2 |        Lap 02 | 2017-07-16 12:00:00 | 2017-07-16 13:00:00 |
|  2 |        Lap 02 | 2017-07-16 13:00:00 | 2017-07-16 14:00:00 |
|  2 |        Lap 02 | 2017-07-16 14:00:00 | 2017-07-16 15:00:00 |
|  2 |        Lap 02 | 2017-07-16 15:00:00 | 2017-07-16 16:00:00 |
|  2 |        Lap 02 | 2017-07-16 16:00:00 | 2017-07-16 17:00:00 |
|  2 |        Lap 02 | 2017-07-16 17:00:00 | 2017-07-16 18:00:00 |
|  2 |        Lap 02 | 2017-07-16 18:00:00 | 2017-07-16 19:00:00 |
|  2 |        Lap 02 | 2017-07-16 19:00:00 | 2017-07-16 20:00:00 |
|  2 |        Lap 02 | 2017-07-16 20:00:00 | 2017-07-16 21:00:00 |
|  2 |        Lap 02 | 2017-07-16 21:00:00 | 2017-07-16 22:00:00 |
|  2 |        Lap 02 | 2017-07-16 22:00:00 | 2017-07-16 23:00:00 |
|  2 |        Lap 02 | 2017-07-16 23:00:00 | 2017-07-17 00:00:00 |
|  3 |        Lap 03 | 2017-07-16 08:00:00 | 2017-07-16 09:00:00 |
|  3 |        Lap 03 | 2017-07-16 09:00:00 | 2017-07-16 10:00:00 |
|  3 |        Lap 03 | 2017-07-16 10:00:00 | 2017-07-16 11:00:00 |
|  3 |        Lap 03 | 2017-07-16 11:00:00 | 2017-07-16 12:00:00 |
|  3 |        Lap 03 | 2017-07-16 12:00:00 | 2017-07-16 13:00:00 |
|  3 |        Lap 03 | 2017-07-16 13:00:00 | 2017-07-16 14:00:00 |
|  3 |        Lap 03 | 2017-07-16 14:00:00 | 2017-07-16 15:00:00 |
|  3 |        Lap 03 | 2017-07-16 17:00:00 | 2017-07-16 18:00:00 |
|  3 |        Lap 03 | 2017-07-16 18:00:00 | 2017-07-16 19:00:00 |
|  3 |        Lap 03 | 2017-07-16 19:00:00 | 2017-07-16 20:00:00 |
|  3 |        Lap 03 | 2017-07-16 20:00:00 | 2017-07-16 21:00:00 |
|  3 |        Lap 03 | 2017-07-16 21:00:00 | 2017-07-16 22:00:00 |
|  3 |        Lap 03 | 2017-07-16 22:00:00 | 2017-07-16 23:00:00 |
|  3 |        Lap 03 | 2017-07-16 23:00:00 | 2017-07-17 00:00:00 |

see: http://sqlfiddle.com/#!9/775f36/4

Answer 3

In the former answer you learn how to generate the required available hours row by row. Once you have that data available, then it can be summarized into ranges using the following technique for MySQL:

set @dt := str_to_date('2017-07-16','%Y-%m-%d')

SELECT id, nama_lapangan, AvailStart, MAX(AvailEndHr) AvailEndHr
FROM (
  SELECT
         mytable.*
       , @fin := IF(@lid<=>id AND TIMESTAMPDIFF(HOUR, @d, AvailStartHr)=1, @fin, AvailStartHr) AS AvailStart
       , @lid := id
       , @d := AvailStartHr
  FROM (
      select
          l.id
        , l.nama_lapangan
        , date_add(@dt,INTERVAL h.StartAt HOUR) AvailStartHr
        , date_add(@dt,INTERVAL h.StartAt+1 HOUR) AvailEndHr
      from lapangan l
      cross join OpenHours h
      left join yfutsal y on l.id = y.lapangan_id
                         and date_add(@dt,INTERVAL h.StartAt HOUR) between date_booking and date_end_booking
      where y.date_booking IS NULL
     ) mytable
  CROSS JOIN (SELECT @lid:=NULL, @d:=NULL, @fin:= NOW()) AS init
  ORDER BY id, AvailStartHr
  ) d
GROUP BY  id, nama_lapangan, AvailStart

Results see sqlfiddle :

| id | nama_lapangan |          AvailStart |          AvailEndHr |
|----|---------------|---------------------|---------------------|
|  1 |        Lap 01 | 2017-07-16 08:00:00 | 2017-07-16 10:00:00 |
|  1 |        Lap 01 | 2017-07-16 13:00:00 | 2017-07-17 00:00:00 |
|  2 |        Lap 02 | 2017-07-16 08:00:00 | 2017-07-17 00:00:00 |
|  3 |        Lap 03 | 2017-07-16 08:00:00 | 2017-07-16 15:00:00 |
|  3 |        Lap 03 | 2017-07-16 17:00:00 | 2017-07-17 00:00:00 |

Also see this previous answer on a similar theme.