Convertir une plage de dates en une description d'intervalle

Dans un projet récent, une exigence consistait à indiquer quand une ressource serait entièrement consommée. En plus de la date du calendrier de l'épuisement, on m'a demandé de montrer le temps restant au format anglais, quelque chose comme "1 an, 3 mois."

La DATEDIFFfonction intégrée

Renvoie le nombre ... des limites de partie de date spécifiées croisées entre la date de début et la date de fin spécifiées.

Si utilisé tel quel, cela pourrait produire des résultats trompeurs ou déroutants. Par exemple, l'utilisation d'un intervalle de YEAR montrerait que 1999-12-31 (YYYY-MM-DD) et 2000-01-01 sont à un an d'intervalle alors que le bon sens dirait que ces dates ne sont séparées que d'un jour. À l'inverse, en utilisant un intervalle de JOUR 1999-12-31 et 2010-12-31 sont séparés de 4 018 jours alors que la plupart des gens verraient "11 ans" comme une meilleure description.

En partant du nombre de jours et en calculant les mois et les années à partir de là, il y aurait des erreurs d'année bissextile et de taille de mois.

Je me suis demandé comment cela pouvait être implémenté dans les différents dialectes SQL? Un exemple de sortie comprend:

create table TestData(
    FromDate date not null,
    ToDate date not null,
    ExpectedResult varchar(100) not null); -- exact formatting is unimportant

insert TestData (FromDate, ToDate, ExpectedResult)
values ('1999-12-31', '1999-12-31', '0 days'),
       ('1999-12-31', '2000-01-01', '1 day'),
       ('2000-01-01', '2000-02-01', '1 month'),
       ('2000-02-01', '2000-03-01', '1 month'),              -- month length not important
       ('2000-01-28', '2000-02-29', '1 month, 1 day'),       -- leap years to be accounted for
       ('2000-01-01', '2000-12-31', '11 months, 30 days'),
       ('2000-02-28', '2000-03-01', '2 days'),
       ('2001-02-28', '2001-03-01', '1 day'),                -- not a leap year
       ('2000-01-01', '2001-01-01', '1 year'),
       ('2000-01-01', '2011-01-01', '11 years'),
       ('9999-12-30', '9999-12-31', '1 day'),                -- catch overflow in date calculations
       ('1900-01-01', '9999-12-31', '8099 years 11 months 30 days');  -- min(date) to max(date)

Il se trouve que j'utilise SQL Server 2008R2, mais je suis intéressé d'apprendre comment d'autres dialectes géreraient cela.

La solution suivante concerne SQL Server. L'approche est similaire à celle de Serg dans la mesure où la requête utilise uniquement les fonctions DATEADD et DATEDIFF. Cependant, il ne tient pas compte des intervalles négatifs ( FromDate > ToDate ), et il dérive les années et les mois de la différence totale de mois:

WITH
  MonthDiff AS
  (
    SELECT
      t.FromDate,
      t.ToDate,
      t.ExpectedResult,
      Months = x.Months - CASE WHEN DAY(t.FromDate) > DAY(t.ToDate) THEN 1 ELSE 0 END
    FROM
      dbo.TestData AS t
      CROSS APPLY (SELECT DATEDIFF(MONTH, t.FromDate, t.ToDate)) AS x (Months)
  )
SELECT
  t.FromDate,
  t.ToDate,
  t.ExpectedResult,
  Result = ISNULL(NULLIF(ISNULL(x.Years  + CASE x.Years  WHEN '1' THEN ' year '  ELSE ' years '  END, '')
                       + ISNULL(x.Months + CASE x.Months WHEN '1' THEN ' month ' ELSE ' months ' END, '')
                       + ISNULL(x.Days   + CASE x.Days   WHEN '1' THEN ' day '   ELSE ' days '   END, ''), ''), '0 days')
FROM
  MonthDiff AS t
  CROSS APPLY
  (
    SELECT
      CAST(NULLIF(t.Months / 12, 0) AS varchar(10)),
      CAST(NULLIF(t.Months % 12, 0) AS varchar(10)),
      CAST(NULLIF(DATEDIFF(DAY, DATEADD(MONTH, t.Months, t.FromDate), t.ToDate), 0) AS varchar(10))
  ) AS x (Years, Months, Days)
;

Production:

FromDate    ToDate      ExpectedResult                 Result
----------  ----------  -----------------------------  -----------------------------
1999-12-31  1999-12-31  0 days                         0 days
1999-12-31  2000-01-01  1 day                          1 day 
2000-01-01  2000-02-01  1 month                        1 month 
2000-02-01  2000-03-01  1 month                        1 month 
2000-01-28  2000-02-29  1 month, 1 day                 1 month 1 day 
2000-01-01  2000-12-31  11 months, 30 days             11 months 30 days 
2000-02-28  2000-03-01  2 days                         2 days 
2001-02-28  2001-03-01  1 day                          1 day 
2000-01-01  2001-01-01  1 year                         1 year 
2000-01-01  2011-01-01  11 years                       11 years 
9999-12-30  9999-12-31  1 day                          1 day 
1900-01-01  9999-12-31  8099 years 11 months 30 days   8099 years 11 months 30 days 

Cette réponse montre une implémentation utilisant une fonction CLR SQL Server (2005+).

-- Enable CLR (if necessary)
EXECUTE sys.sp_configure 
    @configname = 'clr enabled',
    @configvalue = 1;

RECONFIGURE;

Assemblage et fonction

CREATE ASSEMBLY DBA
AUTHORIZATION dbo
FROM 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
WITH PERMISSION_SET = SAFE;
GO
CREATE FUNCTION dbo.IntervalDescription
(
    @From date, 
    @To date
)
RETURNS nvarchar(100)
AS EXTERNAL NAME 
    DBA.UserDefinedFunctions.IntervalDescription;

Usage

SELECT 
    TD.FromDate,
    TD.ToDate,
    TD.ExpectedResult, 
    IntervalDescription = dbo.IntervalDescription(TD.FromDate, TD.ToDate) 
FROM dbo.TestData AS TD;

Résultat

La source

_{Je ne suis pas programmeur C #!}

using Microsoft.SqlServer.Server;
using System;
using System.Text;

public partial class UserDefinedFunctions
{
    [SqlFunction
        (
        DataAccess = DataAccessKind.None,
        SystemDataAccess = SystemDataAccessKind.None,
        IsDeterministic = true,
        IsPrecise = true,
        Name = "IntervalDescription"
        )
    ]
    [return: SqlFacet(IsFixedLength = false, IsNullable = false, MaxSize = 100)]
    public static string IntervalDescription(DateTime From, DateTime To)
    {
        var workDate = From;
        int years = To.Year - From.Year;
        int months = 0;
        int days = 0;

        if (years != 0)
        {
            if (From.Month > To.Month || (From.Month == To.Month && From.Day > To.Day))
            {
                years--;
            }
            workDate = workDate.AddYears(years);
        }

        while (workDate < To && (workDate.Year != DateTime.MaxValue.Year || workDate.Month != DateTime.MaxValue.Month))
        {
            if (workDate.AddMonths(1) <= To)
            {
                months++;
                workDate = workDate.AddMonths(1);
            }
            else
            {
                break;
            }
        }

        while (workDate < To)
        {
            days++;
            workDate = workDate.AddDays(1);
        }

        StringBuilder sb = new StringBuilder(100);

        if (years > 0)
        {
            sb.Append(years);
            sb.Append(years == 1 ? " year" : " years");
            sb.Append((months > 0 || days > 0) ? ", " : string.Empty);
        }

        if (months > 0)
        {
            sb.Append(months);
            sb.Append(months == 1 ? " month" : " months");
            sb.Append(days > 0 ? ", " : string.Empty);
        }

        if (days > 0 || (years == 0 && months == 0))
        {
            sb.Append(days);
            sb.Append(days == 1 ? " day" : " days");
        }

        return
            sb.ToString();

    }
}

Ma version, implémentée dans SQL Server 2008R2 SP2.

CREATE FUNCTION dbo.ReadableInterval(
    @FromDate AS date,
    @ToDate AS date
)
RETURNS TABLE AS RETURN 
(
with YearStep as
(
    select
        max(n1.Number) as YearNumber
    from dbo.Numbers as n1
    where n1.Number <= DATEDIFF(YEAR, @FromDate, @ToDate)  -- see comment (A)
    and DATEADD(YEAR, n1.Number, @FromDate) <= @ToDate     -- see comment (B)
)
, MonthStep as
(
    select
        max(n2.Number) as MonthNumber
    from dbo.Numbers as n2
    cross apply YearStep as y1
    where n2.Number <= DATEDIFF(MONTH, DATEADD(YEAR, y1.YearNumber, @FromDate), @ToDate)
    and DATEADD(MONTH, n2.Number, DATEADD(YEAR, y1.YearNumber, @FromDate)) <= @ToDate
)
, DayStep as
(
    select
        DATEDIFF(day, DATEADD(MONTH, m1.MonthNumber, DATEADD(YEAR, y2.YearNumber, @FromDate)), @ToDate) as DayNumber
    from MonthStep as m1
    cross apply YearStep as y2
)
select
    y.YearNumber,
    m.MonthNumber,
    d.DayNumber
from YearStep as y
cross apply MonthStep as m
cross apply DayStep as d
)

Avec les données d'essai données, les résultats sont

select
    td.FromDate,
    td.ToDate,
    td.ExpectedResult,
    ri.YearNumber as Years,
    ri.MonthNumber as Months,
    ri.DayNumber as [Days]
from dbo.TestData as td
cross apply dbo.ReadableInterval(td.FromDate, td.ToDate) as ri;

FromDate   ToDate     ExpectedResult               Years Months Days
---------- ---------- ---------------------------- ----- ------ ----
1999-12-31 1999-12-31 0 days                           0      0    0
1999-12-31 2000-01-01 1 day                            0      0    1
2000-01-01 2000-02-01 1 month                          0      1    0
2000-02-01 2000-03-01 1 month                          0      1    0
2000-01-28 2000-02-29 1 month, 1 day                   0      1    1
2000-01-01 2000-12-31 11 months, 30 days               0     11   30
2000-02-28 2000-03-01 2 days                           0      0    2
2001-02-28 2001-03-01 1 day                            0      0    1
2000-01-01 2001-01-01 1 year                           1      0    0
2000-01-01 2011-01-01 11 years                        11      0    0
9999-12-30 9999-12-31 1 day                            0      0    1
1900-01-01 9999-12-31 8099 years 11 months 30 days  8099     11   30

Explication

Mon approche générale est d'avancer à partir de la date antérieure, d'abord en années, puis en mois, puis en jours. À chaque niveau de granularité, l'objectif est de se rapprocher le plus possible de la date de fin sans la dépasser, puis de continuer au niveau inférieur suivant.

J'utilise un tableau de nombres pour faciliter le calcul proche mais pas terminé. De ce tableau et DATEADDje peux trouver le plus grand nombre d'années / mois / jours qui précèdent ToDate- commentaire (B) dans le code.

Étant donné que je cherchais le numéro MAX et que ma table Numbers y était regroupée, l'optimiseur effectuait un balayage descendant, alimentant les valeurs en DATEADD. Cela provoquait des erreurs de dépassement de date, car Numbers contient plus de 100 000 lignes. DATEADD(YEAR, 100000, @FromDate)est supérieur à 9999-12-31 et une erreur est générée. Le prédicat (A) donne une limite supérieure à la valeur numérique à partir de laquelle l'analyse vers l'arrière démarre, évitant le dépassement de date. Par conséquent, le plan de requête traverse très peu de lignes, même pour de très grandes plages de dates.

Cette approche est utilisée pour trouver des années et des mois, sauf que le point de départ des mois est avancé par le nombre d'années que j'ai trouvé dans le premier CTE. DAYS est mon niveau de granularité le plus bas, donc un simple DATEDIFF suffit.

Cela pourrait être étendu à une granularité plus fine, renvoyant l'intervalle en heures, minutes et secondes si nécessaire.

PostgreSQL prend en charge la agefonction prête à l'emploi:

select
  FromDate,
  ToDate,
  ExpectedResult,
  age(ToDate, FromDate)
from TestData;

Cela donne le résultat souhaité, donnez ou prenez des valeurs de temps supplémentaires.

FromDate      ToDate        ExpectedResult                  age
----------    ----------    ----------------------------    --------------------------
1999-12-31    1999-12-31    0 days                          00:00:00
1999-12-31    2000-01-01    1 day                           1 day
2000-01-01    2000-02-01    1 month                         1 mon
2000-02-01    2000-03-01    1 month                         1 mon
2000-01-28    2000-02-29    1 month, 1 day                  1 mon 1 day
2000-01-01    2000-12-31    11 months, 30 days              11 mons 30 days
2000-02-28    2000-03-01    2 days                          2 days
2001-02-28    2001-03-01    1 day                           1 day
2000-01-01    2001-01-01    1 year                          1 year
2000-01-01    2011-01-01    11 years                        11 years
9999-12-30    9999-12-31    1 day                           1 day
1900-01-01    9999-12-31    8099 years 11 months 30 days    8099 years 11 mons 30 days

Version sans numbertable ni décompte requis. Donne le même résultat sur les données de test de Michael Green. Ils diffèrent sur les données où @FromDate > @ToDate. ReadableInterval2renvoie des valeurs négatives contrairement aux valeurs nulles.

CREATE FUNCTION dbo.ReadableInterval2(
    @FromDate AS date,
    @ToDate AS date
)
RETURNS TABLE AS RETURN 
(with checkData as (
    select 
       fromDate = case when @FromDate > @ToDate then @ToDate else @FromDate end,
       toDate = case when @FromDate <= @ToDate then @ToDate else @FromDate end,
       k = case when @FromDate > @ToDate then -1 else 1 end
), MonthStep as (
    select k, FromDate, ToDate,
        YearNumber = x.months / 12,
        MonthNumber = x.months % 12
    from checkdata
    cross apply(
        select months = DATEDIFF(MONTH, FromDate, ToDate)
            - case when DAY(FromDate) > DAY(ToDate) then 1 else 0 end
        ) x
)
select YearNumber = k*YearNumber, 
      MonthNumber = k*MonthNumber,
      DayNumber = k*DATEDIFF(day, DATEADD(MONTH, MonthNumber, DATEADD(YEAR, YearNumber, FromDate)), ToDate) 
    from MonthStep 
)