quick question on Oracle To_timestamp.
I have a two fields in a table that are both varchars
1 field contains the YYYYMMDD formatted value
the 2nd field contains the hh24mmss formatted value
Im trying to convert these two fields into a timestamp field,
expected output is DD/MM/YYYY HH24:mm:ss
like '7/23/2015 12:53:04'
however, it gaves me weird result...
like '7/15/0023 12:53:04'
seems year becomes day and day becomes year..
---------- heres my sql ----------------
select
to_date(to_char(to_date('20150723','yyyymmdd'),'yyyymmdd'),'yyyymmdd') dt,
to_char(to_date(SUBSTR('005304000',1,6), 'hh24miss'), 'hh24:mi:ss') tm,
TO_TIMESTAMP(
to_date(to_char(to_date('20150723','yyyymmdd'),'yyyymmdd'),'yyyymmdd') ||
to_char(to_date(SUBSTR('005304000',1,6), 'hh24miss'), 'hh24:mi:ss'), 'yyyy/mm/dd HH24:MI:SS' ) dttm
from dual
You have one layer to much in your conversion chain. And it will be much easier if you concatenate the two columns before converting them:
to_date(date_column||time_column, 'YYYYMMDDHH24MISS')
will return a date and a time (an Oracle DATE includes a time part).
Now that you have a proper date value, you can then format that as you want:
to_char(to_date(date_column||time_column, 'YYYYMMDDHH24MISS'), 'yyyy/mm/dd HH24:MI:SS')
As your "time" column doesn't contain milliseconds there is no real benefit in using to_timestamp.
But you should really fix your table, and store that information in a single DATE (or TIMESTAMP) column. Storing dates in varchar columns is always a very bad choice.
Related
I have a table tab1 in which a column col1 has data type VARCHAR2(50 BYTE) and this column has values like '9/27/21 18:05'
I want to add 1 day to this and I am expecting a result like '9/28/21 18:05'
If I do
TO_TIMESTAMP(col1,'MM/DD/YYYY HH24:MI') + INTERVAL '1' DAY
then I get '28-SEP-21 06.24.00.000000000 PM', and if I do
TO_DATE(col1,'MM/DD/YYYY HH24:MI') + INTERVAL '1' DAY'
then I get '28-SEP-21'.
Please note in both the above cases format is changing.
How can I get the result I want?
DATE and TIMESTAMP values are both binary data types that do NOT have a given format; therefore, when you convert a string to a DATE or a TIMESTAMP then the format you use is NOT stored.
If you want to convert it to a DATE or TIMESTAMP and then back to a string in the same format then, after adding the interval, you want to use TO_CHAR to convert back to a string with the required format.
For example:
SELECT TO_CHAR(
TO_DATE(col1, 'MM/DD/RR HH24:MI') + INTERVAL '1' DAY,
'MM/DD/RR HH24:MI'
)
FROM tab1
Note: If you use YYYY in the format model then 21 will be parsed as 21 AD and not as 2021 AD. Instead, you need to use YY or RR (depending on how you want values from the end of the last century to be handled).
Please note in both the above cases format is changing.
The format of your column is not changing, you have converted the strings to DATE and TIMESTAMP which are binary data type and do not have any format.
The user interface you are using (i.e. SQL/Plus or SQL Developer) tries to be helpful and rather than presenting you, the user, with binary data will use its internal rules to format the binary data as something you can read. SQL/Plus and SQL Developer will use the NLS_DATE_FORMAT session parameter for DATE values and the NLS_TIMESTAMP_FORMAT session parameter for TIMESTAMP values. These parameters can be set to different values for each user in each of their sessions so you should not rely on them to be consistent.
If you want a consistent format then wrap the date/timestamp in TO_CHAR to apply that consistent format.
You are converting your string into a date or timestamp, and adjusting it by a day. Your client then decides how to format that for display, usually using you session setting like NLS_DATE_FORMAT.
If you want to display (or store*) the value in a particular format then you should specify that, with to_char(), e.g.:
TO_CHAR(TO_DATE(col1,'MM/DD/YYYY HH24:MI') + INTERVAL '1' DAY,'MM/DD/YYYY HH24:MI')
09/28/0021 18:05
or if you want to suppress some leading zeros to match your original string you can toggle those with the FM modifier:
TO_CHAR(TO_DATE(col1,'MM/DD/YYYY HH24:MI') + INTERVAL '1' DAY,'FMMM/DD/YYYY HH24:FMMI')
9/28/21 18:05
As you can see in the output of first of those, and as #Aitor mentioned, the year comes out as 0021 rather than 21. That's because you used a four-digit YYYY mask for a 2-digit year value. In the second one the FM suppresses that, so it's less obvious. As you don't seem to care about the century it usually doesn't matter whether you use YY or RR - the exception maybe being if you happen to hit a leap year/day; but it's still better to have the mask match the string, so with RR:
TO_CHAR(TO_DATE(col1,'MM/DD/RR HH24:MI') + INTERVAL '1' DAY,'FMMM/DD/RR HH24:FMMI')
9/28/21 18:05
db<>fiddle
* But you should not be storing dates as strings. They should be stored as dates, and formatted as strings for display only. You shouldn't really be using 2-digit years any more either.
You said your column has this: 9/27/21, but you put a mask like YYYY. Be careful with that, because with YYYY, the year will be 21 BC...
Maybe you want RRRR in your date mask. RRRR means 2-digit years in the range 00 to 49 are assumed to be in the current century:
select to_char(to_date('9/27/2021 18:05','MM/DD/RRRR HH24:MI')+ INTERVAL '1' DAY,'MM/DD/YYYY HH24:MI') result
from dual;
Result: 09/28/2021 18:05
I don't know what is your output format, but anyway, if you want your date formatted like VARCHAR, try this. With your column, is something like that
select to_char(to_date(col1,'MM/DD/RRRR HH24:MI') + 1,'MM/DD/YYYY HH24:MI') result
from your_table;
Also you can use, instead of INTERVATL '1' DAY, a simple +1
I'm having trouble with filtering a date and time for anything two hours before and sooner. I tried this:
SELECT *
FROM
table
where
date >= sysdate - 1
AND
TO_DATE( Time, 'HH24:MI:SS' ) >= TO_DATE( sysdate, 'HH24:MI:SS' ) - 2
But I'm getting an inconsistent type error which is what I thought I was handling with the TO_DATE() function but I guess not.
sysdate is already a date (and time), so TO_DATE( sysdate, 'HH24:MI:SS' ) doesn't make any sense.
You didn't provide your data types for your date and time columns in table, so I'm going to assume they're both varchar2(10) with formats MM/DD/YYYY and HH24:MI:SS respectively.
I'm also going to go ahead and change your example table and column names, since they're invalid names to use in a real query.
-- example data
with my_table as (select '06/13/2019' as date_column, '09:40:34' as time_column from dual)
-- your query
SELECT *
FROM
my_table
where
to_date(date_column || ' ' || time_column, 'MM/DD/YYYY HH24:MI:SS') >= sysdate - 2/24
What I'm doing here is to combine your date and time strings into one date-time string, then converting it to an Oracle date type (actually date+time). Then we compare it to sysdate - 2/24, which says to take the current time and subtract 2/24ths of a day, which is 2 hours.
For this example, you might need to change the example data date_column and time_column values to something from the past 2 hours, depending on when you run this and what time zone you're in.
I have a table name as business_details and column name business_date whose data type is varchar2.
Now i have to find out the data between two different dates and date format like : 12-JUN-18 21:15:13
Means, 12 Jun, 2018.
Kindly help me to write a query which can fetch the data between these two dates :12-JUN-18 21:15:13 and 25-JUN-18 18:15:32
I assume that in table business_details you have an column date or something like that.
Than use something like this:
select business_date from business_details
where date between TO_DATE ('12-JUN-18 21:15:13','dd-MM-yy hh:mi:ss')
AND TO_DATE ('25-JUN-18 18:15:32','dd-MM-yy hh:mi:ss');
Assuming your business_date is actually a string in the format you've shown (and it isn't really a date your client is just showing in that format), you need to convert that to a date type, as well as converting the string literals.
select *
from business_details
where to_date(business_date, 'DD-MON-RR HH24:MI:SS')
between to_date('12-JUN-18 21:15:13', 'DD-MON-RR HH24:MI:SS')
and to_date('25-JUN-18 18:15:32', 'DD-MON-RR HH24:MI:SS');
The format model you tried to use in a comment did this:
to_date('12-JUN-18 21:15:13', 'DD-MM-YYYY HH24:MI:SS')
is using MM rather than MON, which works anyway by default - although using month numbers is safer anyway as they aren't dependent on your session language. But more importantly it uses YYYY. If you pass a 2-digit value like 18 and try to convert with YYYY you get the wrong year:
select to_date('12-JUN-18 21:15:13', 'DD-MON-YYYY HH24:MI:SS') form dual;
TO_DATE('12-JUN-182
-------------------
0018-06-12 21:15:13
In your version your business_date was being converted implicitly so would use NLS settings, which are presumably using RR already. But that means you were comparing a date in 2018 with a range in 0018, which is why nothing matched.
You could also use timestamp literals for the fixed values (unless those strings are actually being passed in from somewhere else):
select *
from business_details
where to_date(business_date, 'DD-MON-RR HH24:MI:SS')
between cast(timestamp '2018-06-12 21:15:13' as date)
and cast(timestamp '2018-06-25 18:15:32' as date);
In Oracle SQL I have merged 4 columns (ex of column names: a_d, a_t, d_d,d_t) into 2 columns (ex of names: a_d_t and d_d_t) that are each of the format: YYYY-MMM-DD HH24:MI. I am trying to find how much time passed (days and hours), for each observation, between a_d_t (the starting time) and d_d_t (the ending time).
I have tried d_d_t - a_d_t and to_date(d_d_t)-to_date(a_d_t), but I got back the following for each: invalid identifier.
For reference the code that I used (which worked), to merge the columns is:
to_char(to_date
( to_char (a_d,'YYYYMMDD')
|| a_t,
'YYYYMMDDHH24MI'
), 'YYYY-MM-DD HH24:MI')
Your inputs are strings, so you must convert them to dates. But it is not sufficient to say to_date, you must also give proper format models.
select to_date(d_d_t, 'yyyy-mm-dd hh24:mi') - to_date(a_d_t, 'yyyy-mm-dd hh24:mi')
from < ... >
will give you the difference in days. Multiply by 24, you will get it in hours.
Considering your inputs d_d_t, a_dare string in format yyyy-mm-dd hh:mi and nls_date_format is different from it, you can try to use following query and see, if it works.
To find the duration between two date, you have to convert your date string to the date format.
select col_a, col_b, to_date(col_a,'yyyy-mm-dd hh24:mi:ss')- to_date(col_b,'yyyy-mm-dd hh24:mi:ss')duration
from x_datedur
replace col_a with ending time, col_b with starting time and x_datedur with your tablename.
I have a table as
create table Dummy (date_created date)
in oracle.I want to store date in 'dd-mon-yyyy' (12-dec-2010) format.
How should i do this.
Please help.
In Oracle a column created with the DATE datatype just stores the date. It doesn't have a particular format, it just stores the day, month, year, hour, minute, and second. You need to convert from whatever format you have using the TO_DATE function. If you have a text string with the date in 'dd-mon-yyyy' format and you want to put this date into your table you'd use something like
INSERT INTO DUMMY (DATE_CREATED)
VALUES (TO_DATE('01-FEB-2011', 'DD-MON-YYYY');
Going the other way (from DATE column value to character string) you'd use the TO_CHAR function. If you were retrieving a value from your table and wanted to convert it to 'DD-MON-YYYY' format you'd use something like
SELECT TO_CHAR(DATE_CREATED, 'DD-MON-YYYY')
FROM DUMMY;
Share and enjoy.
Use to_date() function. In your case, the syntax would be
insert into Dummy values (to_date('08-09-2010', 'dd-mm-yyyy'));
Here is a link to the detailed help.
The DATE datatype will store date and time information (century, year, month, day, hours, minutes, and seconds) in an internal format in the database. When you get it out of the database, you can choose to display it in whatever format you like.
This information is either created using implicit conversion from a string or explicitly using either the TO_DATE function or the ANSI date literal. If you look in the v$nls_parameters view, this will tell you what the NLS_DATE_FORMAT is which is generally used for the implicit conversion. This may often be defined as DD-MON-RR, which might be why the date will come out as 23-DEC-10 when the query select sysdate from dual is run. (Not entirely sure I'm right about the nls stuff. Correct me if I'm wrong.)
However, all the date information is available if you know how to get it. The query select to_char(sysdate, 'dd-mon-yyyy hh24:mi:ss') from dual will return all the date fields.
Likewise, the insert statement shown below will create a row with a date value in it.
insert into dummy (date_created)
values (to_date('12-dec-2010 12:34:56', 'dd-mon-yyyy hh24:mi:ss'))`
This data can then be retrieved.
select date_created from dummy
This will implicitly convert the date to a character string using the NLS_DATE_FORMAT, providing the output below.
DATE_CREA
---------
23-DEC-10
The full date information is available by explicitly converting the date to a character string.
select to_char(date_created, 'DD-MON-YYYY') as date_created from dummy;
select to_char(date_created, 'DD-MON-YYYY HH24:MI:SS') as date_created
from dummy;
This will provide output in the format you require:
DATE_CREATE
-----------
23-DEC-2010
If you always use the TO_DATE and TO_CHAR functions to convert to/from a date datatype, then you will have fewer problems. Implicit conversion is useful but can cause some confusion or problems.
You can keep and eye here
http://www.techonthenet.com/oracle/functions/to_date.php
use to_date function to save a data with the format you need. I suggest to use SYSDATE updating table and when you need to read data from table use something like that:
dbms_output.put_line(TO_CHAR(SYSDATE, 'DD-MON-YYYY HH24:MI:SS'));
to solve your problem use:
to_date('08/JAN/2010', 'DD/MON/YYYY')
Just use TRUNC(YourDate) if date have time part, it will be truncate time part. Oracle have not just 'DATE' type, 'DATE' always have time part.
However if you do not specify time - it will 00:00:00.
SELECT TRUNC(SYSDATE) from dual
Result:
23-12-2010
Oracle does not support DATE without time part.
You can make it always be an integer date by adding a CHECK constraint:
CREATE TABLE dummy (date_created date CHECK (date_created = TRUNC(date_created)))
, insert it in any format you want:
INSERT
INTO dummy (date_created)
VALUES (TO_DATE('23-DEC-2010', 'dd-mon-yyyy'))
and select it in any format you want:
SELECT TO_CHAR(date_created, 'dd-mon-yyyy')
FROM dummy