I have a wordpress sql database that records CREATE_DATE in an integer format
some examples are
1615776024
1615852620
1616024968
1616027330
1616981834
some of the entries were created this month, so the format in which the date is recorded doesn't reveal anything about itself.
I have tried these but all of these statements error
SELECT
CONVERT(DATE,CONVERT(CHAR(8),CREATE_DATE),112)
,CONVERT(DATE,CONVERT(VARCHAR(10), CREATE_DATE))
,CONVERT(DATETIME, CONVERT(VARCHAR(10), CREATE_DATE))
FROM WORDPRESS_TABLE
But they all error with the same error message :
Msg 241, Level 16, State 1, Line 1 Conversion failed when converting
date and/or time from character string.
can any one tell me how I can convert this into the following format please? dd/MM/yyyy in SQL?
Having thought about it a bit more, (like 10 seconds); I wonder if what you see is not epoch. I.e. number of seconds since Jan 1, 1970. In that case you can quite easily convert it back to a proper date by: SELECT DATEADD(S, [your-value], '1970-01-01').
Testing that on your first value in the original question: SELECT DATEADD(S, 1615776024, '1970-01-01'), gives me: 2021-03-15 02:40:24.000.
This should "square" up with what you said above regarding newly created entries this month.
Related
I am trying to convert a date value ingested from SFMC as a string. Attempts and results below:
PARSE_DATE('%m/%d/%Y', '6/22/2017’) returns “invalid date”
2.safe.PARSE_DATE('%m-%d-%Y',Timesent) returns null results
Any insight would be greatly appreciated.
I wondered if it was because the day of month had no leading zero, because according to the docs, %d means the day of the month as a decimal number (01-31).
But your first example worked OK for me in a BigQuery SQL workspace anyway as per below. Your second example has hyphens rather than slashes in the format mask, and BigQuery returns a mismatch error for me.
Are you certain about the date value which is failing for you?
When looking at the History.db from Safari, there's a table named history_visits which has a column named visit_time, which is a REAL value. It has values such as 470799793.096987. What format is that in? I'd like to see it in a format such as 12/08/2015 05:12:05.
It's the number in seconds since 00:00:00 UTC on 1 January 2001. It must be coming from an NSDate.
NSDate objects encapsulate a single point in time, independent of any particular calendrical system or time zone. Date objects are immutable, representing an invariant time interval relative to an absolute reference date (00:00:00 UTC on 1 January 2001).
— NSDate Class Reference
To get a decent human value out of it, you must add 978307200 (the epoch for 2001-01-01 00:00:00).
This query should give you what you want:
.headers on
select datetime(v.visit_time + 978307200, 'unixepoch', 'localtime') as date, v.visit_time + 978307200 as epoch, v.visit_time, i.domain_expansion, i.url
from history_items i left join history_visits v on i.id = v.history_item
order by i.id desc
limit 100;
Example output:
date|epoch|visit_time|domain_expansion|url
2015-12-31 11:51:27|1451562687.28465|473255487.284646|duckduckgo|https://duckduckgo.com/?q=current+timestamp+2015-12-31+11:51&t=osx
PS: Just for future reference, the Safari db file is located at ~/Library/Safari/History.db
To convert the visit_time value in the history.db in an excel spread sheet, open the history.db file in a tool such as DB browser for SQLLite (Windows) and export the history_visits values to a CSV file.
Open the CSV file and create a column where you will populate your values in human readable time adjusted to your time zone, and use the following formula convert your NSDate:
=((((C2+978307200)/60)/60)/24)+DATE(1970,1,1)+(-5/24)
In the above formula, the time value is in cell C2, and my time zone GMT-5. To adjust to your own time zone adjust the statement in the last set of parenthesis. Presently I have (-5/24) to represent GMT-5.
When I first approached this conversion, I mistakenly assumed the time in the history.db to be epoch time, which starts at 1/1/1970, and did not understand why there was such a skew in time. Adding the required conversion factor +978307200 solved the problem.
I found the domain_expansion field to be null in some cases, here's a modified query:
SELECT SUBSTR(
SUBSTR(url, INSTR(url, '/')+2),
1,
INSTR(SUBSTR(url, INSTR(url, '/')+2),'/') - 1
) domain,
url,
datetime(hv.visit_time + 978307200, 'unixepoch', 'localtime') visit_time
FROM history_items hi
JOIN history_visits hv on hi.id = hv.history_item;
I have a scenario where i need to convert a Datefield(joindate) to currentcompany timezone date. And then i need to compare this with anotherdate(startdate). If the difference is more than 365 days i need to give an warning. Can someone help me in this.
Thanks in advance.
You can apply a timezone to an utcdatetime via DateTimeUtil::applyTimeZoneOffset
The company timezone can be retrieved by calling DateTimeUtil::getCompanyTimeZone
Afterwards calculate the difference by calling DateTimeUtil::getDifference, which returns the difference in seconds so you have to compare that with the seconds per year.
To avoid inserting a 'magic' number, use the constants in the macro library TimeConstants.
If Datefield(joindate) is of type date and not utcDateTime then DateTimeUtil::newDateTime() should be used to convert it to utcDateTime:
utcDateTime joinDateTime = DateTimeUtil::newDateTime(joindate, 0,
DateTimeUtil::getCompanyTimeZone());
DateTimeUtil::getDifference() can be used to get the number of seconds between the utcDateTime values.
If both Datefield(joindate) and anotherdate(startdate) are of type date and not utcDateType then no conversion is required at all, and you can check whether the difference is more than 365 as follows:
if (joindate - startdate > 365) {}
If the above assumptions are wrong, see DAXaholic's answer.
Select distinct Format(DateAdd(""s""," & columnname & ",""1/1/1980 12:00:00 AM""), 'dd-MMM-yyyy') as A
I have assumed that the seconds to add and the original date are hard coded values below whilst awaiting clarifications requested in the comments.
To add a number of seconds to a date you can use:
select datetime('1980-01-01 00:00:00', "345000 seconds");
This gives the result: 1980-01-04 23:50:00
The example above is just under 4 days in seconds, if you want to truncate the result to just the date as implied by the query in your questions then you can wrap this inside a date function. However, this would give the result in the format "YYYY-MM-DD" rather than "DD-MMM-YYYY" as your access query does.
Unfortunately I cannot find any native SQLite function to convert a numeric month value to mmm format. You can do this manually with replace (similar to the answer to this question), but this is a bit messy.
If you are happy to live with the numeric months then you can simply use:
select strftime('%d-%m-%Y', '1980-01-01 00:00:00', "345000 seconds");
This gives the result: 04-01-1980
More information on the SQLite date / time functions can be found here.
I am using Oracle SQL Developer 3.0.03. I am trying to upload an Excel file to an Oracle data table. I am getting an error for the date. The column in the database is a timestamp and I don't know what to put into the date format for the 'Data Load Wizard'. If I use the following format (dd-mom-yy hh.mi.ss), SQL Developer will show the following error:
--Insert failed for row 1 TIMESTAMP_COLUMN GDK-05047: A day of the month must be between 1 and the last day of the month.
--Row 1
INSERT INTO TABLE_1 (Column1, Column2, Column3, TIMESTAMP_COLUMN) VALUES ('Some Text','Some Text','Some more text',to_timestamp('40604.0', 'dd-mon-yy hh.mi.ss'));
The default number format IN EXCEL is: 40604.0
Which if you change the cell to a short date format you will get: 3/2/2011
I am trying to figure out what 'Format' I need to put into the FORMAT section of the 'DATA Load Wizard' that will accept the date format that is in EXCEL and upload it to Oracle as a TIMESTAMP.
I ran into the same thing today, and 'fixed' this two ways. The second way probably seems too complex, but it might help someone if they have a hard time automating the formatting of dates to look like Oracle's standard dd-mmm-yy.
Format the date columns in Excel as dd-mmm-yy and import directly into the table.
Highlight the column(s)
Choose "More Number Formats" where the existing format is (In Excel 2010, it says General in a dropbox on the Home tab
Select the last entry "Custom" in the Category box
Manually enter dd-mmm-yy in the Type: box
Format the date columns in Excel as mm/dd/yy, import the table in as text, write a manual insert statement from the temp text table using TO_DATE(date_field,'MM/DD/YYYY')
Highlight the column(s)
Choose "More Number Formats" where the existing format is (In Excel 2010, it says General in a dropbox on the Home tab
Select the "Date" entry in the Category box
Choose "03/14/01" from the list
The Excel "zero" date is January 0 (yes, zero), 1900. Since Oracle can't handle a "zero" day in a month you need to subtract 1 from the Excel value; thus the following should get you close to what you want:
SELECT CAST(TO_DATE('01-JAN-1900 00:00:00', 'DD-MON-YYYY HH24:MI:SS') AS TIMESTAMP) +
NUMTODSINTERVAL(41017.6361109954 - 1, 'DAY')
FROM DUAL
As far as "how to import it" goes, there's no Oracle date format that I'm aware of to do this. The easiest thing would be to import the value into a NUMBER column and then run a script to do the date conversion.
Share and enjoy.
yeah and that's the problem.
"A day of the month must be between 1 and the last day of the month."
1) how are these decimals created?
2) is this "04-06-2004" ? or are these seconds from 1970?
you need to add more detail about this number format!