thinking through parsing multi-line database commands with cr in the data

[Dave Probert]

Hi,

I found that the simplest route for me when working with multi-line field
data was to pre-parse the data on storage and post parse it on extraction,
by changing the 'cr''s to either another seldom used character or html (eg.
'<br>').

So the database code before the INSERT/UPDATE would replace and cr's with
'<br>', and once a SELECT is done either a single function to replace the
'<br>'s with cr in all the data returned (possibly multi row) or to do the
replace at the usage of the individual row/field.

Doing it this way means that the database manager (Postgres/Mysql/Sqlite)
doesn't really know or care about multi-line data.  But the code handling
the data should be carefully constructed to make sure that there are no
instances of the data getting through to the end user (view OR edit) with
the replaceable characters in.

Obviously if you are using actual html content then it's even easier -
simply replace ALL cr's with empty on write.  And replace '<br>' and '</p>'s
with the same AND a cr (for editing use).

The whole process comes down to a couple of lines of code in the end and
making sure you document which fields you do a morphing of the data in - so
that it's easier to remember in a years time ;) - if necessary!

Hope that helps a bit,
Dave



--
View this message in context: http://runtime-revolution.278305.n4.nabble.com/thinking-through-parsing-multi-line-database-commands-with-cr-in-the-data-tp4665431p4665448.html
Sent from the Revolution - User mailing list archive at Nabble.com.