Search the Community

Hi, I am trying to join two tables that have similar fields, but in slightly different formats. I was previously using something along the lines of like '%*%' Then SUBSTRING(b.fieldname,CHARINDEX(b.fieldname,'*',1)+1,9) However the format has now changed and I am struggling to put something together. Table A BD8K1 23161*35*201904231110*QULN Table B 23161*35*1 However the number of characters in both could vary so cant use that. Table A BD8K1 242875*10*201904251015*NBB1 Table B 242875*10*113 Is there some code that could be written that's something like: A – From the space use the value up to and including the 2nd * (BD8K1 is at the beginning of every value in table A so could potentially use from character 7) B – Use the value up to and including the 2nd * Any help would be very much appreciated. Thanks James

I am teaching a course on the ACSL competition to middle school students and my programming skills are little rusty. In the next competition we have a programming challenge where they need to create a program that you can input a string and several substring parameters. The problem is that they need to be able to input a negative number and have it produce the desired result. For example: with the string "Hello world!" input 1: 0, -1 output 1: Hello world input 2: 0,-5 output 2: Hello w input 3: -4,0 output 3: rld! I am at a little bit of a loss as how to go about getting the desired output.

I am trying to process a string which contains various data relating to an elearning record; Moodle SCORM suspend_data if you're familiar with it Since I do not have access to server side SQL/CLR functions through the Moodle interface, I am attempting to reference, and extract, the relevant information from the string by explicitly referencing the character locations of the information I need in the string and then converting them as required so they can be summed e.g. SELECT prefix_user.username,........ ((CASE(SUBSTRING(value_rep,237,1)) WHEN 'c' THEN 1 ELSE 0 END) + ........(CASE(SUBSTRING(value_rep,3659,1)) WHEN 'c' THEN 1 ELSE 0 END)) AS "# Videos watched"........FROM ((prefix_user LEFT JOIN (SELECT prefix_scorm_scoes_track.userid, Replace(value,"''","'XX'") AS value_repFROM prefix_scorm_scoes_track WHERE (((prefix_scorm_scoes_track.element)='cmi.suspend_data'))) AS vid ON prefix_user.id = vid.userid The string contains multiple occurrences of the string patterns '','n' and 'nn' where n is any digit 0-9 which I need to replace with the same string pattern e.g. 'XX', so that the total length of the string as a whole remains the same. I Initially thought that the pattern would always exist as '' or 'nn' so it was relatively simple to use a REPLACE(value,"''","XX") I understand it's not possible to use regex in the REPLACE function and have been investigating the use of PATINDEX with STUFF, but understand this is inly suitable for operating on 1 occurrence of a pattern. The report can be output to a CSV for viewing in Excel, so my conclusion is currently to carry out post-processing with Excel VBA where I will have access to loops and other functions. Is what I am trying to achieve possible without the use of a function on the SQL server?