How to capture the file delimiter information(whether semicolon/pipe/comma separated) from the source file and compare it with the existing template in Database.
That's an interesting problem. If the delimiter can only be one of several things (e.g. comma, semi-colon, etc.), and the data itself doesn't contain a lot of possible delimiters in close sequence, then you could read the first several lines and count the number of commas, etc. in each. If the first 10 lines each contain 20 commas, but only 3 semi-colons and zero pipes, then your delimiter is probably the comma.
The more potential delimiters you have in the actual data, the more rows you need to read to be sure you've found the actual delimiter; even so, if you read 1000 rows, and every single one has exactly 20 commas, then the probability of the comma *not* being the delimiter is vanishingly small (I'm tempted to estimate it based on the relative density of each possible delimiter, but it's late, and I've had a long day).
I have to capture the Delimiter which is coming from the Source (don't know what type of delimiter) and i have to compare the source delimiter value with the Template Delimiter_Type (as above). If both are matching i should process the file further or else i should reject the source file.
