I would like opinions/suggestions for improvement etc on the following function. It will be wrapped around every value being passed into a CSV file on a classic ASP page. As is
Few things:
This section:
If "" & txt = "" Then
insertCSV = ""
Else
If you just want to return an empty string if txt
is empty, you can just do this:
If Len(txt) = 0 Then Exit Function
You don't need to use End If
for single-line If
statements.
This line:
if isNumeric(tmp) AND left(tmp, 1) <> "0" then tmp = tmp end if
You're assigning the value back to itself? What purpose does this serve?
Don't you want to replace just the symbol ©
with ©
? The way you have it written, you're replacing the entire text with ©
(same goes for your other tests).
I would think you'd want to do this instead:
If InStr(tmp, "©") Then tmp = Replace(tmp, "©", "©")
Try making those changes and then post an updated version of your routine and let's see how it looks.
Update (based on latest comments)
If the database already has values stored like
’
then you need to consider that the web application is inserting the data using the wrong encoding and you are ending up with a mismatch in encoding in the database.
- Answer to Convert UTF-8 String Classic ASP to SQL Database
You don't need any of that the issue is Excel interprets the CSV as ASCII
when you are sending it as UTF-8
.
There is a useful trick that fools Excel into using the correct encoding rather then assuming ASCII, I use this technique all the time and works well. It involves writing a BOM (Byte Order Mark) to the stream before displaying your data that tells Excel the data is UTF-8
not ASCII
encoded.
'BOM to dup Excel into encoding the CSV correctly instead of using ASCII!!!
Call Response.BinaryWrite(ChrB(239) & ChrB(187) & ChrB(191))
This should be specified before you use Response.Write()
to output your text from the insertCSV()
function.
Using this kind of code
if instr(tmp, "Â") then
is just wrong when what the output (Â
) is trying to do is tell you something is wrong with the encoding.
Useful Links