Isaac
Lifelong Learner
- Local time
- Today, 03:45
- Joined
- Mar 14, 2017
- Messages
- 8,777
So we have a couple Sharepoint lists. Because we have so many hands in the pot (unfortunately), my manager is (understandably) worried and wants me to back them up nightly. Also unfortunately, I don't have a lot of control over the input data--Meaning specifically, I can't prevent people from entering quotes, tabs, weird symbols, carriage returns etc.
My preference is kind of to export them to a flat, delimited file. I'm writing my own recordset/textstream code because I don't feel TransferText gives me enough control. I'm actually open to TransferText if I learn something that makes me realize it will be good enough.
My question is something that has always been a classic dilemma for me. "What is the SAFEST - most FOOL-PROOF way to delimit this file?"
I generally like to use odd characters, like pipe to delimit columns, caret to quote text. But regardless of whether I go typical (comma, quote) or weird symbols, what's to stop someone from having any of those characters in the natural data??
Is there really any guideline/answer to this question, or is it just, the more cryptic and least likely to occur in the natural data, the better off you are?
(Which is to say, I'd do something really weird, like write code to create a text file with columns delimited & quoting by multi-characters).
Am I over complicating it? If you think "yes", please be specific....as in, what do you suggest that solves the problem of, What if that delimiter occurs naturally in the data?
Any feedback welcome--thanks.
My preference is kind of to export them to a flat, delimited file. I'm writing my own recordset/textstream code because I don't feel TransferText gives me enough control. I'm actually open to TransferText if I learn something that makes me realize it will be good enough.
My question is something that has always been a classic dilemma for me. "What is the SAFEST - most FOOL-PROOF way to delimit this file?"
I generally like to use odd characters, like pipe to delimit columns, caret to quote text. But regardless of whether I go typical (comma, quote) or weird symbols, what's to stop someone from having any of those characters in the natural data??
Is there really any guideline/answer to this question, or is it just, the more cryptic and least likely to occur in the natural data, the better off you are?
(Which is to say, I'd do something really weird, like write code to create a text file with columns delimited & quoting by multi-characters).
Am I over complicating it? If you think "yes", please be specific....as in, what do you suggest that solves the problem of, What if that delimiter occurs naturally in the data?
Any feedback welcome--thanks.