URLEncode + Base64Encode

Lounge (Discussions)
1

From time to time, I've noticed that Base64 encoding alone won't necessarily make it across the wire without some problem character. Therefore, I also wrap that Base64 encoded file with a GetAsUrlEncoded() function.

That extra GetAsUrlEncoded() adds to the time it takes to get the file encoded and on its way, so I was just wondering if anyone has noticed this problem character behavior using Base64 encoding alone.

Thanks,

2

I remember having this kind of issue 2 years ago. A Xojo Web app communicating with FMS experienced that - I think that was when FMS sent back the Base64 encoded file.

In my case it had to do with CR-LF sequences that were breaking decoding at the other end.

I googled to find the thread I crated on the Community, but didn't find it :rage:.

If I find something about my issue I will post back.

1 Like
3

Cool, yeah, the getAsUrlEncoding() is a workaround and I've never had a problem wrapping Base64 with GetAsUrlEncoded, but I was just wondering if anybody else had experienced this same issue.

To wit, GetAsUrlEncoded(Base64Encode(<container_field>)) always works.

Appreciate your reply.

Thanks, P.

4

Hi @anon45965781,

Just a note that, subsequent to the introduction of the Base64Encode function, FMP introduced the function: Base64EncodeRFC, which I believe should allow for encoding without line breaks interspersed in the output. I would think that should help, but I do not do enough work that involves Base64 encoding to be able to definitively make such a claim.

Hope this helps!

-steve

4 Likes
5

Thanks Steve! I'm not sure in my case it's a line break issue, but that's super good information to be reminded about. Perhaps this RFC idea warrants a test. If I can get away with not needing to use an extra encoding function, I would probably save a bit of time.

Perhaps @WimDecorte will jump in here also? :slight_smile:

FWIW, I removed the GetAsUrlEncoded and just used Base64EncodeRFC() and no problems transferring file types like PDF, ZIP, HTML, and Open Document Format. Don't see a performance improvement, but that's OK.

Thanks again.

4 Likes
6

Hi @anon45965781,

Agreed that the source of the issue might not be limited to just the presence of line-breaks,

Edit: Good on you for giving it a spin! Glad to see that it is working.

All the best,

-Steve

2 Likes
7

Thanks Steve.

Since I control the service side too, it's not a big deal to remove the code that does the UrlDecode() and just rebuild and redeploy. :slight_smile:

I'll try tomorrow with other file types, but it looks like there's no problem.

These FMP functions are rock solid.

4 Likes