So I'm bugfixing on a program that allows a user to create an Excel document, which it first zip compresses before serving to the customer. One of the client complaints is that, while the zip file can always be downloaded, sometimes the user must copy the Excel file out of the archive (or extract it) before opening it, using the standard Windows compressed tools. Standard "File name too long" error.
The algorithm is, of course, setting the .zip archive name to the same as the report - which can be 100+ character long strings.
The solution I'm implementing is to check the length of the potential report to see it if it violate MAX_PATH, naturally, and truncate the .zip name as needed.
Testing it on Windows 7, this works perfectly. But something odd happens when testing it under 8.1
It still throws an error trying to open from the archive - but this error is a bit more enigmatic.
"Sorry, we couldn't find C:\Users{My user name}\AppData\Local\Temp\Temp1_{Rest of the truncated archive name}.zip{Full file name}.xlsx. Is it possible it was moved, renamed or deleted?"
This error keeps popping up, regardless of whether the file is "Open"ed or "Save"d from the browser.
Normally, I'd just try further tweaking, but testing on the Windows 8.1 platform involves a lot of overhead at the moment, and it doesn't look like the problem is the path name length.
What is going on? Does Windows 8 have problem with the length of names within archives, or...?
Also, the reason I am posting this here, and not, say, User Experience, is because I feel like the solution will be programmatic - something in the code of the program. I readily concede that "allow the end user to skip hitting the Extract button" is about providing an easy user experience, but truncating the length of the entire path solved the problem on Windows 7.
Just not 8.1, and googling/searching the SO site family provides no help - ironically, because of the keyword Excel.
