.cfignore fails silently in non-UTF8 encodings

[jkillian]

If a .cfignore file is in an encoding other than UTF8, the CLI tool fails to read it properly and the files that should be ignored are uploaded with no indications that anything is wrong.

For example, on Windows, if a developer runs:

echo node_modules > .cfignore

the resulting file will by default be in an encoding other than UTF8. It's quite confusing for developers on Windows to figure out then why their .cfignore isn't being respected.

I realize this isn't an easy problem to fix, especially for Unicode files without a BOM. But perhaps we could at least look for some of the more common symptoms of different encodings and print out a warning message if they show up?

For example, say a user accidentally has a UTF16 .cfignore. The CLI could notice the unusual presence of lots of null bytes and print out a warning: "Warning: .cfignore file may be encoded in UTF16 and might not work properly."

Also see https://developer.ibm.com/answers/questions/29227/why-is-cf-push-choking-on-a-file-thats-in-cfignore.html

[cf-gitbot]

We have created an issue in Pivotal Tracker to manage this. You can view the current status of your issue at: https://www.pivotaltracker.com/story/show/81869448.

[goehmen]

@jkillian Thanks for raising this issue. The CF CLI team agreed with you but couldn't find a 100% accurate way to confirm encoding of a file across all supported platforms. We've acknowledged that this is a known issue and added a new section to the CLI's github readme.md file to accomodate this it is in Known issues section under Troubleshooting/FAQs.

[jkillian]

Agreed, I couldn't think up any pretty solutions either. Seems like a reasonable resolution to me, thanks for looking into it