Handle ZWNBSP characters

[JackStuart]

Fixes handling ZWNBSP characters at the start of a file - Seems to be very common in XML files

Renamed File2 to File1 due to there being no File1.xml - So there was no point skipping to File2

Fixed typo of declaratioon

Resolves the issue here - mandiant/flare-vm#640

[flwyd]

Thanks for highlighting this issue!

I think a little more work needed to properly handle byte order marks. Skipping the first line if the file starts with a BOM isn't necessarily correct: it works for file types covered by the head/hashBang list (like XML and shell scripts). But for types without this first-line-is-important feature, the copyright should go on the first line, right after the BOM.

Please add a unit test which covers at least

• XML: {BOM}<?xml\nLicense\n<!-- Normal comment -->
• Shell script or other language with a shebang: {BOM}#!/bin/sh\nLicense\n# Normal comment
• Octothorpe without a shebang: {BOM}License\n# Normal comment
• Non-octothorpe language, e.g. C++: {BOM}License\n// Normal comment

Since Go strings are UTF-8, these tests will need to use byte arrays directly, with big-endian and little-endian versions of the BOM sequence.

[flwyd]

Friendly ping @JackStuart regarding the comments above.