Add detect encoding with BOM: UTF-7 and GB-18030

[rstm-sf]

Resolve #79

Add detect

• [x] UTF-7 BOM
• [x] GB-18030 BOM

[rstm-sf]

Simplification of checks in FindCharSetByBom because len <= buf.Length always

So in the end it is called from the following places https://github.com/CharsetDetector/UTF-unknown/blob/cb3dca2a51d4ccf769687cd18eebd12d60e8874b/src/CharsetDetector.cs#L142

https://github.com/CharsetDetector/UTF-unknown/blob/cb3dca2a51d4ccf769687cd18eebd12d60e8874b/src/CharsetDetector.cs#L199-L201

via

https://github.com/CharsetDetector/UTF-unknown/blob/cb3dca2a51d4ccf769687cd18eebd12d60e8874b/src/CharsetDetector.cs#L269

https://github.com/CharsetDetector/UTF-unknown/blob/cb3dca2a51d4ccf769687cd18eebd12d60e8874b/src/CharsetDetector.cs#L283

https://github.com/CharsetDetector/UTF-unknown/blob/cb3dca2a51d4ccf769687cd18eebd12d60e8874b/src/CharsetDetector.cs#L297-L299

[rstm-sf]

It's bug or feature?

[rstm-sf]

At least add verification will be more efficient

[rstm-sf]

It seemed like it would be better, because otherwise, you need to add a length check every time

[304NotModified]

Thanks for the refactor also :)