Add some safeties around charset detection and transliteration

[uberbrady]

In some environments, you need to specify character sets for iconv() in UPPERCASE. So this adds that.

In addition, sometimes the character set transformation can fail (such as when it's not in uppercase), and while that only usually raises an E_NOTICE, Laravel will tend to transform those notices into Exceptions, so we need to catch those.

And finally, if you don't happen to have iconv installed at all, we should gracefully detect that and default to the 'regular' behavior.

[what-the-diff]

PR Summary

• Improved Encoding Conversion Process
  The team has updated the way our system converts encoding to better handle errors that might occur during the process.

• Error Logging Added to Transliteration
  We have added a logging feature to the transliteration process. Now, if the shift from a detected encoding to UTF-8 fails for any reason, we will still have a record of the incident.

• Adjusted Formatting of Detected Encoding Strings
  We've made changes to how our system deals with the detected encoding strings. Now, it converts these strings to uppercase before handing them to the conversion function, which should help with uniformity.

• Replaced Transliteration Code Block
  The old piece of code in charge of changing file contents was removed and replaced with a more robust error-handling structure. This should enhance the resilience and efficiency of the process.

• Memory Optimization Post-Processing
  Finally, we've inserted a function to nullify the variable holding file contents once they have been processed. This change will help optimize the memory usage of our system.