Added encoding argument to TextReuseCorpus and TextReuseTextDocument

[davidfuhry]

When using Windows readLines will expect text files to be encoded in Windows-1252.

This would add an optional encoding argument to TextReuseCorpus as well as TextReuseTextDocument which can be used to explicitly specify the encoding of the input files (mostly UTF-8).

As it defaults to "unknown" which is the default for readLines this should maintain backward compatability.

Edit: I forgot to mention that this specific issue can be worked around by setting options(encoding = "UTF-8") before creating the corpus however this has some side effects so I still think having an encoding argument is the better way to deal with this.

[codecov-io]

Codecov Report

Merging #89 into master will increase coverage by 0.04%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #89      +/-   ##
==========================================
+ Coverage   86.77%   86.81%   +0.04%     
==========================================
  Files          25       25              
  Lines         658      660       +2     
==========================================
+ Hits          571      573       +2     
  Misses         87       87

Impacted Files	Coverage Δ
R/TextReuseTextDocument.R	91.52% <100%> (ø)	⬆️
R/TextReuseCorpus.R	82.69% <100%> (+0.33%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a96fec3...5d68a16. Read the comment docs.