Remove data.table

[kbenoit]

Removes the dependency on data.table, which was previously used in

• textstat_lexdiv()
• textstat_readability()
• textstat_frequency()

(It was also used in legacy textstat_keyness() functions, but these were removed in #30.)

I checked these against the data.table versions and the performance is equivalent. textstat_frequency() is still slightly faster with the older data.table due to the efficiency of frank(), but it is still acceptably fast using base::rank().

Solves #5 (and one more for a bonus)

[codecov[bot]]

Codecov Report

Merging #31 (66f6054) into master (63111eb) will decrease coverage by 0.16%. The diff coverage is 98.42%.

@@            Coverage Diff             @@
##           master      #31      +/-   ##
==========================================
- Coverage   98.46%   98.29%   -0.17%     
==========================================
  Files          12       12              
  Lines         780      761      -19     
==========================================
- Hits          768      748      -20     
- Misses         12       13       +1     

Impacted Files	Coverage Δ
R/textstat-methods.R	88.88% <66.66%> (ø)
R/textstat_readability.R	99.50% <98.80%> (-0.50%)	:arrow_down:
R/nscrabble.R	100.00% <100.00%> (ø)
R/nsyllable-methods.R	100.00% <100.00%> (ø)
R/textstat_entropy.R	100.00% <100.00%> (ø)
R/textstat_frequency.R	100.00% <100.00%> (ø)
R/textstat_lexdiv.R	97.90% <100.00%> (ø)
R/textstat_simil.R	97.95% <100.00%> (ø)

---

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 63111eb...c7a49b1. Read the comment docs.