Towards a Faster Incremental Packrat Parser

This work is an extension to the research done by Dubroy and Warth [1] on incremental packrat parsing. It attempts to further decrease the parse time of the incremental parser using two techniques from experimental algorithmics: algorithm and code tuning. Two modifications were introduced to the original incremental packrat parser: reduction of copy operations, and the use of an array of overlapping entries. Initial results showed that there is a significant difference between the parse times of the original parser (μ = 15.64ms, σ = 69.29ms) and the modified parser (μ = 4.60ms, σ = 69.53ms), with a t-score of t(890)=52.26 and a p -value of 9.41 ×10
^-274
. This means that faster parse times are achieved when using the modified parser.