BojanFaletic / parseTree

Parser

Efficient string decoding algorithm. It is implemented as parse tree with minimal memory allocations. This algorithm runs in linear time as function of input length. Naive searching would take N compares to all possible combinations, parser dynamically prune imposable strings thus reduce search time.

Warning

This is still in development and not yet sufficiently tested for deploying in production.

  parser_t *root;
  parser_init(&root);

  // set parsing values
  parser_add("Hello", 1, root);
  parser_add("Hey", 2, root);

  parser_parse("Hello", root);
  parser_parse("Hey", root);

  parser_free(root);

Benchmark

Benchmark	Time	CPU	Iterations
ParserParse	10.3 ns	10.3 ns	64872132
NaiveParse	12.3 ns	12.3 ns	56028453
Load_10_000_elements	223447185 ns	223327596 ns	3
Parse_10_000_elements	33.8 ns	33.8 ns	20078692
Parse_10_000_elements_naive	14606 ns	14600 ns	47617

Result of running Google Benchmark on test/benchmark.cpp. Parser is faster than naive strcmp.