html-parser-benchmark

HTML parsers benchmark

• HTMLCleaner
• HTMLUnit
• NekoHTML
• JFiveParse
• JSoup
• Jerry
• Jericho

Running the benchmark

1. Download the source code and build it (mvn clean install)
2. Run the entire benchmark suite with java -jar target/html-parser-bench.jar
3. (Optional) To save resutls to csv use java -jar target/html-parser-bench.jar -rff results.csv -rf csv

(This benchmark is based on https://github.com/mbosecke/template-benchmark but changed to benchmark HTML parsers instead of templating engines)

Interpreting the Results

The benchmarks measure throughput, given in "ops/time". The time unit used is seconds. The tests benchmark how fast a small html fragment (h2 element with id "mp-dyk-h2") is extracted from the wikipedia.html (wikipedia's main page) file. The score represents the number of times this operation is done per second; the higher the score, the better.

Example Results

Benchmark                    Mode  Cnt     Score   Error  Units
HTMLCleanerBench.benchmark  thrpt        127,902          ops/s
HtmlUnitBench.benchmark     thrpt        184,343          ops/s
NekoHTMLBench.benchmark     thrpt        465,414          ops/s
JFiveParseBench.benchmark   thrpt        788,918          ops/s
JsoupBench.benchmark        thrpt        904,262          ops/s
JerryBench.benchmark        thrpt        979,093          ops/s
JerichoBench.benchmark      thrpt       3156,872          ops/s