sys.path[0:0]= [os.path.join(os.path.dirname(__file__),"..")]

defbench_parse(fh,treebuilder):

fh.seek(0)

html5lib.parse(fh,treebuilder=treebuilder,useChardet=False)

defbench_serialize(loops,fh,treebuilder):

fh.seek(0)

doc=html5lib.parse(fh,treebuilder=treebuilder,useChardet=False)

range_it=range(loops)

t0=pyperf.perf_counter()

forloopsinrange_it:

html5lib.serialize(doc,tree=treebuilder,encoding="ascii",inject_meta_charset=False)

returnpyperf.perf_counter()-t0

BENCHMARKS= ["parse","serialize"]

defadd_cmdline_args(cmd,args):

ifargs.benchmark:

cmd.append(args.benchmark)

if__name__=="__main__":

runner=pyperf.Runner(add_cmdline_args=add_cmdline_args)

runner.metadata["description"]="Run benchmarks based on Anolis"

runner.argparser.add_argument("benchmark",nargs="?",choices=BENCHMARKS)

args=runner.parse_args()

ifargs.benchmark:

benchmarks= (args.benchmark,)

else:

withopen(os.path.join(os.path.dirname(__file__),"data","html.html"),"rb")asfh:

source=io.BytesIO(fh.read())

if"parse"inbenchmarks:

fortbin ("etree","dom","lxml"):

runner.bench_func("html_parse_%s"%tb,bench_parse,source,tb)

if"serialize"inbenchmarks:

fortbin ("etree","dom","lxml"):

runner.bench_time_func("html_serialize_%s"%tb,bench_serialize,source,tb)

sys.path[0:0]= [os.path.join(os.path.dirname(__file__),"..")]

defbench_html5lib(fh):

fh.seek(0)

html5lib.parse(fh,treebuilder="etree",useChardet=False)

defadd_cmdline_args(cmd,args):

ifargs.benchmark:

cmd.append(args.benchmark)

BENCHMARKS= {}

forroot,dirs,filesinos.walk(os.path.join(os.path.dirname(os.path.abspath(__file__)),"data","wpt")):

forfinfiles:

iff.endswith(".html"):

BENCHMARKS[f[:-len(".html")]]=os.path.join(root,f)

if__name__=="__main__":

runner=pyperf.Runner(add_cmdline_args=add_cmdline_args)

runner.metadata["description"]="Run parser benchmarks from WPT"

runner.argparser.add_argument("benchmark",nargs="?",choices=sorted(BENCHMARKS))

args=runner.parse_args()

ifargs.benchmark:

benchmarks= (args.benchmark,)

else:

benchmarks=sorted(BENCHMARKS)

forbenchinbenchmarks:

path=BENCHMARKS[bench]

withopen(path,"rb")asfh:

fh2=io.BytesIO(fh.read())

runner.bench_func(name,bench_html5lib,fh2)

The files in this data are derived from:

*`html.html`: from[html](http://github.com/whatwg/html), revision

77db356a293f2b152b648c836b6989d17afe42bb. This is the first 5000 lines of`source`. (This is

representative of the input to[Anolis](https://bitbucket.org/ms2ger/anolis/); first 5000 lines

chosen to make it parse in a reasonable time.)

*`wpt`: see`wpt/README.md`.