benchmarks/json/bench_json.sh

   1 #!/bin/bash
   2 # This file is part of NIT ( http://www.nitlanguage.org ).
   3 #
   4 # Licensed under the Apache License, Version 2.0 (the "License");
   5 # you may not use this file except in compliance with the License.
   6 # You may obtain a copy of the License at
   7 #
   8 #     http://www.apache.org/licenses/LICENSE-2.0
   9 #
  10 # Unless required by applicable law or agreed to in writing, software
  11 # distributed under the License is distributed on an "AS IS" BASIS,
  12 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13 # See the License for the specific language governing permissions and
  14 # limitations under the License.
  15
  16 # Shell script to bench json parsers over different documents
  17
  18 source ../bench_common.sh
  19 source ../bench_plot.sh
  20
  21 ## CONFIGURATION OPTIONS ##
  22
  23 # Default number of times a command must be run with bench_command
  24 # Can be overrided with 'the option -n'
  25 count=5
  26
  27 ## HANDLE OPTIONS ##
  28
  29 function init_repo()
  30 {
  31         echo "Preparing submodules"
  32         git submodule update --init
  33         echo "Submodules ready"
  34         mkdir -p inputs
  35         echo "Preparing data for benchmarks"
  36         if [ ! -e inputs/large_escaped.json ]; then
  37                 echo "Downloading file 1/4"
  38                 wget -O inputs/large_escaped.json https://github.com/seductiveapps/largeJSON/blob/master/100mb.json?raw=true
  39         fi
  40         echo "File 1/4 ready"
  41         if [ ! -e inputs/magic.json ]; then
  42                 echo "Downloading file 2/4"
  43                 wget -O inputs/magic.json http://mtgjson.com/json/AllSets-x.json
  44         fi
  45         echo "File 2/4 ready"
  46         if [ ! -e inputs/big_twitter.json ]; then
  47                 echo "Downloading file 3/4"
  48                 wget -O inputs/twitter.json https://github.com/miloyip/nativejson-benchmark/raw/master/data/twitter.json
  49                 cd inputs
  50                 ./multiply_twitter.sh
  51                 rm twitter.json
  52                 cd ..
  53         fi
  54         echo "File 3/4 ready"
  55         if [ ! -e inputs/big_gov_data.json ]; then
  56                 echo "Downloading file 4/4"
  57                 wget -O inputs/gov_data.json https://edg.epa.gov/data.json
  58                 cd inputs
  59                 ./multiply_gov.sh
  60                 rm gov_data.json
  61                 cd ..
  62         fi
  63         echo "File 4/4 ready"
  64 }
  65
  66 function usage()
  67 {
  68         echo "run_bench: ./bench_json.sh [options]"
  69         echo "  -v: verbose mode"
  70         echo "  -n count: number of execution for each bar (default: $count)"
  71         echo "  -fast: check only Nit"
  72         echo "  -h: this help"
  73 }
  74
  75 stop=false
  76 fast=
  77 while [ "$stop" = false ]; do
  78         case "$1" in
  79                 -v) verbose=true; shift;;
  80                 --fast) fast=true; shift;;
  81                 -h) usage; exit;;
  82                 -n) count="$2"; shift; shift;;
  83                 *) stop=true
  84         esac
  85 done
  86
  87 init_repo
  88
  89 mkdir -p out
  90 html="index.html"
  91 echo >"$html" "<html><head></head><body>"
  92
  93 echo "Compiling engines"
  94
  95 if [ -z "$fast" ]; then
  96         declare -a script_names=('C' 'Python 3' 'Python 2' 'Go' 'Nit Ad-hoc UTF-8 No Ropes' 'Nit Ad-hoc UTF-8 + Ropes' 'Ruby ext')
  97         declare -a script_cmds=('./scripts/c_parser' 'python3 scripts/python.py' 'python2 scripts/python.py' './scripts/json_parse' './scripts/nit_adhoc_utf_noropes' './scripts/nit_adhoc_utf_ropes' 'ruby scripts/json_ext.rb')
  98
  99         echo "C JSON Parser"
 100
 101         gcc -O2 -I thirdparty/ujson4c/src -I thirdparty/ujson4c/3rdparty/ thirdparty/ujson4c/3rdparty/ultrajsondec.c scripts/c_parser.c -o scripts/c_parser -lm
 102
 103         echo "Go JSON Parser"
 104
 105         go build -o scripts/json_parse scripts/json_parse.go
 106 else
 107         declare -a script_names=('Nit Ad-hoc UTF-8 No Ropes' 'Nit Ad-hoc UTF-8 + Ropes')
 108         declare -a script_cmds=('./scripts/nit_adhoc_utf_noropes' './scripts/nit_adhoc_utf_ropes')
 109
 110 fi
 111
 112 echo "Nit/NitCC Parser"
 113
 114 nitc --semi-global scripts/nitcc_parser.nit -o scripts/nitcc_parser
 115
 116 echo "Nit/Ad-Hoc UTF-8 Parser, No Ropes"
 117
 118 nitc --semi-global scripts/nit_adhoc_utf_noropes.nit -o scripts/nit_adhoc_utf_noropes
 119
 120 echo "Nit/Ad-Hoc UTF-8 Parser, With Ropes"
 121
 122 nitc --semi-global scripts/nit_adhoc_utf_ropes.nit -o scripts/nit_adhoc_utf_ropes
 123
 124
 125 for script in `seq 1 ${#script_cmds[@]}`; do
 126         echo "Preparing res for ${script_names[$script - 1]}"
 127         prepare_res "./out/${script_names[$script - 1]}.dat" "${script_names[$script - 1]}" "${script_names[$script - 1]}"
 128         for file in inputs/*.json; do
 129                 fname=`basename $file .json`
 130                 bench_command $file "Benching file $file using ${script_cmds[$script - 1]} parser" ${script_cmds[$script - 1]} $file
 131         done;
 132 done;
 133
 134 rm scripts/nitcc_parser
 135 rm scripts/json_parse
 136 rm scripts/c_parser
 137 rm scripts/nit_adhoc_utf_noropes
 138 rm scripts/nit_adhoc_utf_ropes
 139
 140 plot out/bench_json.gnu
 141
 142 echo >>"$html" "</body></html>"
 143
 144 if test -n "$died"; then
 145         echo "Some commands failed"
 146         exit 1
 147 fi
 148 exit 0