Parsing a string with regex to build JSON

Question

I'm pretty sure the code is of mild quality but not entirely sure as I'm not much of a JavaScript programmer. Everything works as expected. A quick rundown of the code. It simply parses a string and builds some game data for saving into a database. One thing I'm not sure of is how I should handle data that doesn't conform to how it's expected.

// var diplo = `!report Game Type: Diplo Early Bird
// 1: <@12321321421412421>
// 2: <@23423052352342334>
// 3: <@45346346345343453>
// 4: <@23423423423523523>`

var diplo = `!report Game Type: Team Modded
Team: 1
    <@23452342359234534>
    <@34642342359232352>
Team: 2
    <@34534342359234879>
    <@58493457202345054>`


var report = {};


function gameType(game) {
    let data = {};
    let game_types = [{id: 1, name: 'Diplo'}, {id: 2, name: 'Always War'}, {id: 3, name: 'FFA'}
        , {id: 4, name: 'No Diplo'}, {id: 5, name: 'Team'}, {id: 6, name: 'Duel'}, {id: 7, name: 'No War'}];
    let o = {}
    let reType = /!report Game Type:[\s+]?(\d|\w+)\s?(\w+\s?(\w+)?)?\n/gi
    let match = reType.exec(game)
    if(match[1]){
        for (var j = 0; j < game_types.length; j++) {
            if ((game_types[j].name).toLowerCase() === (match[1]).toLowerCase()) {
                data.type = game_types[j].id;
                break;
            }
        }
    }
    if(match[2]){
        data.moddifier = match[2]
    }
    diplo = game.slice(reType.lastIndex);
    return (data)
}

function getPlayers(game) {
    let players = [];
    rePlayer = /(\d+):\s?<@(\d+)>[\s\S]?(\w+\s?\w+)?/gi;
        do {
            let o = {};
            player = rePlayer.exec(game);
            if (player) {
                o.position = player[1]
                o.discord_id = player[2]
                players.push(o);
            }
        } while (player);
        return players;
}

function getTeamPlayers(game) {
    let players = [];
    reTeam = /Team:\s(\d)[\s\S](\s+?<@(\d+)>)+/gi;
    rePlayer = /\s+?<@(\d+)>/gi;
    let i = 1;
    do {
        team = reTeam.exec(game);
        if(team){
            do {
                player = rePlayer.exec(team[0]);
                if(player){         
                    let o = {}; 
                    o.team_id = i;
                    o.position = team[1];
                    o.discord_id = player[1];
                    players.push(o);
                }
            }while(player)
        }
        i++;
    }while(team)
    return players;
}

function main() {

    console.log("Reading game data...")
    try {
        report = gameType(diplo)
        if(report.type === 5){
            report.players = getTeamPlayers(diplo);
        }else{
            report.players = getPlayers(diplo);
        }
        console.log(report)
    } catch (err) {
        return 'Error Processing Report!';
    }        
}

main();

woxxom · Accepted Answer · 2017-09-16 21:51:25Z

I would manipulate RegExp.lastIndex with the sticky flag to avoid slicing the source string:

function parseGame(text) {
  const RX_HEADER = /!report Game Type:\s*(\d|\w+)(?:\s*(\w.*))?/y;
  const RX_TEAM = /Team:\s*(\d+)/y;
  const RX_DISCORD = /<@(\d+)>/y;
  const RX_POSITION_DISCORD = /(\d+):\s*<@(\d+)>\s*(?:\w+\s?\w+)?/y;
  const RX_WHITESPACE = /\s+/y;

  let pos = 0;
  const data = {players: []};

  function match(rx) {
    rx.lastIndex = pos;
    const matches = rx.exec(text);
    if (!matches) {
      return [];
    }
    pos = rx.lastIndex;
    if (pos < text.length) {
      RX_WHITESPACE.lastIndex = pos;
      RX_WHITESPACE.exec(text);
      pos = RX_WHITESPACE.lastIndex || pos;
    }
    return matches;
  }

  function readHeader() {
    const ID_NAME_MAP = {
      1: 'Diplo',
      2: 'Always War',
      3: 'FFA',
      4: 'No Diplo',
      5: 'Team',
      6: 'Duel',
      7: 'No War',
    };
    let [_, gameType, gameMod] = match(RX_HEADER);
    if (!gameType) {
      return;
    }
    gameType = gameType.toLowerCase();
    for (const type in ID_NAME_MAP) {
      if (type === gameType || ID_NAME_MAP[type].toLowerCase() === gameType) {
        data.type = Number(type);
        break;
      }
    }
    if (gameMod) {
      data.moddifier = gameMod;
    }
  }

  function readTeams() {
    let teamId;
    while ((teamId = Number(match(RX_TEAM)[1]))) {
      let position = 0;
      let discordId;
      while ((discordId = match(RX_DISCORD)[1])) {
        data.players.push({
          teamId,
          discordId,
          position: ++position,
        });
      }
    }
  }

  function readPlayers() {
    while (true) {
      const [_, position, discordId] = match(RX_POSITION_DISCORD);
      if (position) {
        data.players.push({
          discordId,
          position: Number(position),
        });
      } else {
        break;
      }
    }
  }

  readHeader();
  if (data.type === 5) {
    readTeams();
  } else {
    readPlayers();
  }
  return data;
}

And just to nitpick:

[\s+]? matches whitespace and a + which seems weird so I replaced it with \s*
[\s\S]? in rePlayer is ambiguous as it allows any single character, which seems weird.
I would expect only whitespace and new line(s) allowed e.g. \s*
I usually specify a match group is unused via ?: e.g. (?:foo)?
Indent the code consistently
Consistently use spaces before opening braces e.g. if (foo) { instead of if(foo){
No need for var generally if you use ES6 anyway
Use const for variables you don't reassign (more info)
Use a consistent type for the id and position instead of mixing strings & numbers
Use a JS linter like eslint to catch the errors like a global player variable
Use camelCase for properties and variable names
Use object literals like arr.push({foo: 'bar'}) without an interim variable

Sweet, thanks for all the insight. Does the sticky flag, apply through different functions? — user1620152
– user1620152, Commented Sep 17, 2017 at 1:20
Yes, match and replace also respect lastIndex which is very handy in performance-critical code where you can do pos=str.indexOf('foo') first and then proceed with a complex regexp at the found position without slicing the string. — woxxom
– woxxom, Commented Sep 17, 2017 at 9:07

Stack Exchange Network

Parsing a string with regex to build JSON

1 Answer 1

You must log in to answer this question.

Hot Network Questions

Parsing a string with regex to build JSON

1 Answer 1

You must log in to answer this question.

Related

Hot Network Questions