Clever Geek Handbook

Enough already writing these regulars

Hello, my name is Dmitry Karlovsky and before I also used Perl for frontend development. Just look at what concise code you can parse, for example, an email:







/^(?:((?:[\w!#\$%&'\*\+\/=\?\^`\{\|\}~-]){1,}(?:\.(?:[\w!#\$%&'\*\+\/=\?\^`\{\|\}~-]){1,}){0,})|("(?:((?:(?:([\u{1}-\u{8}\u{b}\u{c}\u{e}-\u{1f}\u{21}\u{23}-\u{5b}\u{5d}-\u{7f}])|(\\[\u{1}-\u{9}\u{b}\u{c}\u{e}-\u{7f}]))){0,}))"))@(?:((?:[\w!#\$%&'\*\+\/=\?\^`\{\|\}~-]){1,}(?:\.(?:[\w!#\$%&'\*\+\/=\?\^`\{\|\}~-]){1,}){0,}))$/gsu


Here, however, several errors crept in. Well, nothing, we will fix it in the next release!







Jokes aside









As they grow, the regulars very quickly lose their clarity. It is not for nothing that there are dozens of services on the Internet for debugging regulars. Here are just a few of them:









, :







/(?<>(?<>\p{Script=Cyrillic})\p{Script=Cyrillic}+)/gimsu


, , . 5 :







/\t/
/\ci/
/\x09/
/\u0009/
/\u{9}/u


JS , ?







const text = 'lol;)'

// SyntaxError: Invalid regular expression: /^(lol;)){2}$/: Unmatched ')'
const regexp = new RegExp( `^(${ text }){2}$` )


, , :







const VISA = /(?<type>4)\d{12}(?:\d{3})?/
const MasterCard = /(?<type>5)[12345]\d{14}/

// Invalid regular expression: /(?<type>4)\d{12}(?:\d{3})?|(?<type>5)[12345]\d{14}/: Duplicate capture group name
const CardNumber = new RegExp( VISA.source + '|' + MasterCard.source )


, , , ! ?









JS. XRegExp:







  • API .
  • .
  • .
  • .
  • .
  • IDE.


, , , .









DSL, JS . PEG.js:







  • .
  • — .
  • .
  • IDE.
  • 2 .


.







, . .









TypeScript $mol_regexp:







  • .
  • IDE.
  • .
  • , .


. - ..











- , .







const {
    char_only, latin_only, decimal_only,
    begin, tab, line_end, end,
    repeat, repeat_greedy, from,
} = $mol_regexp


, NPM



import { $mol_regexp: {
    char_only, decimal_only,
    begin, tab, line_end,
    repeat, from,
} } from 'mol_regexp'
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    




// /4(?:\d){12,}?(?:(?:\d){3,}?){0,1}/gsu
const VISA = from([
    '4',
    repeat( decimal_only, 12 ),
    [ repeat( decimal_only, 3 ) ],
])

// /5[12345](?:\d){14,}?/gsu
const MasterCard = from([
    '5',
    char_only( '12345' ),
    repeat( decimal_only, 14 ),
])


:







  • .
  • .
  • .
  • . .
  • ( ).




// /(?:(4(?:\d){12,}?(?:(?:\d){3,}?){0,1})|(5[12345](?:\d){14,}?))/gsu
const CardNumber = from({ VISA, MasterCard })
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    




// /^(?:\t){0,}?(?:((?:(4(?:\d){12,}?(?:(?:\d){3,}?){0,1})|(5[12345](?:\d){14,}?))))(?:((?:\r){0,1}\n)|(\r))/gmsu
const CardRow = from(
    [ begin, repeat( tab ), {CardNumber}, line_end ],
    { multiline: true },
)
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    




const cards = `
    3123456789012
    4123456789012
    551234567890123
    5512345678901234
`
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    




for( const token of cards.matchAll( CardRow ) ) {

    if( !token.groups ) {
        if( !token[0].trim() ) continue
        console.log( ' ', token[0].trim() )
        continue
    }

    const type = ''
        || token.groups.VISA && ' VISA'
        || token.groups.MasterCard && 'MasterCard'

    console.log( type, token.groups.CardNumber )

}


, , . matchAll



, . $mol_regexp



. groups



. , , .









  3123456789012
 VISA 4123456789012
  551234567890123
MasterCard 5512345678901234


.







E-Mail



:







const {
    begin, end,
    char_only, char_range,
    latin_only, slash_back,
    repeat_greedy, from,
} = $mol_regexp

//      
const atom_char = char_only( latin_only, "!#$%&'*+/=?^`{|}~-" )
const atom = repeat_greedy( atom_char, 1 )
const dot_atom = from([ atom, repeat_greedy([ '.', atom ]) ])

//      
const name_letter = char_only(
    char_range( 0x01, 0x08 ),
    0x0b, 0x0c,
    char_range( 0x0e, 0x1f ),
    0x21,
    char_range( 0x23, 0x5b ),
    char_range( 0x5d, 0x7f ),
)

//     
const quoted_pair = from([
    slash_back,
    char_only(
        char_range( 0x01, 0x09 ),
        0x0b, 0x0c,
        char_range( 0x0e, 0x7f ),
    )
])

//   
const name = repeat_greedy({ name_letter, quoted_pair })
const quoted_name = from([ '"', {name}, '"' ])

//   :   
const local_part = from({ dot_atom, quoted_name })
const domain = dot_atom

// ,     
const mail = from([ begin, local_part, '@', {domain}, end ])


— . !







//  SyntaxError: Wrong param: dot_atom=foo..bar
mail.generate({
    dot_atom: 'foo..bar',
    domain: 'example.org',
})


, … :







// foo.bar@example.org
mail.generate({
    dot_atom: 'foo.bar',
    domain: 'example.org',
})


:







// "foo..bar"@example.org
mail.generate({
    name: 'foo..bar',
    domain: 'example.org',
})


.









, "" /snjat-dvushku/s-remontom/v-vihino



. , :







const translit = char_only( latin_only, '-' )
const place = repeat_greedy( translit )

const action = from({ rent: 'snjat', buy: 'kupit' })
const repaired = from( 's-remontom' )

const rooms = from({
    one_room: 'odnushku',
    two_room: 'dvushku',
    any_room: 'kvartiru',
})

const route = from([
    begin,
    '/', {action}, '-', {rooms},
    [ '/', {repaired} ],
    [ '/v-', {place} ],
    end,
])


:







// `/snjat-dvushku/v-vihino`.matchAll(route).next().value.groups
{
    action: "snjat",
    rent: "snjat",
    buy: "",
    rooms: "dvushku",
    one_room: "",
    two_room: "dvushku",
    any_room: "",
    repaired: "",
    place: "vihino",
}


, :







// /kupit-kvartiru/v-moskve
route.generate({
    buy: true,
    any_room: true,
    repaired: false,
    place: 'moskve',
})


true



, . false



, .







, , , .







?



, , . 2 , . . groups



:







// time.source == "((\d{2}):(\d{2}))"
// time.groups == [ 'time', 'hours', 'minutes' ]
const time = from({
    time: [
        { hours: repeat( decimal_only, 2 ) },
        ':',
        { minutes: repeat( decimal_only, 2 ) },
    ],
)


, exec



- groups



:







{
    time: '12:34',
    hours: '12,
    minutes: '34',
}


, , , , :







// time.source == "((\d{2}):(\d{2}))"
// time.groups == [ 'time', 'minutes' ]
const time = wrong_from({
    time: [
        /(\d{2})/,
        ':',
        { minutes: repeat( decimal_only, 2 ) },
    ],
)
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    


{
    time: '12:34',
    hours: '34,
    minutes: undefined,
}


, , "" "0", "1" . — , , :







new RegExp( '|' + regexp.source ).exec('').length - 1


, String..match



String..matchAll



exec



. , , Symbol.match



Symbol.matchAll



. :







*[Symbol.matchAll] (str:string) {
    const index = this.lastIndex
    this.lastIndex = 0
    while ( this.lastIndex < str.length ) {
        const found = this.exec(str)
        if( !found ) break
        yield found
    }
    this.lastIndex = index
}


, , :







interface RegExpMatchArray {
    groups?: {
        [key: string]: string
    }
}


, :







interface String {

    match< RE extends RegExp >( regexp: RE ): ReturnType<
        RE[ typeof Symbol.match ]
    >

    matchAll< RE extends RegExp >( regexp: RE ): ReturnType<
        RE[ typeof Symbol.matchAll ]
    >

}


TypeScript groups



, - .







, .











— , , , - ( ) .



















More articles:


All Articles