overleaf-cep/services/web/frontend/js/features/source-editor/lezer-bibtex/bibtex.grammar

@tokens {
  whiteSpace { $[\t\n\r ]+ }

  // Anything outside of an entry at the top level is a BibTeX free-form comment.
  // A word may contain a "%" but may not start with one, so a "%" at a token
  // boundary still opens a comment: btparse parses the entry in
  // "Junk% @article{a}" but treats "Junk % @article{a}" as a comment.
  junkWord { ![@ \t\n\r%] ![@ \t\n\r]* }
  Junk { junkWord (whiteSpace junkWord)* }

  // BibLaTeX supports comments starting with a % and extending to the end of
  // the line
  Comment { "%" ![\n\r]* }

  // btparse derives a name by "removing the magic 10 from printable ASCII, then
  // further excluding @, \ and ~", the magic 10 being " # % ' ( ) , = { } and
  // space. Stating that by exclusion rather than by listing the allowed
  // punctuation also admits non-ASCII: btparse's own lexer is ASCII-only here,
  // but biber accepts UTF-8 throughout, including in citation keys.
  // See https://metacpan.org/dist/Text-BibTeX/view/btparse/doc/bt_language.pod
  identifierChar { ![\u{0}-\u{20}\u{7f}"#%'(),=@\\{}~] }

  // Identifiers cannot start with a digit, so that a value like 1995 is a
  // NumberLiteral rather than a macro name. Citation keys can.
  identifierStart { ![\u{0}-\u{20}\u{7f}"#%'(),=@\\{}~0-9] }

  identifier { identifierStart identifierChar* }

  EntryType { identifier }

  CitationKey { identifierChar+ }

  NumberLiteral { @digit+ }

  "#" "="
}

@external specialize {EntryType} specializeEntryType from "./tokens.mjs" {
  stringKeyword,
  preambleKeyword,
  commentKeyword
}

@skip { whiteSpace | Comment }

@top Bibliography {
  (Junk | Entry | CommentEntry | PreambleEntry | StringEntry)*
}

StringEntry { StringCommand StringBody }
PreambleEntry { PreambleCommand PreambleBody }
CommentEntry { CommentCommand CommentBody }
Entry { EntryCommand EntryBody }

EntryCommand { "@" EntryType }
EntryBody { commandBody<entryContents> }
StringCommand { "@" stringKeyword }
StringBody { commandBody<fields> }
PreambleCommand { "@" preambleKeyword }
PreambleBody { commandBody<Value> }
CommentCommand { "@" commentKeyword }

commandBody<contents> {
  (
    BodyOpen[closedBy="BodyClose"] { "{" }
    contents
    BodyClose[openedBy="BodyOpen"] { "}" }
  ) | (
    BodyOpen[closedBy="BodyClose"] { "(" }
    contents
    BodyClose[openedBy="BodyOpen"] { ")" }
  )
}

entryContents { CitationKey ("," fields)?  }

fields { (Field ("," Field)* ","?)?  }

Field { FieldName "=" Value }
FieldName { identifier }
Value { simpleValue ("#" simpleValue)* }
simpleValue { StringLiteral | NumberLiteral | StringName }
StringName { identifier }

@local tokens {
  strOpenBrace { "{" }
  strCloseBrace { "}" }
  strOpenParen { "(" }
  strCloseParen { ")" }
  strQuote { "\"" }
  @else strOther
}

@skip {} {
  StringLiteral {
    (
      StringOpen[closedBy="StringClose"] { "\"" }
      StringContents { quotedStringContents }
      StringClose[openedBy="StringOpen"] { strQuote }
    ) | (
      StringOpen[closedBy="StringClose"] { "{" }
      StringContents { bracedStringContents }
      StringClose[openedBy="StringOpen"] { strCloseBrace }
    )
  }

  CommentBody {
    (
      BodyOpen[closedBy="BodyClose"] { "{" }
      CommentContents { bracedStringContents }
      BodyClose[openedBy="BodyOpen"] { strCloseBrace }
    ) | (
      BodyOpen[closedBy="BodyClose"] { "(" }
      CommentContents { parenthesizedStringContents }
      BodyClose[openedBy="BodyOpen"] { strCloseParen }
    )
  }

  quotedStringContents {
    (
      strOpenBrace bracedStringContents strCloseBrace |
      strOpenParen |
      strCloseParen |
      strCloseBrace |
      strOther
    )*
  }

  parenthesizedStringContents {
    (
      strOpenBrace bracedStringContents strCloseBrace |
      strQuote |
      strOpenParen |
      strCloseBrace |
      strOther
    )*

  }

  bracedStringContents {
    (
      strOpenBrace bracedStringContents strCloseBrace |
      strQuote |
      strOpenParen |
      strCloseParen |
      strOther
    )*
  }
}

@external propSource highlighting from "./highlight.mjs"