Repository Public
1// ORIX Docs2// File: Packages/ORIXDocsKit/Sources/DocReferences/NameParser.swift3// Description: BibTeX name-list parsing ("First von Last", "von Last, First", "von Last, Jr, First", "and others", braced corporate names)4// and the inverse serialisation. Also the conventions the engine uses to fit name suffixes into `PersonName`.5//6// Author: Simon-Pierre Boucher7// Contact: contact@spboucher.ai8import Foundation9import DocModel1011extension PersonName {12 /// Marker produced by BibTeX "and others": the list is known to be incomplete. Formatters render it as "et al.".13 public static let others = PersonName(family: "", given: "", literal: "others")14 public var isOthers: Bool { family.isEmpty && given.isEmpty && literal == "others" }15 /// True for an institutional name kept verbatim.16 public var isLiteral: Bool { !isOthers && !(literal ?? "").isEmpty }1718 /// `PersonName` has no suffix slot: a generational suffix is stored after the given name, "Martin Luther, Jr.".19 public var givenAndSuffix: (given: String, suffix: String?) {20 guard let r = given.range(of: ", ", options: .backwards) else { return (given, nil) }21 let suffix = String(given[r.upperBound...])22 return suffix.isEmpty ? (given, nil) : (String(given[..<r.lowerBound]), suffix)23 }24}2526public enum NameParser {27 /// Parses a BibTeX name list. `decode` converts one raw name part to display text (LaTeX decoding for .bib input).28 public static func parseList(_ raw: String, decode: (String) -> String = NameParser.stripBraces) -> [PersonName] {29 split(raw).compactMap { parseOne($0, decode: decode) }30 }3132 /// Serialises names in BibTeX syntax. `encode` escapes one name part (LaTeX encoding for .bib output).33 public static func string(_ names: [PersonName], encode: (String) -> String = { $0 }) -> String {34 names.map { n -> String in35 if n.isOthers { return "others" }36 if let l = n.literal, !l.isEmpty { return "{" + encode(l) + "}" }37 var family = encode(n.family)38 if n.family.contains(",") || n.family.range(of: " and ", options: .caseInsensitive) != nil { family = "{" + family + "}" }39 let (given, suffix) = n.givenAndSuffix40 if let suffix { return family + ", " + encode(suffix) + ", " + encode(given) }41 if given.isEmpty { return n.family.contains(" ") ? family + "," : family }42 return family + ", " + encode(given)43 }.joined(separator: " and ")44 }4546 public static func stripBraces(_ s: String) -> String {47 s.contains("{") || s.contains("}") ? s.replacingOccurrences(of: "{", with: "").replacingOccurrences(of: "}", with: "") : s48 }4950 /// Splits on the word "and" at brace depth 0.51 static func split(_ raw: String) -> [String] {52 let c = Array(raw)53 var parts: [String] = []54 var depth = 0, start = 0, i = 055 func isSpace(_ ch: Character) -> Bool { ch == " " || ch == "\n" || ch == "\t" || ch == "\r" || ch == "\r\n" || ch == "~" }56 while i < c.count {57 let ch = c[i]58 if ch == "{" { depth += 1 } else if ch == "}" { depth = max(0, depth - 1) }59 else if depth == 0, isSpace(ch), i + 4 < c.count, (c[i + 1] == "a" || c[i + 1] == "A"), (c[i + 2] == "n" || c[i + 2] == "N"),60 (c[i + 3] == "d" || c[i + 3] == "D"), isSpace(c[i + 4]) {61 parts.append(String(c[start..<i]))62 i += 463 start = i + 164 }65 i += 166 }67 if start <= c.count { parts.append(String(c[min(start, c.count)...])) }68 return parts.map { $0.trimmingCharacters(in: .whitespacesAndNewlines) }.filter { !$0.isEmpty }69 }7071 private static func splitTopLevel(_ raw: String, isSeparator: (Character) -> Bool, keepEmpty: Bool) -> [String] {72 var parts: [String] = []73 var cur = ""74 var depth = 075 for ch in raw {76 if ch == "{" { depth += 1 } else if ch == "}" { depth = max(0, depth - 1) }77 if depth == 0, isSeparator(ch) {78 if keepEmpty || !cur.isEmpty { parts.append(cur) }79 cur = ""80 } else { cur.append(ch) }81 }82 if keepEmpty || !cur.isEmpty { parts.append(cur) }83 return parts84 }8586 private static func isFullyBraced(_ w: String) -> Bool {87 guard w.hasPrefix("{"), w.hasSuffix("}") else { return false }88 var depth = 089 var idx = w.startIndex90 while idx < w.endIndex {91 if w[idx] == "{" { depth += 1 } else if w[idx] == "}" { depth -= 1; if depth == 0, w.index(after: idx) != w.endIndex { return false } }92 idx = w.index(after: idx)93 }94 return true95 }9697 static func parseOne(_ raw: String, decode: (String) -> String) -> PersonName? {98 let trimmed = raw.trimmingCharacters(in: .whitespacesAndNewlines)99 guard !trimmed.isEmpty else { return nil }100 if trimmed.lowercased() == "others" { return .others }101 func clean(_ s: String) -> String {102 LaTeXText.collapse(decode(s).replacingOccurrences(of: "\u{00A0}", with: " "))103 }104 let parts = splitTopLevel(trimmed, isSeparator: { $0 == "," }, keepEmpty: true).map { $0.trimmingCharacters(in: .whitespacesAndNewlines) }105 if parts.count == 1 {106 let words = splitTopLevel(parts[0], isSeparator: { $0 == " " || $0 == "\n" || $0 == "\t" || $0 == "~" || $0 == "\r" || $0 == "\r\n" }, keepEmpty: false)107 if words.count == 1 {108 if isFullyBraced(words[0]), !words[0].hasPrefix("{\\") { return PersonName(family: "", given: "", literal: clean(String(words[0].dropFirst().dropLast()))) }109 return PersonName(family: clean(words[0]))110 }111 func isLower(_ w: String) -> Bool {112 if w.hasPrefix("{") && !w.hasPrefix("{\\") { return false }113 return clean(w).first(where: { $0.isLetter })?.isLowercase ?? false114 }115 let n = words.count116 if let k = (0..<(n - 1)).first(where: { isLower(words[$0]) }) {117 // Everything from the first lower-case word on is "von Last"; the particle stays with the family name.118 return PersonName(family: clean(words[k...].joined(separator: " ")), given: clean(words[..<k].joined(separator: " ")))119 }120 return PersonName(family: clean(words[n - 1]), given: clean(words[..<(n - 1)].joined(separator: " ")))121 }122 let family = clean(parts[0])123 if parts.count == 2 { return PersonName(family: family, given: clean(parts[1])) }124 let suffix = clean(parts[1])125 let given = clean(parts[2...].joined(separator: ", "))126 return PersonName(family: family, given: [given, suffix].filter { !$0.isEmpty }.joined(separator: ", "))127 }128}129