@retorquere/bibtex-parser 6.1.4 → 6.3.2

package/index.d.ts

@@ -1,5 +1,6 @@
 import * as bibtex from './grammar';
 import { JabRefMetadata } from './jabref';
+export declare function toSentenceCase(text: string, ignoreHTML?: boolean): string;
 declare type Node = bibtex.Bibliography | bibtex.Math | bibtex.Environment | bibtex.Block | bibtex.RegularCommand | bibtex.DiacriticCommand | bibtex.Entry | bibtex.Field | bibtex.NumberValue | bibtex.StringDeclaration | bibtex.StringReference | bibtex.SubscriptCommand | bibtex.SuperscriptCommand | bibtex.SymbolCommand | bibtex.TextValue | bibtex.PreambleExpression | bibtex.BracedComment | bibtex.LineComment | bibtex.Markup | bibtex.NonEntryText;
 export interface Name {
     /**

package/index.js

@@ -1,6 +1,6 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.jabref = exports.chunker = exports.promises = exports.ast = exports.parse = exports.fields = void 0;
+exports.jabref = exports.chunker = exports.promises = exports.ast = exports.parse = exports.fields = exports.toSentenceCase = void 0;
 const tslib_1 = require("tslib");
 /* eslint-disable @typescript-eslint/no-unsafe-argument */
 const bibtex = tslib_1.__importStar(require("./grammar"));
@@ -8,6 +8,95 @@ const chunker = tslib_1.__importStar(require("./chunker"));
 const jabref_1 = require("./jabref");
 const unicode2latex_1 = require("unicode2latex");
 const crossref_json_1 = tslib_1.__importDefault(require("./crossref.json"));
+const xregexp_1 = tslib_1.__importDefault(require("xregexp"));
+function restore(text, orig, preserve) {
+    for (const { start, end } of preserve) {
+        text = text.substring(0, start) + orig.substring(start, end) + text.substring(end);
+    }
+    return text;
+}
+class SentenceCaser {
+    constructor() {
+        this.acronym = (0, xregexp_1.default)('^(\\p{Lu}[.])+(?=$|[\\P{L}])');
+        this.quoted = (0, xregexp_1.default)('^"[^"]+"(?=$|[\\P{L}])');
+        this.innerCaps = (0, xregexp_1.default)('\\p{Ll}\\p{Lu}');
+        this.allCaps = (0, xregexp_1.default)('^\\p{Lu}+$');
+        this.aint = (0, xregexp_1.default)("^\\p{L}n't(?=$|[\\P{L}])"); // isn't
+        this.word = (0, xregexp_1.default)('^\\p{L}+(-\\p{L}+)*'); // also match gallium-nitride as one word
+    }
+    convert(text, ignoreHTML = false) {
+        this.input = text;
+        this.result = '';
+        this.sentenceStart = true;
+        const preserve = [];
+        if (ignoreHTML) {
+            let replace = true;
+            while (replace) {
+                replace = false;
+                // always keep the leading char, but skip markup
+                this.input = this.input.replace(/[^<>]<[^>]+>/, (match, i) => {
+                    preserve.push({ start: i + 1, end: i + match.length });
+                    replace = true;
+                    return match[0].repeat(match.length);
+                });
+            }
+            replace = true;
+            while (replace) {
+                replace = false;
+                // always keep the leading char, but skip markup
+                this.input = this.input.replace(/<[^>]+>[^<>]/, (match, i) => {
+                    preserve.push({ start: i, end: i + match.length - 1 });
+                    replace = true;
+                    return match[match.length - 1].repeat(match.length);
+                });
+            }
+        }
+        let m;
+        while (this.input) {
+            if (m = xregexp_1.default.exec(this.input, this.quoted)) { // "Hello There"
+                this.add(m[0], true);
+            }
+            else if (m = xregexp_1.default.exec(this.input, this.acronym)) { // U.S.
+                this.add(m[0], m[0].length > 2);
+            }
+            else if (m = xregexp_1.default.exec(this.input, this.aint)) { // isn't
+                this.add(m[0]);
+            }
+            else if (m = xregexp_1.default.exec(this.input, this.word)) {
+                this.add(m[0]);
+            }
+            else {
+                this.add(this.input[0]);
+            }
+        }
+        return restore(this.result, text, preserve);
+    }
+    add(word, keep = false) {
+        word = word.split('-').map((part, i, parts) => {
+            if ((keep || this.sentenceStart) && i === 0)
+                return part;
+            if (xregexp_1.default.exec(part, this.acronym) && part.length === 2)
+                return part.toLowerCase(); // name initial?
+            if (part === 'A' && parts.length === 1)
+                return part.toLowerCase();
+            if (xregexp_1.default.exec(part, this.innerCaps))
+                return part;
+            if (xregexp_1.default.exec(part, this.allCaps))
+                return part;
+            return part.toLowerCase();
+        }).join('-');
+        this.result += word;
+        this.input = this.input.substr(word.length);
+        if (!word.match(/^\s+$/)) {
+            this.sentenceStart = !!word.match(/^[.?!]$/);
+        }
+    }
+}
+const sentenceCaser = new SentenceCaser;
+function toSentenceCase(text, ignoreHTML = false) {
+    return sentenceCaser.convert(text, ignoreHTML);
+}
+exports.toSentenceCase = toSentenceCase;
 class ParserError extends Error {
     constructor(message, node) {
         super(message); // 'Error' breaks prototype chain here
@@ -353,7 +442,7 @@ class Parser {
           if (start < last.start) throw new Error(JSON.stringify({...last, new: { start, end, reason }, text: this.field.text}))
         }
         */
-        this.field.preserveRanges.push({ start, end, reason });
+        this.field.preserveRanges.push({ start, end, description: reason });
     }
     parseChunk(chunk) {
         this.chunk = chunk.text;
@@ -1276,18 +1365,7 @@ class Parser {
     convertToSentenceCase(text) {
         if (!this.field.preserveRanges)
             return text;
-        // always keep the leading char, but skip markup
-        const lead = text.match(/^(<[^>]+>)*./);
-        if (lead) {
-            this.preserve(lead[0].length - 1, lead[0].length);
-        }
-        else {
-            this.preserve(0, 1);
-        }
-        let sentenceCased = text.toLowerCase().replace(/(([?!]\s*|^)(['"¡¿“‘„«\s]+)?[^\s])/g, x => x.toUpperCase());
-        for (const { start, end } of this.field.preserveRanges) {
-            sentenceCased = sentenceCased.substring(0, start) + text.substring(start, end) + sentenceCased.substring(end);
-        }
+        const sentenceCased = restore(sentenceCaser.convert(text, true), text, this.field.preserveRanges);
         if (text !== sentenceCased)
             this.entry.sentenceCased = true;
         return sentenceCased;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@retorquere/bibtex-parser",
-  "version": "6.1.4",
+  "version": "6.3.2",
   "description": "BibTeX parser",
   "main": "index.js",
   "scripts": {
@@ -32,27 +32,27 @@
   "author": "Emiliano Heyns",
   "license": "ISC",
   "devDependencies": {
-    "@types/node": "^17.0.21",
-    "@typescript-eslint/eslint-plugin": "^5.14.0",
-    "@typescript-eslint/parser": "^5.14.0",
+    "@types/node": "^17.0.25",
+    "@typescript-eslint/eslint-plugin": "^5.20.0",
+    "@typescript-eslint/parser": "^5.20.0",
     "command-line-args": "^5.2.1",
     "csv": "^6.0.5",
-    "digital-tree": "^2.0.3",
+    "digital-tree": "^2.0.4",
     "ejs": "^3.1.6",
-    "eslint": "^8.10.0",
-    "eslint-plugin-import": "^2.25.4",
+    "eslint": "^8.13.0",
+    "eslint-plugin-import": "^2.26.0",
     "eslint-plugin-prefer-arrow": "^1.2.3",
-    "glob": "^7.2.0",
+    "glob": "^8.0.1",
     "mkdirp": "^1.0.4",
-    "node-fetch": "^3.2.2",
+    "node-fetch": "^3.2.3",
     "peggy": "^1.2.0",
     "piscina": "^3.2.0",
     "slug": "^5.3.0",
-    "tap": "^16.0.0",
+    "tap": "^16.0.1",
     "ts-node": "^10.7.0",
     "tslib": "^2.3.1",
-    "typedoc": "^0.22.13",
-    "typescript": "^4.6.2"
+    "typedoc": "^0.22.15",
+    "typescript": "^4.6.3"
   },
   "dependencies": {
     "unicode2latex": "^3.0.0",