@retorquere/bibtex-parser 6.4.19 → 7.0.2

package/index.d.ts

@@ -205,6 +205,10 @@ export interface ParserOptions {
      * If you pass an empty array, or `false`, no sentence casing will be applied (even when there's no language field).
      */
     sentenceCase?: string[] | boolean;
+    /**
+     * If you have sentence-casing on, you can independently choose whether quoted titles within a title are preserved as-is (true) or also sentence-cased(false)
+     */
+    sentenceCasePreserveQuoted?: boolean;
     /**
      * Some bibtex has titles in sentence case, or all-uppercase. If this is on, and there is a field that would normally have sentence-casing applied in which more words are all-`X`case
      * (where `X` is either lower or upper) than mixed-case, it is assumed that you want them this way, and no sentence-casing will be applied to that field

package/index.js

@@ -1228,6 +1228,12 @@ class Parser {
         }
         return parsed;
     }
+    restore(text, orig, preserve) {
+        for (const { start, end } of preserve) {
+            text = text.substring(0, start) + orig.substring(start, end) + text.substring(end);
+        }
+        return text;
+    }
     convert_entry(node) {
         this.entry = {
             key: node.id,
@@ -1324,10 +1330,14 @@ class Parser {
             }
             else {
                 if (this.field.preserveRanges) {
-                    if (this.options.guessAlreadySentenceCased && Math.max(this.field.words.upper, this.field.words.lower) > (this.field.words.other + Math.min(this.field.words.upper, this.field.words.lower))) {
+                    const allCaps = this.field.text.match(/\s/) && this.field.words.upper && !this.field.words.lower && !this.field.words.other;
+                    const seemsSentenceCased = Math.max(this.field.words.upper, this.field.words.lower) > (this.field.words.other + Math.min(this.field.words.upper, this.field.words.lower));
+                    if (!allCaps && seemsSentenceCased && this.options.guessAlreadySentenceCased) {
                         this.preserve(null, null); // , 'mostly sentence cased already')
                     }
                     else {
+                        if (allCaps)
+                            this.field.preserveRanges = [];
                         const txt = this.field.text.replace(preserveCase.markup, markup => marker.markup.repeat(markup.length));
                         let match;
                         preserveCase.sentenceStart.lastIndex = 0;
@@ -1357,7 +1367,7 @@ class Parser {
     convertToSentenceCase(text) {
         if (!this.field.preserveRanges)
             return text;
-        const sentenceCased = (0, sentence_case_2.restore)((0, sentence_case_2.toSentenceCase)(text), text, this.field.preserveRanges);
+        const sentenceCased = this.restore((0, sentence_case_2.toSentenceCase)(text, this.options.sentenceCasePreserveQuoted || false), 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": "6.4.19",
+  "version": "7.0.2",
   "description": "BibTeX parser",
   "main": "index.js",
   "typings": "./index.d.ts",
@@ -9,7 +9,7 @@
     "prestart": "npm run lint && peggy grammar.peggy",
     "start": "tsc -d",
     "tsc": "tsc -d",
-    "poststart": "echo npm run docs",
+    "poststart": "npm run docs",
     "pretest": "npm start",
     "test": "tap test/tap --no-timeout --no-coverage",
     "snap": "npm test --snapshot",

package/sentence-case.d.ts

@@ -1,8 +1,2 @@
-export type TextRange = {
-    start: number;
-    end: number;
-    description?: string;
-};
-export declare function restore(text: string, orig: string, preserve: TextRange[]): string;
-export declare function toSentenceCase(text: string): string;
+export declare function toSentenceCase(sentence: string, preserveQuoted?: boolean): string;
 //# sourceMappingURL=sentence-case.d.ts.map

package/sentence-case.js

@@ -1,6 +1,6 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.toSentenceCase = exports.restore = void 0;
+exports.toSentenceCase = void 0;
 /* eslint-disable @typescript-eslint/no-unsafe-argument */
 const categories = require("xregexp/tools/output/categories");
 const cat = name => categories.find(c => c.name === name).bmp;
@@ -10,13 +10,6 @@ const Ll = cat('Ll');
 const N = cat('N');
 const No = cat('No');
 const Pc = cat('Pc');
-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;
-}
-exports.restore = restore;
 const re = {
     acronym: new RegExp(`([${Lu}][.])+$`),
     innerCaps: new RegExp(`.[${Lu}]`),
@@ -30,9 +23,11 @@ const re = {
     // private word = XRegExp('^\\p{L}+([-.]\\p{L}+)*') // also match gallium-nitride as one word
     // private and = XRegExp('^\\p{Lu}&\\p{Lu}(?=$|[\\P{L}])') // Q&A
 };
-function lowercase(word) {
+function lowercase(word, allcaps) {
     if (!word)
         return word;
+    if (allcaps)
+        return word.toLowerCase();
     const unmasked = word.replace(/\uFFFD/g, '');
     if (unmasked.match(re.skipWords))
         return word.toLowerCase();
@@ -47,43 +42,47 @@ function lowercase(word) {
         return word;
     return word.toLowerCase();
 }
-function toSentenceCase(text) {
+function toSentenceCase(sentence, preserveQuoted = false) {
+    const allcaps = sentence === sentence.toUpperCase();
     const preserve = [];
-    text.replace(/“.*?”/g, (match, i) => {
-        preserve.push({ start: i, end: i + match.length, description: 'quoted' });
-        return '';
-    });
-    text.replace(/‘.*?’/g, (match, i) => {
-        preserve.push({ start: i, end: i + match.length, description: 'quoted' });
-        return '';
-    });
-    text.replace(/(["]).*?\1/g, (match, _quote, i) => {
-        preserve.push({ start: i, end: i + match.length, description: 'quoted' });
-        return '';
-    });
-    text.replace(/([.?!][\s]+)(<[^>]+>)?([A-Z])/g, (match, end, markup, char, i) => {
-        if (!text.substring(0, i + 1).match(re.acronym)) {
-            preserve.push({ start: i + end.length + ((markup === null || markup === void 0 ? void 0 : markup.length) || 0), end: i + end.length + ((markup === null || markup === void 0 ? void 0 : markup.length) || 0) + char.length, description: 'sub-sentence-start' });
+    sentence.replace(/([.?!][\s]+)(<[^>]+>)?([A-Z])/g, (match, end, markup, char, i) => {
+        if (!sentence.substring(0, i + 1).match(re.acronym)) {
+            preserve.push({ pos: i + end.length + ((markup === null || markup === void 0 ? void 0 : markup.length) || 0), text: char, description: 'sub-sentence-start' });
         }
         return '';
     });
-    text.replace(/^(<[^>]+>)?([A-Z])/, (match, markup, char) => {
-        preserve.push({ start: ((markup === null || markup === void 0 ? void 0 : markup.length) || 0), end: ((markup === null || markup === void 0 ? void 0 : markup.length) || 0) + char.length, description: 'sentence-start' });
+    sentence.replace(/^(<[^>]+>)?([A-Z])/, (match, markup, char) => {
+        preserve.push({ pos: ((markup === null || markup === void 0 ? void 0 : markup.length) || 0), text: char, description: 'sentence-start' });
         return '';
     });
-    text.replace(/<span class="nocase">.*?<\/span>|<nc>.*?<\/nc>/gi, (match, i) => {
-        preserve.push({ start: i, end: i + match.length, description: 'nocase' });
+    sentence.replace(/<span class="nocase">.*?<\/span>|<nc>.*?<\/nc>/gi, (text, pos) => {
+        preserve.push({ pos, text, description: 'nocase' });
         return '';
     });
-    let masked = text.replace(/<[^>]+>/g, (match, i) => {
-        preserve.push({ start: i, end: i + match.length, description: 'markup' });
-        return '\uFFFD'.repeat(match.length);
+    let masked = sentence.replace(/<[^>]+>/g, (text, pos) => {
+        preserve.push({ pos, text, description: 'markup' });
+        return '\uFFFD'.repeat(text.length);
     });
+    // last because we're potentially modifying the original
+    for (const q of [/(“)(.*?)”/g, /(‘)(.*?)’/g, /(["])(.*?)\1/g]) {
+        sentence.replace(q, (text, quote, quoted, pos) => {
+            preserve.push({
+                pos: pos + (preserveQuoted ? 0 : quote.length),
+                text: preserveQuoted ? text : toSentenceCase(quoted, preserveQuoted),
+                description: `quoted with ${q}`,
+            });
+            return '';
+        });
+    }
     masked = masked
         .replace(/[;:]\uFFFD*\s+\uFFFD*A\s/g, match => match.toLowerCase())
         .replace(/[–—]\uFFFD*\s*\uFFFD*A\s/g, match => match.toLowerCase())
-        .replace(re.words, word => lowercase(word));
-    return restore(masked, text, preserve);
+        .replace(re.words, word => lowercase(word, allcaps));
+    sentence = masked;
+    for (const { pos, text } of preserve) {
+        sentence = sentence.substring(0, pos) + text + sentence.substring(pos + text.length);
+    }
+    return sentence;
 }
 exports.toSentenceCase = toSentenceCase;
 //# sourceMappingURL=sentence-case.js.map