feat: 抽取 html 标签包裹的单行文本（通常出现于表格中）

aio/tools/translator/extractor.spec.ts

@@ -30,6 +30,21 @@ describe('从对照翻译文件中采集生成字典', () => {
     expect(result).eql([{original: 'a', translation: '一'}]);
   });
 
+  it('处理 html 标签包裹的翻译文本', () => {
+    const result = gatherTranslations(`
+    <p>
+    a
+    </p>
+
+    <p>
+    一
+</p>
+    
+    `);
+    expect(result).eql([{original: 'a', translation: '一'}]);
+
+  });
+
   it('从真实的文件中采集（测试）', function () {
     const fs = require('fs');
     const content = fs.readFileSync(dirs.content + 'guide/forms.md', 'utf-8');

aio/tools/translator/extractor.ts

@@ -2,6 +2,7 @@ import * as globby from 'globby';
 import { DictEntry } from './dict-entry';
 import {
   isNotCnPages,
+  isOnlyTag,
   normalizeLines,
   originalIsNotChinese,
   originalIsNotTag,
@@ -28,7 +29,20 @@ export function gatherTranslations(text: string): DictEntry[] {
     const translation = purifyText(lines[i]);
     if (isTranslation(translation)) {
       const original = purifyText(lines[i - 1]);
-      result.push({original, translation});
+      // 对于包裹在 html tag 中的翻译文本进行特殊处理
+      if (isOnlyTag(original)) {
+        const prevTag = lines[i - 4].trim();
+        const prevEndTag = lines[i - 2].trim();
+        const thisEndTag = lines[i + 1].trim();
+        if (original === prevTag && prevEndTag === thisEndTag) {
+          result.push({
+            original: lines[i - 3],
+            translation: lines[i],
+          });
+        }
+      } else {
+        result.push({original, translation});
+      }
     }
   }
   return result

aio/tools/translator/utils.ts

@@ -9,12 +9,20 @@ export function originalIsNotChinese(entry: DictEntry): boolean {
   return !isTranslation(entry.original);
 }
 
+export function isTagLine(text: string) {
+  return /^\s*<\/?\w+.*/.test(text);
+}
+
 export function originalIsNotTag(entry: DictEntry): boolean {
-  return !/^\s*<div.*/.test(entry.original);
+  return !isTagLine(entry.original);
+}
+
+export function isOnlyTag(text: string) {
+  return /^\s*<\w+>\s*$/.test(text);
 }
 
 export function originalIsOnlyTag(entry: DictEntry): boolean {
-  return !/^\s*<\w+>\s*$/.test(entry.original);
+  return !isOnlyTag(entry.original);
 }
 
 export function isNotImg(entry: DictEntry): boolean {
@@ -40,10 +48,11 @@ export function isHead(line: string): boolean {
 export function normalizeLines(text: string): string {
   // 列表、标题等自带换行含义的markdown
   const blockElementPattern = /(?=\n *(\d+\.|-|\*|#|<) )\n/g;
-  const htmlTagPattern = /\n(\s*<.*?>\s*?)\n/g;
-  return text.replace(blockElementPattern, '\n\n')
-    .replace(htmlTagPattern, '\n\n$1\n\n')
-    .replace(/\n\n+/, '\n\n');
+  const htmlTagPattern = /(\s*<.*?>\s*?)\n/g;
+  text = text.replace(blockElementPattern, '\n\n');
+  text = text.replace(htmlTagPattern, '\n\n$1\n\n');
+  text = text.replace(/\n\n+/, '\n\n');
+  return text;
 }
 
 export function indentOf(line): number {