gtx-cli 2.1.20 → 2.1.21

package/CHANGELOG.md

@@ -1,5 +1,11 @@
 # gtx-cli
 
+## 2.1.21
+
+### Patch Changes
+
+- [#645](https://github.com/generaltranslation/gt/pull/645) [`58cfaee`](https://github.com/generaltranslation/gt/commit/58cfaee5cc1dcd187f0b72b2761f96c19b4f313e) Thanks [@fernando-aviles](https://github.com/fernando-aviles)! - Escaping HTML to avoid parsing issues from MDX consumers
+
 ## 2.1.20
 
 ### Patch Changes

package/dist/utils/addExplicitAnchorIds.js

@@ -5,7 +5,7 @@ import remarkFrontmatter from 'remark-frontmatter';
 import remarkStringify from 'remark-stringify';
 import { visit } from 'unist-util-visit';
 import { logWarning } from '../console/logging.js';
-import { encodeAnglePlaceholders } from './encodePlaceholders.js';
+import { escapeHtmlInTextNodes } from './escapeHtml.js';
 /**
  * Generates a slug from heading text
  */
@@ -176,7 +176,7 @@ function applyInlineIds(translatedContent, idMappings) {
         const stringifyProcessor = unified()
             .use(remarkFrontmatter, ['yaml', 'toml'])
             .use(remarkMdx)
-            .use(encodeAnglePlaceholders)
+            .use(escapeHtmlInTextNodes)
             .use(remarkStringify, {
             handlers: {
                 // Custom handler to prevent escaping of {#id} syntax

package/dist/utils/escapeHtml.d.ts

@@ -0,0 +1,8 @@
+import type { Plugin } from 'unified';
+import type { Root } from 'mdast';
+/**
+ * Escape HTML-sensitive characters (`&`, `<`, `>`, `"`, `'`) in text nodes,
+ * leaving code, math, MDX expressions, and front-matter untouched.
+ * Ensures literals render safely without altering already-escaped entities.
+ */
+export declare const escapeHtmlInTextNodes: Plugin<[], Root>;

package/dist/utils/escapeHtml.js

@@ -0,0 +1,32 @@
+import { findAndReplace } from 'mdast-util-find-and-replace';
+const IGNORE_PARENTS = [
+    'code',
+    'inlineCode',
+    'mdxFlowExpression',
+    'mdxTextExpression',
+    'mdxjsEsm',
+    'heading',
+    'yaml',
+    'toml',
+    'math',
+    'inlineMath',
+];
+// & that is NOT already an entity: &word;  {  ᨫ
+const AMP_NOT_ENTITY = /&(?![a-zA-Z][a-zA-Z0-9]*;|#\d+;|#x[0-9A-Fa-f]+;)/g;
+/**
+ * Escape HTML-sensitive characters (`&`, `<`, `>`, `"`, `'`) in text nodes,
+ * leaving code, math, MDX expressions, and front-matter untouched.
+ * Ensures literals render safely without altering already-escaped entities.
+ */
+export const escapeHtmlInTextNodes = function () {
+    return (tree) => {
+        findAndReplace(tree, [
+            // Order matters: & first (idempotency), then the rest
+            [AMP_NOT_ENTITY, '&amp;'],
+            [/</g, '&lt;'],
+            [/>/g, '&gt;'],
+            [/"/g, '&quot;'],
+            [/'/g, '&#39;'],
+        ], { ignore: IGNORE_PARENTS });
+    };
+};

package/dist/utils/localizeStaticUrls.js

@@ -7,7 +7,7 @@ import remarkMdx from 'remark-mdx';
 import remarkFrontmatter from 'remark-frontmatter';
 import remarkStringify from 'remark-stringify';
 import { visit } from 'unist-util-visit';
-import { encodeAnglePlaceholders } from './encodePlaceholders.js';
+import { escapeHtmlInTextNodes } from './escapeHtml.js';
 const { isMatch } = micromatch;
 /**
  * Localizes static urls in content files.
@@ -373,7 +373,7 @@ function transformMdxUrls(mdxContent, defaultLocale, targetLocale, hideDefaultLo
         const stringifyProcessor = unified()
             .use(remarkFrontmatter, ['yaml', 'toml'])
             .use(remarkMdx)
-            .use(encodeAnglePlaceholders)
+            .use(escapeHtmlInTextNodes)
             .use(remarkStringify, {
             handlers: {
                 // Handler to prevent escaping (avoids '<' -> '\<')

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "gtx-cli",
-  "version": "2.1.20",
+  "version": "2.1.21",
   "main": "dist/index.js",
   "bin": "dist/main.js",
   "files": [

package/dist/utils/encodePlaceholders.d.ts

@@ -1,11 +0,0 @@
-import type { Plugin } from 'unified';
-import type { Root } from 'mdast';
-/**
- * Re-encode angle-bracket placeholders like <accountName> -> &lt;accountName&gt;
- * in plain text, to prevent MDX JSX parsing on re-parse.
- *
- * - Only touches Text nodes.
- * - Idempotent: already-encoded entities are left unchanged.
- * - Does NOT affect actual MDX/JSX elements because those are not text nodes.
- */
-export declare const encodeAnglePlaceholders: Plugin<[], Root>;

package/dist/utils/encodePlaceholders.js

@@ -1,30 +0,0 @@
-import { findAndReplace } from 'mdast-util-find-and-replace';
-const IGNORE_PARENTS = [
-    'code',
-    'inlineCode',
-    'mdxFlowExpression',
-    'mdxTextExpression',
-    'mdxjsEsm',
-    'yaml',
-    'toml',
-    'math',
-    'inlineMath',
-];
-/**
- * Re-encode angle-bracket placeholders like <accountName> -> &lt;accountName&gt;
- * in plain text, to prevent MDX JSX parsing on re-parse.
- *
- * - Only touches Text nodes.
- * - Idempotent: already-encoded entities are left unchanged.
- * - Does NOT affect actual MDX/JSX elements because those are not text nodes.
- */
-export const encodeAnglePlaceholders = function () {
-    return (tree) => {
-        findAndReplace(tree, [
-            [
-                /<([A-Za-z][\w.-]*)>/g, // <content>, <con-tent>, <con.tent>, <con_tent>
-                (_match, name) => `&lt;${name}&gt;`,
-            ],
-        ], { ignore: IGNORE_PARENTS });
-    };
-};