diff --git a/scripts/test-website.py b/scripts/test-website.py
index 5c1aef876e..df4790702b 100644
--- a/scripts/test-website.py
+++ b/scripts/test-website.py
@@ -15,7 +15,7 @@ from urllib.request import urlopen
import xml.etree.ElementTree as ET
import polib
-from website_i18n import Document, build_translations, checked_catalog, extract_catalogs, languages, reading_pages, read_runtime
+from website_i18n import Document, build_translations, checked_catalog, extract_catalogs, is_language_code, languages, reading_pages, read_runtime
ROOT = Path(__file__).resolve().parent.parent
@@ -252,6 +252,19 @@ class WebsiteMaintenanceChecks(unittest.TestCase):
class WebsiteTranslationChecks(unittest.TestCase):
"""Exercise source edits, translated evidence and language editions at their consumers."""
+ def test_script_subtags_preserve_the_hinglish_edition(self):
+ for code in ("en", "zh-CN", "hi-Latn", "hi-Latn-IN"):
+ with self.subTest(code=code):
+ self.assertTrue(is_language_code(code))
+ for code in ("hi-latn", "hi-LATN", "hi-IN-Latn", "hi/Latn"):
+ with self.subTest(code=code):
+ self.assertFalse(is_language_code(code))
+ hinglish = next(item for item in languages(ROOT) if item["readme"] == "README_hi-Latn.md")
+ self.assertEqual(hinglish["code"], "hi-Latn")
+ catalog = polib.pofile(ROOT / "website/locales/hi-Latn/messages.po")
+ self.assertEqual(catalog.metadata["Language"], "hi-Latn")
+ self.assertFalse((ROOT / "website/locales/hi/messages.po").exists())
+
@contextlib.contextmanager
def translated_project(self):
with tempfile.TemporaryDirectory() as directory:
diff --git a/scripts/verify-website.py b/scripts/verify-website.py
index 10af305dcd..58d9ee9e5c 100644
--- a/scripts/verify-website.py
+++ b/scripts/verify-website.py
@@ -11,6 +11,7 @@ import xml.etree.ElementTree as ET
from zipfile import BadZipFile, ZipFile
from website import NOTES_FIXTURES, SITE_ORIGIN, authored_markdown, page_url, sync_pages
+from website_i18n import is_language_code
LEGACY_ORIGIN = "morluto.github.io"
SITE_TEXT_URL = re.compile(
@@ -179,7 +180,7 @@ def check_search_metadata(site, pages):
if not title:
errors.append(f"{label}: needs exactly one nonempty head title")
first = path.relative_to(site).parts[0]
- expected_language = first if re.fullmatch(r"[a-z]{2}(?:-[A-Z]{2})?", first) else "en"
+ expected_language = first if is_language_code(first) else "en"
if page.translation_fallback:
if page.lang != "en" or not page.noindex():
errors.append(f"{label}: English translation fallback needs lang=en and noindex")
diff --git a/scripts/website_i18n.py b/scripts/website_i18n.py
index 928594df76..082f75a23f 100644
--- a/scripts/website_i18n.py
+++ b/scripts/website_i18n.py
@@ -25,6 +25,14 @@ LITERALS = {"REA", "Reverse Engineer Anything", "REA · Reverse Engineer Anythin
FALLBACK_NOTICE = "Translation is being updated. Read this page in English for now."
+LANGUAGE_CODE = re.compile(r"[a-z]{2}(?:-[A-Z][a-z]{3})?(?:-[A-Z]{2})?")
+
+
+def is_language_code(value):
+ """Recognize canonical language, optional script and region inventory codes."""
+ return LANGUAGE_CODE.fullmatch(value) is not None
+
+
def languages(root):
"""Load the website language inventory and check it against README editions."""
path = root / "website/locales/languages.json"
@@ -41,7 +49,7 @@ def languages(root):
raise ValueError("Each language needs code, name, readme and dir")
if not all(isinstance(value, str) and value for value in item.values()):
raise ValueError("Language fields must be nonempty strings")
- if not re.fullmatch(r"[a-z]{2}(?:-[A-Z]{2})?", item["code"]) or item["code"] in codes:
+ if not is_language_code(item["code"]) or item["code"] in codes:
raise ValueError(f"Invalid or duplicate language: {item['code']}")
if item["dir"] not in ("ltr", "rtl"):
raise ValueError(f"Invalid text direction: {item['code']}")
diff --git a/website/locales/hi/messages.po b/website/locales/hi-Latn/messages.po
similarity index 99%
rename from website/locales/hi/messages.po
rename to website/locales/hi-Latn/messages.po
index 88f0337568..15c7c552f3 100644
--- a/website/locales/hi/messages.po
+++ b/website/locales/hi-Latn/messages.po
@@ -2,7 +2,7 @@
msgid ""
msgstr ""
"Project-Id-Version: REA website\n"
-"Language: hi\n"
+"Language: hi-Latn\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n"
"Content-Transfer-Encoding: 8bit\n"
diff --git a/website/locales/languages.json b/website/locales/languages.json
index 31ab11f981..0fdb902d2b 100644
--- a/website/locales/languages.json
+++ b/website/locales/languages.json
@@ -49,7 +49,7 @@
},
{ "code": "it", "name": "Italiano", "readme": "README_it.md", "dir": "ltr" },
{
- "code": "hi",
+ "code": "hi-Latn",
"name": "Hinglish",
"readme": "README_hi-Latn.md",
"dir": "ltr"
diff --git a/website/public/blog/index.html b/website/public/blog/index.html
index a1a2428b0e..bb3dd9bf65 100644
--- a/website/public/blog/index.html
+++ b/website/public/blog/index.html
@@ -51,7 +51,7 @@
-
+
@@ -119,7 +119,7 @@
FAQ