From 775625948e1dd23edceb5cf0815c41c6a3972060 Mon Sep 17 00:00:00 2001
From: "James D. Forrester" <jforrester@wikimedia.org>
Date: Mon, 3 Aug 2026 19:39:10 -0400
Subject: [PATCH] Add ar-001, rhg-arab and rhg-rohg

Wikifunctions carries these three codes as languages users can write
labels in, and WikiLambda reads label directionality from ULS via
$.uls.data.getDir(). ULS answers 'ltr' for any code it does not know,
with no error and no fallback to the base subtag, so all three currently
render left-to-right when they should render right-to-left.

ar-001 is CLDR's code for Modern Standard Arabic and rhg-rohg is the
Hanifi-script Rohingya that the bare rhg code already denotes, so both
are redirects. rhg-arab needs an entry of its own; its autonym is taken
from the corresponding Wikifunctions object.

Twenty-one further right-to-left codes that Wikifunctions carries are
still missing, and seven of them use scripts absent from rtlscripts
(Avst, Mand, Orkh, Phli, Phnx, Samr and Sogd). Those are not included
here because none of them has an autonym available from a source that
has already been reviewed, and guessing at native-language names for
extinct languages seemed worse than leaving them out.

Assisted-by: Claude Opus 5
---
 data/langdb.yaml        |  6 ++++++
 data/language-data.json | 15 ++++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/data/langdb.yaml b/data/langdb.yaml
index ff10d36..5e4ae5e 100644
--- a/data/langdb.yaml
+++ b/data/langdb.yaml
@@ -47,6 +47,9 @@ languages:
   apc-latn: [Latn, [ME], šāmi]
   apw: [Latn, [AM], "Ndee biyati'"]
   ar: [Arab, [AF, ME], العربية]
+ # ar-001 is CLDR's code for Modern Standard Arabic, which for the purposes of
+ # this database is the same thing as ar:
+  ar-001: [ar]
   arc: [Syrc, [ME], ܐܪܡܝܐ]
   arn: [Latn, [AM], mapudungun]
   aro: [Latn, [AM], Araona]
@@ -675,6 +678,9 @@ languages:
   rej: [Latn, [AS, PA], Jang]
   rgn: [Latn, [EU], Rumagnôl]
   rhg: [Rohg, [AS], 𐴌𐴟𐴇𐴥𐴝𐴚𐴒𐴙𐴝]
+  rhg-arab: [Arab, [AS], رُحَ࣪ڠۡگَ࣪ࢬ عَرࣤبِي لࣦكَ࣪]
+ # The bare code rhg already denotes Hanifi-script Rohingya:
+  rhg-rohg: [rhg]
   rif: [Latn, [AF], Tarifit]
   rki: [Mymr, [AS], ရခိုင်]
   rm: [Latn, [EU], rumantsch]
diff --git a/data/language-data.json b/data/language-data.json
index 03898b5..585dd8a 100644
--- a/data/language-data.json
+++ b/data/language-data.json
@@ -302,6 +302,9 @@
             ],
             "العربية"
         ],
+        "ar-001": [
+            "ar"
+        ],
         "arc": [
             "Syrc",
             [
@@ -4323,6 +4326,16 @@
             ],
             "𐴌𐴟𐴇𐴥𐴝𐴚𐴒𐴙𐴝"
         ],
+        "rhg-arab": [
+            "Arab",
+            [
+                "AS"
+            ],
+            "رُحَ࣪ڠۡگَ࣪ࢬ عَرࣤبِي لࣦكَ࣪"
+        ],
+        "rhg-rohg": [
+            "rhg"
+        ],
         "rif": [
             "Latn",
             [
@@ -7831,5 +7844,5 @@
             "tn"
         ]
     },
-    "version": "2026-06-26T13:18:35+00:00"
+    "version": "2026-08-03T23:38:01+00:00"
 }