From 3f8fd90c31e290c8fdd6fbf5370f03ce5b1feea7 Mon Sep 17 00:00:00 2001
From: Vadim Zeitlin <vadim@wxwidgets.org>
Date: Wed, 1 Sep 2021 01:02:00 +0200
Subject: [PATCH] Support creating Unix wxUILocale for languages without region

Unlike MSW and macOS, creating wxUILocale for e.g. "en" doesn't
necessarily succeed under Linux even if "en_US" is available, so try to
find a working language+region combination if just language doesn't
work.

This requires access to the languages database, so add a private
wxGetLanguageInfos() to avoid having to depend on wxLocale just for
this.
---
 include/wx/private/uilocale.h |  6 ++++
 src/common/intl.cpp           | 12 ++++++--
 src/unix/uilocale.cpp         | 56 +++++++++++++++++++++++++----------
 3 files changed, 57 insertions(+), 17 deletions(-)

diff --git a/include/wx/private/uilocale.h b/include/wx/private/uilocale.h
index abe4f13b14..e2ab82279d 100644
--- a/include/wx/private/uilocale.h
+++ b/include/wx/private/uilocale.h
@@ -12,6 +12,12 @@
 
 #include "wx/localedefs.h"
 #include "wx/string.h"
+#include "wx/vector.h"
+
+typedef wxVector<wxLanguageInfo> wxLanguageInfos;
+
+// Return the vector of all known languages.
+const wxLanguageInfos& wxGetLanguageInfos();
 
 // Function returning hard-coded values for the "C" locale.
 wxString wxGetStdCLocaleInfo(wxLocaleInfo index, wxLocaleCategory cat);
diff --git a/src/common/intl.cpp b/src/common/intl.cpp
index 795110052f..d8f66cd703 100644
--- a/src/common/intl.cpp
+++ b/src/common/intl.cpp
@@ -51,7 +51,8 @@
 #include "wx/stdpaths.h"
 #include "wx/hashset.h"
 #include "wx/uilocale.h"
-#include "wx/vector.h"
+
+#include "wx/private/uilocale.h"
 
 #ifdef __WIN32__
     #include "wx/msw/private/uilocale.h"
@@ -193,9 +194,16 @@ wxString wxLanguageInfo::GetLocaleName() const
 // wxLocale
 // ----------------------------------------------------------------------------
 
-static wxVector<wxLanguageInfo> gs_languagesDB;
+static wxLanguageInfos gs_languagesDB;
 static bool gs_languagesDBInitialized = false;
 
+const wxLanguageInfos& wxGetLanguageInfos()
+{
+    wxLocale::CreateLanguagesDB();
+
+    return gs_languagesDB;
+}
+
 /*static*/ void wxLocale::CreateLanguagesDB()
 {
     if (!gs_languagesDBInitialized)
diff --git a/src/unix/uilocale.cpp b/src/unix/uilocale.cpp
index a470f2a3c6..2334e7010e 100644
--- a/src/unix/uilocale.cpp
+++ b/src/unix/uilocale.cpp
@@ -80,6 +80,22 @@ inline locale_t TryCreateLocale(const wxLocaleIdent& locId)
     return newlocale(LC_ALL_MASK, locId.GetName().mb_str(), NULL);
 }
 
+// Wrapper around newlocale() also trying to append UTF-8 codeset (and
+// modifying its wxLocaleIdent argument if it succeeds).
+locale_t TryCreateLocaleWithUTF8(wxLocaleIdent& locId)
+{
+    locale_t loc = TryCreateLocale(locId);
+    if ( !loc && locId.GetCharset().empty() )
+    {
+        wxLocaleIdent locIdUTF8 = wxLocaleIdent(locId).Charset("UTF-8");
+        loc = TryCreateLocale(locIdUTF8);
+        if ( loc )
+            locId = locIdUTF8;
+    }
+
+    return loc;
+}
+
 #endif // HAVE_LOCALE_T
 
 } // anonymous namespace
@@ -337,7 +353,7 @@ wxUILocaleImpl* wxUILocaleImpl::CreateForLocale(const wxLocaleIdent& locIdOrig)
     // Make a copy of it because it can be modified below.
     wxLocaleIdent locId = locIdOrig;
 
-    locale_t loc = TryCreateLocale(locId);
+    locale_t loc = TryCreateLocaleWithUTF8(locId);
     if ( !loc )
     {
         // Try to find a variant of this locale available on this system: first
@@ -345,28 +361,38 @@ wxUILocaleImpl* wxUILocaleImpl::CreateForLocale(const wxLocaleIdent& locIdOrig)
         // does _not_ work under Linux, so try adding one if we don't have it.
         if ( locId.GetRegion().empty() )
         {
-            const wxLanguageInfo* const info =
-                wxLocale::FindLanguageInfo(locId.GetLanguage());
-            if ( info )
+            const wxString lang = locId.GetLanguage();
+
+            const wxLanguageInfos& infos = wxGetLanguageInfos();
+            for ( wxLanguageInfos::const_iterator it = infos.begin();
+                  it != infos.end();
+                  ++it )
             {
-                wxString region = info->CanonicalName.AfterFirst('_');
-                if ( !region.empty() )
+                const wxString& fullname = it->CanonicalName;
+                if ( fullname.BeforeFirst('_') == lang )
                 {
                     // We never have encoding in our canonical names, but we
                     // can have modifiers, so get rid of them if necessary.
-                    region = region.BeforeFirst('@');
+                    const wxString&
+                        region = fullname.AfterFirst('_').BeforeFirst('@');
+                    if ( !region.empty() )
+                    {
+                        loc = TryCreateLocaleWithUTF8(locId.Region(region));
+                        if ( loc )
+                        {
+                            // We take the first available region, we don't
+                            // have enough data to know how to prioritize them
+                            // (and wouldn't want to start any geopolitical
+                            // disputes).
+                            break;
+                        }
+                    }
 
-                    loc = TryCreateLocale(locId.Region(region));
+                    // Don't bother reverting region to the old value as it will
+                    // be overwritten during the next loop iteration anyhow.
                 }
             }
         }
-
-        // And sometimes the locale without encoding is not available, but one
-        // with UTF-8 encoding is, so try this too.
-        if ( !loc && locId.GetCharset().empty() )
-        {
-            loc = TryCreateLocale(locId.Charset("UTF-8"));
-        }
     }
 
     if ( !loc )