[core] Port is-supported-script to native.

Native port is much simpler because RTL text support is always enabled.

3 files changed, 42 insertions, 2 deletions

diff --git a/src/mbgl/style/expression/compound_expression.cpp b/src/mbgl/style/expression/compound_expression.cpp
index c36ffa33e3..c257dbf2bb 100644
--- a/src/mbgl/style/expression/compound_expression.cpp
+++ b/src/mbgl/style/expression/compound_expression.cpp
@@ -3,6 +3,7 @@
 #include <mbgl/style/expression/util.hpp>
 #include <mbgl/tile/geometry_tile_data.hpp>
 #include <mbgl/math/log2.hpp>
+#include <mbgl/util/i18n.hpp>
 #include <mbgl/util/ignore.hpp>
 #include <mbgl/util/string.hpp>
 #include <mbgl/util/platform.hpp>
@@ -379,6 +380,10 @@ std::unordered_map<std::string, CompoundExpressionRegistry::Definition> initiali
     
     define("!", [](bool e) -> Result<bool> { return !e; });
     
+    define("is-supported-script", [](const std::string& x) -> Result<bool> {
+        return util::i18n::isStringInSupportedScript(x);
+    });
+
     define("upcase", [](const std::string& input) -> Result<std::string> {
         return platform::uppercase(input);
     });
diff --git a/src/mbgl/util/i18n.cpp b/src/mbgl/util/i18n.cpp
index 1fc13bfb7d..5530796915 100644
--- a/src/mbgl/util/i18n.cpp
+++ b/src/mbgl/util/i18n.cpp
@@ -1,4 +1,5 @@
-#include "i18n.hpp"
+#include <mbgl/util/i18n.hpp>
+#include <mbgl/util/utf.hpp>
 
 #include <algorithm>
 #include <map>
@@ -65,7 +66,7 @@ DEFINE_IS_IN_UNICODE_BLOCK(UnifiedCanadianAboriginalSyllabics, 0x1400, 0x167F)
 // DEFINE_IS_IN_UNICODE_BLOCK(Hanunoo, 0x1720, 0x173F)
 // DEFINE_IS_IN_UNICODE_BLOCK(Buhid, 0x1740, 0x175F)
 // DEFINE_IS_IN_UNICODE_BLOCK(Tagbanwa, 0x1760, 0x177F)
-// DEFINE_IS_IN_UNICODE_BLOCK(Khmer, 0x1780, 0x17FF)
+DEFINE_IS_IN_UNICODE_BLOCK(Khmer, 0x1780, 0x17FF)
 // DEFINE_IS_IN_UNICODE_BLOCK(Mongolian, 0x1800, 0x18AF)
 DEFINE_IS_IN_UNICODE_BLOCK(UnifiedCanadianAboriginalSyllabicsExtended, 0x18B0, 0x18FF)
 // DEFINE_IS_IN_UNICODE_BLOCK(Limbu, 0x1900, 0x194F)
@@ -581,6 +582,38 @@ std::u16string verticalizePunctuation(const std::u16string& input) {
 char16_t verticalizePunctuation(char16_t chr) {
     return verticalPunctuation.count(chr) ? verticalPunctuation.at(chr) : 0;
 }
+    
+bool charInSupportedScript(char16_t chr) {
+    // This is a rough heuristic: whether we "can render" a script
+    // actually depends on the properties of the font being used
+    // and whether differences from the ideal rendering are considered
+    // semantically significant.
+    
+    // Even in Latin script, we "can't render" combinations such as the fi
+    // ligature, but we don't consider that semantically significant.n false;
+    if ((chr >= 0x0900 && chr <= 0x0DFF) ||
+        // Main blocks for Indic scripts and Sinhala
+        (chr >= 0x0F00 && chr <= 0x109F) ||
+        // Main blocks for Tibetan and Myanmar
+        isInKhmer(chr)) {
+        // These blocks cover common scripts that require
+        // complex text shaping, based on unicode script metadata:
+        // http://www.unicode.org/repos/cldr/trunk/common/properties/scriptMetadata.txt
+        // where "Web Rank <= 32" "Shaping Required = YES"
+        return false;
+    }
+    return true;
+}
+    
+bool isStringInSupportedScript(const std::string& input) {
+    auto u16string = util::utf8_to_utf16::convert(input);
+    for (char16_t chr : u16string) {
+        if (!charInSupportedScript(chr)) {
+            return false;
+        }
+    }
+    return true;
+}
 
 } // namespace i18n
 } // namespace util
diff --git a/src/mbgl/util/i18n.hpp b/src/mbgl/util/i18n.hpp
index b3960c743c..a74215a134 100644
--- a/src/mbgl/util/i18n.hpp
+++ b/src/mbgl/util/i18n.hpp
@@ -72,6 +72,8 @@ std::u16string verticalizePunctuation(const std::u16string& input);
 
     @return The character’s specialized vertical form; 0 if not applicable. */
 char16_t verticalizePunctuation(char16_t chr);
+    
+bool isStringInSupportedScript(const std::string& input);
 
 } // namespace i18n
 } // namespace util