Add support for hi-IN Orthography Information (#2945)

* Add support for hi-IN OrthographyInformation

* Default to english parsing

* Move MAUI to .Net 9 to resolve build error

* Remove bom

* Add trace warning macro when not using SDK macros

Author: rhurey

samples/cpp/intent-recognition/samples/intent_recognizer/integer_entity.cpp

@@ -86,10 +86,10 @@ void CSpxIntegerEntity::Init(const std::string& name, const OrthographyInformati
     else
     {
         // fall back to default and use English
-        SPX_TRACE_ERROR(
-            "No explicit integer parser for '%s' language. Disabling integer parsing",
+        SPX_TRACE_WARNING(
+            "No explicit integer parser for '%s' language. Fallig back to English",
             orthography.Name.c_str());
-        m_integerParser = std::static_pointer_cast<ISpxIntegerParser>(std::make_shared<NoOpIntegerParser>());
+        m_integerParser = std::static_pointer_cast<ISpxIntegerParser>(std::make_shared<CSpxENIntegerParser>());
     }
 }
 

samples/cpp/intent-recognition/samples/intent_recognizer/locale_information.cpp

@@ -17,7 +17,7 @@ namespace Intent {
 namespace Impl {
 namespace Locales {
 
-    static const std::array<const OrthographyInformation, 6> ORTHOGRAPHY_INFORMATION =
+    static const std::array<const OrthographyInformation, 7> ORTHOGRAPHY_INFORMATION =
     {
         // English (the default one to use) should be first
         OrthographyInformation{
@@ -73,6 +73,15 @@ namespace Locales {
             "！。？?", // SentenceEndCharacters
             { }, // WordBoundary
             false // RightToLeft
+        },
+        OrthographyInformation{
+            "hi",
+            "\t\r\n ", // Whitespace
+            "[]{}()!?।॥.,:;'\"@#$₹%&*+-=<>/\\^_=`~‑—‒⁻₋−➖﹣－…''""§@*/\\&#†‡′″£₤￡﹩＄€؉‰٪﹪％ʼ՚᾽᾿'＇∶︓﹕：․।︒﹒．｡¥￥₩￦₨،٫⹁、︐︑﹐﹑،､⁺₊➕﬩﹢＋", // InputPunctuation
+            "!?।॥.,:;'\"@#$₹%&*+-=<>/\\^_=`~‑—‒⁻₋−➖﹣－…''""§@*/\\&#†‡′″£₤￡﹩＄€؉‰٪﹪％ʼ՚᾽᾿'＇∶︓﹕：․।︒﹒．｡¥￥₩￦₨،٫⹁、︐︑﹐﹑،､⁺₊➕﬩﹢＋", // PatternPunctuation
+            "।!.?", // SentenceEndCharacters
+            { ' ' }, // WordBoundary
+            false // RightToLeft
         }
     };
 

samples/cpp/intent-recognition/samples/intent_recognizer/stdafx.h

@@ -32,6 +32,11 @@
 #endif
 #define SPX_TRACE_ERROR(...)
 
+#ifdef SPX_TRACE_WARNING
+#undef SPX_TRACE_WARNING
+#endif
+#define SPX_TRACE_WARNING(...)
+
 #ifdef UNUSED
 #undef UNUSED
 #endif

samples/cpp/intent-recognition/samples/samples.cpp

@@ -2,6 +2,7 @@
 // Copyright (c) Microsoft. All rights reserved.
 // Licensed under the MIT license. See https://aka.ms/csspeech/license for the full license information.
 //
+#include "intent_recognizer/stdafx.h"
 
 #include <speechapi_cxx.h> // from Speech SDK
 #include <intentapi_cxx.h> // from this project
@@ -5603,3 +5604,23 @@ TEST_CASE("IntentRecognizer::PatternMatching::ZH Prebuilt integer entities", "[z
     intentResult = intentRecognizer->RecognizeOnceAsync(" ").get();
     RequireIntentId(intentResult, "");
 }
+
+
+TEST_CASE("IntentRecognizer::PatternMatching::IntentList for hi", "[hi]")
+{
+    SECTION("Basic Hindi intents")
+    {
+        auto phrase = std::string("पिछले सात कैरेक्टर को बोल्ड करें।");
+        auto intent = std::string("पिछले सात कैरेक्टर को बोल्ड करें");
+
+        auto intentRecognizer = IntentRecognizer::FromLanguage("hi-IN");
+        intentRecognizer->AddIntent(intent, "intent");
+
+        auto intentResult = intentRecognizer->RecognizeOnceAsync(phrase).get();
+
+        // Log the results
+        SPX_TRACE_INFO("Intent Id: %s", intentResult->IntentId.c_str());
+        SPX_TRACE_INFO("Recognized Text: %s", intentResult->GetDetailedResult().c_str());
+        RequireIntentId(intentResult, "intent");
+    }
+}

samples/cpp/intent-recognition/samples/test_utils.cpp

@@ -5,6 +5,9 @@
 
 #include <ajv.h>
 #include <intentapi_cxx.h>
+#include <string>
+#include <sstream>
+#include <iomanip>
 
 #include "catch2/catch_amalgamated.hpp"
 #include "test_utils.h"
@@ -79,3 +82,84 @@ void RequireAlternateCount(std::shared_ptr<IntentRecognitionResult> result, int
     std::cout << "*** Intent Count: " << actualCount << "\n";
 #endif
 }
+
+
+std::string stringToHex(const std::string& str) {
+    std::ostringstream oss;
+    oss << std::hex << std::setfill('0');
+
+    for (unsigned char c : str) {
+        oss << std::setw(2) << static_cast<int>(c) << " ";
+    }
+
+    return oss.str();
+}
+
+std::string dumpStringToUTF8(const std::string& str, bool escapeAscii) {
+    std::ostringstream oss;
+
+    for (size_t i = 0; i < str.size(); ) {
+        unsigned char c = str[i];
+        uint32_t codepoint = 0;
+        int numBytes = 0;
+
+        // Determine number of bytes in this UTF-8 character
+        if ((c & 0x80) == 0x00) {
+            // 1-byte character (ASCII)
+            codepoint = c;
+            numBytes = 1;
+        }
+        else if ((c & 0xE0) == 0xC0) {
+            // 2-byte character
+            codepoint = c & 0x1F;
+            numBytes = 2;
+        }
+        else if ((c & 0xF0) == 0xE0) {
+            // 3-byte character
+            codepoint = c & 0x0F;
+            numBytes = 3;
+        }
+        else if ((c & 0xF8) == 0xF0) {
+            // 4-byte character
+            codepoint = c & 0x07;
+            numBytes = 4;
+        }
+        else {
+            // Invalid UTF-8, skip this byte
+            oss << "\\x" << std::hex << std::setw(2) << std::setfill('0')
+                << static_cast<int>(c);
+            i++;
+            continue;
+        }
+
+        // Read continuation bytes
+        for (int j = 1; j < numBytes && (i + j) < str.size(); j++) {
+            unsigned char cont = str[i + j];
+            if ((cont & 0xC0) != 0x80) {
+                // Invalid continuation byte
+                break;
+            }
+            codepoint = (codepoint << 6) | (cont & 0x3F);
+        }
+
+        // Output the character
+        if (numBytes == 1 && !escapeAscii && std::isprint(c)) {
+            // Keep printable ASCII as-is
+            oss << static_cast<char>(c);
+        }
+        else if (codepoint <= 0xFFFF) {
+            // Use \uHHHH for BMP characters
+            oss << "\\u" << std::hex << std::setw(4) << std::setfill('0')
+                << codepoint;
+        }
+        else {
+            // Use \UHHHHHHHH for characters outside BMP
+            oss << "\\U" << std::hex << std::setw(8) << std::setfill('0')
+                << codepoint;
+        }
+
+        i += numBytes;
+    }
+
+    return oss.str();
+}

samples/cpp/intent-recognition/samples/test_utils.h

@@ -13,6 +13,8 @@ void RequireEntity(std::shared_ptr<IntentRecognitionResult> result, std::string
 void RequireNoEntity(std::shared_ptr<IntentRecognitionResult> result, std::string expectedEntityId);
 void RequireAlternateIntentId(std::shared_ptr<IntentRecognitionResult> result, std::string expectedIntentId);
 void RequireAlternateCount(std::shared_ptr<IntentRecognitionResult> result, int expectedCount);
+std::string stringToHex(const std::string& str);
+std::string dumpStringToUTF8(const std::string& str, bool escapeAscii = false);
 
 inline bool exists(const std::string& name) {
     return std::ifstream(name.c_str()).good();

samples/csharp/maui/speech-to-text/speech-to-text/speech-to-text.csproj

@@ -23,7 +23,7 @@
         <ApplicationDisplayVersion>1.0</ApplicationDisplayVersion>
         <ApplicationVersion>1</ApplicationVersion>
 
-        <SupportedOSPlatformVersion Condition="$([MSBuild]::GetTargetPlatformIdentifier('$(TargetFramework)')) == 'ios'">11.0</SupportedOSPlatformVersion>
+        <SupportedOSPlatformVersion Condition="$([MSBuild]::GetTargetPlatformIdentifier('$(TargetFramework)')) == 'ios'">15.0</SupportedOSPlatformVersion>
         <SupportedOSPlatformVersion Condition="$([MSBuild]::GetTargetPlatformIdentifier('$(TargetFramework)')) == 'maccatalyst'">13.1</SupportedOSPlatformVersion>
         <SupportedOSPlatformVersion Condition="$([MSBuild]::GetTargetPlatformIdentifier('$(TargetFramework)')) == 'android'">21.0</SupportedOSPlatformVersion>
         <SupportedOSPlatformVersion Condition="$([MSBuild]::GetTargetPlatformIdentifier('$(TargetFramework)')) == 'windows'">10.0.19041.0</SupportedOSPlatformVersion>