Repo created

libs/routing/turns_tts_text_i18n.cpp

@@ -0,0 +1,167 @@
+#include "base/logging.hpp"
+#include "base/string_utils.hpp"
+#include "unicode/uchar.h"
+
+#include <string>
+
+namespace routing::turns::sound
+{
+
+void HungarianBaseWordTransform(std::string & hungarianString)
+{
+  if (hungarianString.length() == 0)
+    return;
+
+  strings::UniString myUniStr = strings::MakeUniString(hungarianString);
+
+  std::pair<char32_t, char32_t> constexpr kToReplace[] = {{U'e', U'é'}, {U'a', U'á'}, {U'ö', U'ő'}, {U'ü', U'ű'}};
+  auto & lastChar = myUniStr.back();
+  for (auto [base, harmonized] : kToReplace)
+  {
+    if (lastChar == base)
+    {
+      lastChar = harmonized;
+      hungarianString = strings::ToUtf8(myUniStr);
+      return;
+    }
+  }
+}
+
+bool EndsInAcronymOrNum(strings::UniString const & myUniStr)
+{
+  if (myUniStr.empty())
+    return false;
+
+  bool allUppercaseNum = true;
+  strings::UniString lowerStr = strings::MakeLowerCase(myUniStr);
+  for (long i = myUniStr.size() - 1; i > 0; i--)
+  {
+    // if we've reached a space, we're done here
+    if (myUniStr[i] == ' ')
+      break;
+    // we've found a char that is already lowercase and not a number,
+    // therefore the string is not exclusively uppercase/numeric
+    else if (myUniStr[i] == lowerStr[i] && !u_isdigit(myUniStr[i]))
+    {
+      allUppercaseNum = false;
+      break;
+    }
+  }
+  return allUppercaseNum;
+}
+
+uint8_t CategorizeHungarianAcronymsAndNumbers(std::string const & hungarianString)
+{
+  if (hungarianString.empty())
+    return 2;
+
+  std::array<std::string_view, 14> constexpr backNames = {
+      "A",  // a
+      "Á",  // á
+      "H",  // há
+      "I",  // i
+      "Í",  // í
+      "K",  // ká
+      "O",  // o
+      "Ó",  // ó
+      "U",  // u
+      "Ű",  // ú
+      "0",  // nulla or zéró
+      "3",  // három
+      "6",  // hat
+      "8",  // nyolc
+  };
+
+  std::array<std::string_view, 31> constexpr frontNames = {
+      // all other letters besides H and K
+      "B", "C", "D", "E", "É", "F", "G", "J", "L", "M", "N", "Ö", "Ő",
+      "P", "Q", "R", "S", "T", "Ú", "Ü", "V", "W", "X", "Y", "Z",
+      "1",  // egy
+      "2",  // kettő
+      "4",  // négy
+      "5",  // öt
+      "7",  // hét
+      "9",  // kilenc
+  };
+
+  std::array<std::string_view, 5> constexpr specialCaseFront = {
+      "10",  // tíz special case front
+      "40",  // negyven front
+      "50",  // ötven front
+      "70",  // hetven front
+      "90",  // kilencven front
+  };
+
+  std::array<std::string_view, 4> constexpr specialCaseBack = {
+      "20",  // húsz back
+      "30",  // harminc back
+      "60",  // hatvan back
+      "80",  // nyolcvan back
+  };
+
+  // "100", // száz back, handled below
+
+  // Compare the end of our string with the views above. The order is chosen
+  // in priority of what would most likely cause an ending vowel sound change.
+  for (auto myCase : specialCaseFront)
+    if (hungarianString.ends_with(myCase))
+      return 1;
+  for (auto myCase : specialCaseBack)
+    if (hungarianString.ends_with(myCase))
+      return 2;
+  if (hungarianString.ends_with("100"))
+    return 2;
+  for (auto myCase : frontNames)
+    if (hungarianString.ends_with(myCase))
+      return 1;
+  for (auto myCase : backNames)
+    if (hungarianString.ends_with(myCase))
+      return 2;
+  if (hungarianString.ends_with(' '))
+    return 2;
+
+  LOG(LWARNING, ("Unable to find Hungarian front/back for", hungarianString));
+  return 2;
+}
+
+uint8_t CategorizeHungarianLastWordVowels(std::string const & hungarianString)
+{
+  if (hungarianString.empty())
+    return 2;
+
+  strings::UniString myUniStr = strings::MakeUniString(hungarianString);
+
+  // scan for acronyms and numbers first (i.e. characters spoken differently than words)
+  // if the last word is an acronym/number like M5, check those instead
+  if (EndsInAcronymOrNum(myUniStr))
+    return CategorizeHungarianAcronymsAndNumbers(hungarianString);
+
+  bool foundIndeterminate = false;
+  strings::MakeLowerCaseInplace(myUniStr);  // this isn't an acronym, so match based on lowercase
+
+  std::u32string_view constexpr front{U"eéöőüű"};
+  std::u32string_view constexpr back{U"aáoóuú"};
+  std::u32string_view constexpr indeterminate{U"ií"};
+
+  // find last vowel in last word
+  for (size_t i = myUniStr.size() - 1; i > 0; i--)
+  {
+    if (front.find(myUniStr[i]) != std::string::npos)
+      return 1;
+    if (back.find(myUniStr[i]) != std::string::npos)
+      return 2;
+    if (indeterminate.find(myUniStr[i]) != std::string::npos)
+      foundIndeterminate = true;
+    // if we've hit a space with only indeterminates, it's back
+    if (myUniStr[i] == U' ' && foundIndeterminate == true)
+      return 2;
+    // if we've hit a space with no vowels at all, check for numbers and acronyms
+    if (myUniStr[i] == U' ' && foundIndeterminate == false)
+      return CategorizeHungarianAcronymsAndNumbers(hungarianString);
+  }
+  // if we got here, are we even reading Hungarian words?
+  LOG(LWARNING, ("Hungarian word not found:", hungarianString));
+  return 2;  // default
+}
+
+}  // namespace routing::turns::sound