Support UTF8 to UTF16 LE

This commit is contained in:
2026-03-31 14:10:21 +11:00
parent 0c24096c80
commit 70be68d0fc
14 changed files with 790 additions and 11 deletions

View File

@@ -1,4 +1,8 @@
#pragma once
#ifdef _WIN32
#define NOMINMAX
#include <windows.h>
#endif
#include "ANSODEngine.h"
#include "ANSYOLOOD.h"
#include "ANSTENSORRTOD.h"
@@ -1078,6 +1082,32 @@ namespace ANSCENTER {
return boundingBox;
}
static std::string DoubleEscapeUnicode(const std::string& utf8Str) {
bool hasNonAscii = false;
for (unsigned char c : utf8Str) {
if (c >= 0x80) { hasNonAscii = true; break; }
}
if (!hasNonAscii) return utf8Str;
std::string result;
result.reserve(utf8Str.size() * 2);
size_t i = 0;
while (i < utf8Str.size()) {
unsigned char c = static_cast<unsigned char>(utf8Str[i]);
if (c < 0x80) { result += utf8Str[i++]; continue; }
uint32_t cp = 0;
if ((c & 0xE0) == 0xC0 && i + 1 < utf8Str.size()) {
cp = ((c & 0x1F) << 6) | (static_cast<unsigned char>(utf8Str[i + 1]) & 0x3F); i += 2;
} else if ((c & 0xF0) == 0xE0 && i + 2 < utf8Str.size()) {
cp = ((c & 0x0F) << 12) | ((static_cast<unsigned char>(utf8Str[i + 1]) & 0x3F) << 6) | (static_cast<unsigned char>(utf8Str[i + 2]) & 0x3F); i += 3;
} else if ((c & 0xF8) == 0xF0 && i + 3 < utf8Str.size()) {
cp = ((c & 0x07) << 18) | ((static_cast<unsigned char>(utf8Str[i + 1]) & 0x3F) << 12) | ((static_cast<unsigned char>(utf8Str[i + 2]) & 0x3F) << 6) | (static_cast<unsigned char>(utf8Str[i + 3]) & 0x3F); i += 4;
} else { i++; continue; }
if (cp <= 0xFFFF) { char buf[8]; snprintf(buf, sizeof(buf), "\\u%04x", cp); result += buf; }
else { cp -= 0x10000; char buf[16]; snprintf(buf, sizeof(buf), "\\u%04x\\u%04x", 0xD800 + (uint16_t)(cp >> 10), 0xDC00 + (uint16_t)(cp & 0x3FF)); result += buf; }
}
return result;
}
std::string ANSCENTER::ANSUtilityHelper::VectorDetectionToJsonString(const std::vector<Object>& dets)
{
if (dets.empty()) {
@@ -1091,14 +1121,14 @@ namespace ANSCENTER {
{"class_id", std::to_string(det.classId)},
//{"track_id", std::to_string(det.trackId)},
{"track_id", "0"},
{"class_name", det.className},
{"class_name", DoubleEscapeUnicode(det.className)},
{"prob", std::to_string(det.confidence)},
{"x", std::to_string(det.box.x)},
{"y", std::to_string(det.box.y)},
{"width", std::to_string(det.box.width)},
{"height", std::to_string(det.box.height)},
{"mask", ""}, // TODO: convert masks to comma separated string
{"extra_info",det.extraInfo},
{"mask", ""},
{"extra_info", det.extraInfo},
{"camera_id", det.cameraId},
{"polygon", PolygonToString(det.polygon)},
{"kps", KeypointsToString(det.kps)}
@@ -2212,6 +2242,144 @@ namespace ANSCENTER {
return polygon;
}
// Unicode conversion utilities for LabVIEW wrapper classes
extern "C" ANSENGINE_API int ANSEngine_ConvertUTF8ToUTF16LE(const char* utf8Str, LStrHandle result) {
try {
if (!utf8Str || !result) return -1;
int len = (int)strlen(utf8Str);
if (len == 0) return 0;
bool hasUnicodeEscapes = false;
bool hasNonAscii = false;
for (int i = 0; i < len; i++) {
if ((unsigned char)utf8Str[i] >= 0x80) hasNonAscii = true;
if (i + 1 < len && utf8Str[i] == '\\' && utf8Str[i + 1] == 'u') hasUnicodeEscapes = true;
}
if (!hasNonAscii && !hasUnicodeEscapes) {
MgErr error = DSSetHandleSize(result, sizeof(int32) + len * sizeof(uChar));
if (error != noErr) return -2;
(*result)->cnt = len;
memcpy((*result)->str, utf8Str, len);
return 1;
}
if (hasUnicodeEscapes) {
std::string utf16le;
utf16le.reserve(len * 2);
for (int i = 0; i < len; ) {
if (i + 5 < len && utf8Str[i] == '\\' && utf8Str[i + 1] == 'u') {
char hex[5] = { utf8Str[i + 2], utf8Str[i + 3], utf8Str[i + 4], utf8Str[i + 5], 0 };
uint16_t cp = (uint16_t)strtoul(hex, nullptr, 16);
utf16le += static_cast<char>(cp & 0xFF);
utf16le += static_cast<char>((cp >> 8) & 0xFF);
i += 6;
} else {
utf16le += utf8Str[i];
utf16le += '\0';
i++;
}
}
int size = (int)utf16le.size();
MgErr error = DSSetHandleSize(result, sizeof(int32) + size * sizeof(uChar));
if (error != noErr) return -2;
(*result)->cnt = size;
memcpy((*result)->str, utf16le.data(), size);
return 1;
}
#ifdef _WIN32
int wideLen = MultiByteToWideChar(CP_UTF8, 0, utf8Str, len, nullptr, 0);
if (wideLen <= 0) return 0;
std::wstring wideStr(wideLen, 0);
MultiByteToWideChar(CP_UTF8, 0, utf8Str, len, &wideStr[0], wideLen);
int size = wideLen * (int)sizeof(wchar_t);
MgErr error = DSSetHandleSize(result, sizeof(int32) + size * sizeof(uChar));
if (error != noErr) return -2;
(*result)->cnt = size;
memcpy((*result)->str, wideStr.data(), size);
return 1;
#else
return 0;
#endif
}
catch (...) { return -1; }
}
extern "C" ANSENGINE_API int ANSEngine_ConvertUTF16LEToUTF8(const unsigned char* utf16leBytes, int byteLen, LStrHandle result) {
try {
if (!utf16leBytes || byteLen <= 0 || !result) return -1;
bool isUtf16le = (byteLen >= 2 && byteLen % 2 == 0);
if (isUtf16le) {
bool isAscii = true;
for (int i = 1; i < byteLen; i += 2) {
if (utf16leBytes[i] != 0x00) { isAscii = false; break; }
}
if (isAscii) {
int asciiLen = byteLen / 2;
MgErr error = DSSetHandleSize(result, sizeof(int32) + asciiLen * sizeof(uChar));
if (error != noErr) return -2;
(*result)->cnt = asciiLen;
for (int i = 0; i < asciiLen; i++) (*result)->str[i] = utf16leBytes[i * 2];
return 1;
}
}
#ifdef _WIN32
int wideLen = byteLen / (int)sizeof(wchar_t);
const wchar_t* wideStr = reinterpret_cast<const wchar_t*>(utf16leBytes);
int utf8Len = WideCharToMultiByte(CP_UTF8, 0, wideStr, wideLen, nullptr, 0, nullptr, nullptr);
if (utf8Len <= 0) return 0;
std::string utf8Str(utf8Len, 0);
WideCharToMultiByte(CP_UTF8, 0, wideStr, wideLen, &utf8Str[0], utf8Len, nullptr, nullptr);
MgErr error = DSSetHandleSize(result, sizeof(int32) + utf8Len * sizeof(uChar));
if (error != noErr) return -2;
(*result)->cnt = utf8Len;
memcpy((*result)->str, utf8Str.data(), utf8Len);
return 1;
#else
return 0;
#endif
}
catch (...) { return -1; }
}
std::string ANSUtilityHelper::DecodeJsonUnicodeToUTF16LE(const std::string& escapedStr) {
std::string result;
result.reserve(escapedStr.size() * 2);
size_t i = 0;
while (i < escapedStr.size()) {
if (i + 5 < escapedStr.size() && escapedStr[i] == '\\' && escapedStr[i + 1] == 'u') {
// Parse \uXXXX
char hex[5] = { escapedStr[i + 2], escapedStr[i + 3], escapedStr[i + 4], escapedStr[i + 5], 0 };
uint16_t codepoint = (uint16_t)strtoul(hex, nullptr, 16);
// UTF-16LE: low byte first, high byte second
result += static_cast<char>(codepoint & 0xFF);
result += static_cast<char>((codepoint >> 8) & 0xFF);
i += 6;
} else {
// ASCII character -> UTF-16LE (2 bytes: char, 0x00)
result += escapedStr[i];
result += '\0';
i++;
}
}
return result;
}
std::string ANSUtilityHelper::ConvertUTF8ToUTF16LE(const std::string& utf8Str) {
#ifdef _WIN32
if (utf8Str.empty()) return "";
// First call: get required buffer size
int wideLen = MultiByteToWideChar(CP_UTF8, 0, utf8Str.c_str(), (int)utf8Str.size(), nullptr, 0);
if (wideLen <= 0) return "";
// Allocate wide string buffer
std::wstring wideStr(wideLen, 0);
MultiByteToWideChar(CP_UTF8, 0, utf8Str.c_str(), (int)utf8Str.size(), &wideStr[0], wideLen);
// Convert wchar_t buffer to raw bytes (UTF-16LE on Windows)
const char* rawBytes = reinterpret_cast<const char*>(wideStr.data());
return std::string(rawBytes, wideLen * sizeof(wchar_t));
#else
// Non-Windows: return UTF-8 as-is (LabVIEW is primarily Windows)
return utf8Str;
#endif
}
float ANSUtilityHelper::calculate_intersection_area(const cv::Rect& box1, const cv::Rect& box2) {
int xx1 = std::max(box1.x, box2.x);
int yy1 = std::max(box1.y, box2.y);