EngineX-Iluvatar/enginex_bi_series-sherpa-onnx
Archived
16
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

Add on-device tex-to-speech (TTS) demo for HarmonyOS (#1590)

Browse Source
This commit is contained in:
Fangjun Kuang
2024-12-04 14:27:12 +08:00
committed by GitHub
parent 47a2dd4cf8
commit 74a8735f7a
61 changed files with 1930 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

47
sherpa-onnx/csrc/lexicon.cc
View File

@@ -7,6 +7,7 @@
#include <algorithm>
#include <cctype>
#include <fstream>
#include <iomanip>
#include <memory>
#include <sstream>
#include <strstream>
@@ -159,17 +160,26 @@ std::vector<TokenIDs> Lexicon::ConvertTextToTokenIdsChinese(
words = ProcessHeteronyms(words);
if (debug_) {
fprintf(stderr, "Input text in string: %s\n", text.c_str());
fprintf(stderr, "Input text in bytes:");
std::ostringstream os;
os << "Input text in string: " << text << "\n";
os << "Input text in bytes:";
for (uint8_t c : text) {
fprintf(stderr, " %02x", c);
os << " 0x" << std::setfill('0') << std::setw(2) << std::right << std::hex
<< c;
}
fprintf(stderr, "\n");
fprintf(stderr, "After splitting to words:");
os << "\n";
os << "After splitting to words:";
for (const auto &w : words) {
fprintf(stderr, " %s", w.c_str());
os << " " << w;
}
fprintf(stderr, "\n");
os << "\n";
#if __OHOS__
SHERPA_ONNX_LOGE("%{public}s", os.str().c_str());
#else
SHERPA_ONNX_LOGE("%s", os.str().c_str());
#endif
}
std::vector<TokenIDs> ans;
@@ -259,17 +269,26 @@ std::vector<TokenIDs> Lexicon::ConvertTextToTokenIdsNotChinese(
std::vector<std::string> words = SplitUtf8(text);
if (debug_) {
fprintf(stderr, "Input text (lowercase) in string: %s\n", text.c_str());
fprintf(stderr, "Input text in bytes:");
std::ostringstream os;
os << "Input text (lowercase) in string: " << text << "\n";
os << "Input text in bytes:";
for (uint8_t c : text) {
fprintf(stderr, " %02x", c);
os << " 0x" << std::setfill('0') << std::setw(2) << std::right << std::hex
<< c;
}
fprintf(stderr, "\n");
fprintf(stderr, "After splitting to words:");
os << "\n";
os << "After splitting to words:";
for (const auto &w : words) {
fprintf(stderr, " %s", w.c_str());
os << " " << w;
}
fprintf(stderr, "\n");
os << "\n";
#if __OHOS__
SHERPA_ONNX_LOGE("%{public}s", os.str().c_str());
#else
SHERPA_ONNX_LOGE("%s", os.str().c_str());
#endif
}
int32_t blank = token2id_.at(" ");