Fix splitting utf8 string into words (#385)

This commit is contained in:
Fangjun Kuang
2023-10-25 11:49:27 +08:00
committed by GitHub
parent 1249710e1d
commit 6e5efa48c5
3 changed files with 26 additions and 90 deletions

View File

@@ -16,7 +16,7 @@
#include <utility>
#include <vector>
#include "source/utf8.h"
#include "sherpa-onnx/csrc/macros.h"
// This file is copied/modified from
// https://github.com/kaldi-asr/kaldi/blob/master/src/util/text-utils.cc
@@ -163,56 +163,39 @@ template bool SplitStringToFloats(const std::string &full, const char *delim,
std::vector<double> *out);
std::vector<std::string> SplitUtf8(const std::string &text) {
char *begin = const_cast<char *>(text.c_str());
char *end = begin + text.size();
const uint8_t *begin = reinterpret_cast<const uint8_t *>(text.c_str());
const uint8_t *end = begin + text.size();
std::vector<std::string> ans;
std::string buf;
while (begin < end) {
uint32_t code = utf8::next(begin, end);
auto start = begin;
while (start < end) {
uint8_t c = *start;
uint8_t i = 0x80;
int32_t num_bytes = 0;
// 1. is punctuation
if (std::ispunct(code)) {
if (!buf.empty()) {
ans.push_back(std::move(buf));
}
char s[5] = {0};
utf8::append(code, s);
ans.push_back(s);
continue;
// see
// https://en.wikipedia.org/wiki/UTF-8
for (; c & i; i >>= 1) {
++num_bytes;
}
// 2. is space
if (std::isspace(code)) {
if (!buf.empty()) {
ans.push_back(std::move(buf));
}
continue;
if (num_bytes == 0) {
// this is an ascii
ans.emplace_back(reinterpret_cast<const char *>(start), 1);
++start;
} else if (2 <= num_bytes && num_bytes <= 4) {
ans.emplace_back(reinterpret_cast<const char *>(start), num_bytes);
start += num_bytes;
} else {
SHERPA_ONNX_LOGE("Invalid byte at position: %d",
static_cast<int32_t>(start - begin));
// skip this byte
++start;
}
// 3. is alpha
if (std::isalpha(code)) {
buf.push_back(code);
continue;
}
if (!buf.empty()) {
ans.push_back(std::move(buf));
}
// for others
char s[5] = {0};
utf8::append(code, s);
ans.push_back(s);
}
if (!buf.empty()) {
ans.push_back(std::move(buf));
}
return ans;
}
} // namespace sherpa_onnx