Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
#include "result-json.h"
#include <stdexcept>
#include "nlohmann/json.hpp"
namespace funasr {
namespace {
nlohmann::json ParseArrayOrEmpty(const std::string& value) {
if (value.empty()) {
return nlohmann::json::array();
}
auto parsed = nlohmann::json::parse(value, nullptr, false);
if (parsed.is_discarded() || !parsed.is_array()) {
return nlohmann::json::array();
}
return parsed;
}
} // namespace
OutputFormat ParseOutputFormat(const std::string& value) {
if (value == "log") {
return OutputFormat::kLog;
}
if (value == "jsonl") {
return OutputFormat::kJsonl;
}
throw std::invalid_argument("unsupported output format '" + value +
"'; expected 'log' or 'jsonl'");
}
std::string FormatResultJson(const std::string& key, const std::string& mode,
const std::string& text, const std::string& timestamp,
const std::string& stamp_sents) {
nlohmann::json record = {
{"key", key},
{"mode", mode},
{"text", text},
{"timestamp", ParseArrayOrEmpty(timestamp)},
{"stamp_sents", ParseArrayOrEmpty(stamp_sents)},
};
return record.dump();
}
} // namespace funasr