diff options
author | Devtools Arcadia <arcadia-devtools@yandex-team.ru> | 2022-02-07 18:08:42 +0300 |
---|---|---|
committer | Devtools Arcadia <arcadia-devtools@mous.vla.yp-c.yandex.net> | 2022-02-07 18:08:42 +0300 |
commit | 1110808a9d39d4b808aef724c861a2e1a38d2a69 (patch) | |
tree | e26c9fed0de5d9873cce7e00bc214573dc2195b7 /library/cpp/codecs/static/builder.cpp | |
download | ydb-1110808a9d39d4b808aef724c861a2e1a38d2a69.tar.gz |
intermediate changes
ref:cde9a383711a11544ce7e107a78147fb96cc4029
Diffstat (limited to 'library/cpp/codecs/static/builder.cpp')
-rw-r--r-- | library/cpp/codecs/static/builder.cpp | 39 |
1 files changed, 39 insertions, 0 deletions
diff --git a/library/cpp/codecs/static/builder.cpp b/library/cpp/codecs/static/builder.cpp new file mode 100644 index 0000000000..93e34a3edb --- /dev/null +++ b/library/cpp/codecs/static/builder.cpp @@ -0,0 +1,39 @@ +#include "builder.h" +#include "common.h" + +#include <library/cpp/codecs/static/static_codec_info.pb.h> + +#include <library/cpp/codecs/codecs.h> + +#include <util/generic/yexception.h> +#include <util/string/subst.h> + +namespace NCodecs { + TStaticCodecInfo BuildStaticCodec(const TVector<TString>& trainingData, const TCodecBuildInfo& info) { + TStaticCodecInfo result; + TCodecPtr codec = ICodec::GetInstance(info.CodecName); + Y_ENSURE_EX(codec, TCodecException() << "empty codec is not allowed"); + + codec->LearnX(trainingData.begin(), trainingData.end(), info.SampleSizeMultiplier); + { + TStringOutput sout{*result.MutableStoredCodec()}; + ICodec::Store(&sout, codec); + } + + auto& debugInfo = *result.MutableDebugInfo(); + debugInfo.SetStoredCodecHash(DataSignature(result.GetStoredCodec())); + debugInfo.SetCodecName(info.CodecName); + debugInfo.SetSampleSizeMultiplier(info.SampleSizeMultiplier); + debugInfo.SetTimestamp(info.Timestamp); + debugInfo.SetRevisionInfo(info.RevisionInfo); + debugInfo.SetTrainingSetComment(info.TrainingSetComment); + debugInfo.SetTrainingSetResId(info.TrainingSetResId); + return result; + } + + TString GetStandardFileName(const TStaticCodecInfo& info) { + TString cName = info.GetDebugInfo().GetCodecName(); + SubstGlobal(cName, ':', '.'); + return TStringBuilder() << cName << "." << info.GetDebugInfo().GetTimestamp() << ".codec_info"; + } +} |