mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2025-01-22 01:30:49 +08:00
Code Issues Actions 11 Packages Projects Releases Wiki Activity

Fix serialization for vector of RecodedCharID

Browse Source 
Signed-off-by: Stefan Weil <sw@weilnetz.de>
This commit is contained in:
Stefan Weil 2021-03-05 22:47:06 +01:00
parent 33868a52ae
commit 6cfe604d58
2 changed files with 42 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
src/ccutil/serialis.h
View File

@ -28,8 +28,6 @@
namespace tesseract {
class TBOX;
/***********************************************************************
QUOTE_IT MACRO DEFINITION
===========================
@ -90,7 +88,6 @@ class TESS_API TFile {
bool DeSerializeSize(int32_t* data);
bool DeSerialize(std::string& data);
bool DeSerialize(std::vector<char>& data);
bool DeSerialize(TBOX& data);
template <typename T> bool DeSerialize(std::vector<T>& data);
template <typename T>
bool DeSerialize(T *data, size_t count = 1) {
@ -106,29 +103,6 @@ class TESS_API TFile {
bool Serialize(const T *data, size_t count = 1) {
return FWrite(data, sizeof(T), count) == static_cast<int>(count);
}
template <typename T>
bool SerializeClasses(const std::vector<T> &data) {
int32_t sz = data.size();
if (FWrite(&sz, sizeof(sz), 1) != 1)
return false;
for (auto &d : data) {
if (!d.Serialize(this))
return false;
}
return true;
}
template <typename T>
bool DeSerializeClasses(std::vector<T> &data) {
int32_t sz = data.size();
if (FRead(&sz, sizeof(sz), 1) != 1)
return false;
data.resize(sz);
for (auto &d : data) {
if (!d.DeSerialize(this))
return false;
}
return true;
}
// Skip data.
bool Skip(size_t count);

45
src/ccutil/unicharcompress.cpp
View File

@ -83,6 +83,46 @@ static bool DecodeRadicalTable(STRING* radical_data, RSMap* radical_map) {
return true;
}
template <>
bool TFile::DeSerialize(std::vector<RecodedCharID>& data) {
uint32_t size;
if (!DeSerialize(&size)) {
return false;
}
// Arbitrarily limit the number of elements to protect against bad data.
const uint32_t limit = 50000000;
assert(size <= limit);
if (size > limit) {
return false;
}
if (size == 0) {
data.clear();
} else {
// TODO: optimize.
data.resize(size);
for (auto& item : data) {
if (!item.DeSerialize(this)) {
return false;
}
}
}
return true;
}
template <>
bool TFile::Serialize(const std::vector<RecodedCharID>& data) {
uint32_t size = data.size();
if (!Serialize(&size)) {
return false;
}
for (auto& item : data) {
if (!item.Serialize(this)) {
return false;
}
}
return true;
}
UnicharCompress::UnicharCompress() : code_range_(0) {}
UnicharCompress::UnicharCompress(const UnicharCompress& src) { *this = src; }
UnicharCompress::~UnicharCompress() { Cleanup(); }
@ -298,13 +338,12 @@ int UnicharCompress::DecodeUnichar(const RecodedCharID& code) const {
// Writes to the given file. Returns false in case of error.
bool UnicharCompress::Serialize(TFile* fp) const {
return fp->SerializeClasses(encoder_);
return fp->Serialize(encoder_);
}
// Reads from the given file. Returns false in case of error.
bool UnicharCompress::DeSerialize(TFile* fp) {
if (!fp->DeSerializeClasses(encoder_))
return false;
if (!fp->DeSerialize(encoder_)) return false;
ComputeCodeRange();
SetupDecoder();
return true;