320 lines
12 KiB
C++
320 lines
12 KiB
C++
#include "pq_app.h"
|
||
#include "tinyxml2.h"
|
||
|
||
#include <windows.h>
|
||
|
||
#include <algorithm>
|
||
#include <cctype>
|
||
#include <cstdlib>
|
||
#include <filesystem>
|
||
#include <fstream>
|
||
#include <sstream>
|
||
#include <stdexcept>
|
||
#include <string>
|
||
#include <unordered_map>
|
||
#include <vector>
|
||
|
||
/*
|
||
* icd_mapper.cpp
|
||
*
|
||
* 功能：把外部 ICD/XML 中 LN/DOI 的 desc 描述补写到工作簿
|
||
* “未匹配的原始数据”工作表的“icd中代表的字段”列。
|
||
*
|
||
* 输入：
|
||
* icdPath 外部 ICD/XML 文件，可能是 UTF-8，也可能声明/实际使用 GBK/GB2312。
|
||
* workbookPath main.cpp 生成的 Excel 2003 XML 格式 .xls。
|
||
*
|
||
* 输出：
|
||
* 直接修改 workbookPath。message 返回解析结果、匹配数量或错误原因。
|
||
*
|
||
* 核心匹配规则：
|
||
* 1. 从 ICD 的每个 LN 节点读取 lnClass + inst，并遍历其 DOI 子节点；
|
||
* 2. 组成 DO key：lnClass + inst + "$MX$" + DOI/@name；
|
||
* 3. 工作簿未匹配页的“原始路径”如果等于该 key，或以 key + "$" 开头，
|
||
* 就认为该原始 MMS 路径属于这个 DOI，把 DOI/@desc 写入目标列。
|
||
*/
|
||
|
||
namespace fs = std::filesystem;
|
||
|
||
namespace {
|
||
|
||
std::string readAllBinary(const fs::path& path) {
|
||
std::ifstream ifs(path, std::ios::binary);
|
||
if (!ifs) throw std::runtime_error("Failed to open file: " + path.string());
|
||
std::ostringstream ss;
|
||
ss << ifs.rdbuf();
|
||
return ss.str();
|
||
}
|
||
|
||
void writeAllBinary(const fs::path& path, const std::string& text) {
|
||
std::ofstream ofs(path, std::ios::binary);
|
||
if (!ofs) throw std::runtime_error("Failed to write file: " + path.string());
|
||
ofs << text;
|
||
}
|
||
|
||
std::string lowerAscii(std::string s) {
|
||
std::transform(s.begin(), s.end(), s.begin(),
|
||
[](unsigned char c) { return static_cast<char>(std::tolower(c)); });
|
||
return s;
|
||
}
|
||
|
||
bool declaresGbEncoding(const std::string& text) {
|
||
std::string head = lowerAscii(text.substr(0, std::min<size_t>(text.size(), 256)));
|
||
return head.find("gb2312") != std::string::npos ||
|
||
head.find("gbk") != std::string::npos ||
|
||
head.find("gb18030") != std::string::npos;
|
||
}
|
||
|
||
bool isLikelyUtf8(const std::string& text) {
|
||
// 只做 UTF-8 字节合法性判断，不判断语义；合法就优先按 UTF-8 给 tinyxml2。
|
||
int remaining = 0;
|
||
for (unsigned char c : text) {
|
||
if (remaining == 0) {
|
||
if ((c & 0x80) == 0) continue;
|
||
if ((c & 0xE0) == 0xC0) remaining = 1;
|
||
else if ((c & 0xF0) == 0xE0) remaining = 2;
|
||
else if ((c & 0xF8) == 0xF0) remaining = 3;
|
||
else return false;
|
||
} else {
|
||
if ((c & 0xC0) != 0x80) return false;
|
||
--remaining;
|
||
}
|
||
}
|
||
return remaining == 0;
|
||
}
|
||
|
||
std::string ansiToUtf8(const std::string& text) {
|
||
// 936 是 Windows 简体中文代码页，兼容常见 GBK/GB2312 ICD 文件。
|
||
if (text.empty()) return "";
|
||
int wideLen = MultiByteToWideChar(936, 0, text.data(), (int)text.size(), nullptr, 0);
|
||
if (wideLen <= 0) return text;
|
||
std::wstring wide(wideLen, L'\0');
|
||
MultiByteToWideChar(936, 0, text.data(), (int)text.size(), wide.data(), wideLen);
|
||
int utf8Len = WideCharToMultiByte(CP_UTF8, 0, wide.data(), wideLen, nullptr, 0, nullptr, nullptr);
|
||
std::string out(utf8Len, '\0');
|
||
WideCharToMultiByte(CP_UTF8, 0, wide.data(), wideLen, out.data(), utf8Len, nullptr, nullptr);
|
||
return out;
|
||
}
|
||
|
||
std::string loadXmlText(const fs::path& path) {
|
||
std::string raw = readAllBinary(path);
|
||
// ICD 声明 GB 编码或字节本身不像 UTF-8 时，先转成 UTF-8 再交给 tinyxml2。
|
||
if (declaresGbEncoding(raw) || !isLikelyUtf8(raw)) return ansiToUtf8(raw);
|
||
return raw;
|
||
}
|
||
|
||
std::string localName(const char* name) {
|
||
if (!name) return "";
|
||
std::string s(name);
|
||
size_t pos = s.find(':');
|
||
return pos == std::string::npos ? s : s.substr(pos + 1);
|
||
}
|
||
|
||
std::string attr(const tinyxml2::XMLElement* e, const char* name) {
|
||
const char* v = e ? e->Attribute(name) : nullptr;
|
||
return v ? v : "";
|
||
}
|
||
|
||
void collectIcdDescriptions(const tinyxml2::XMLElement* element,
|
||
std::unordered_map<std::string, std::string>& descByDo) {
|
||
if (!element) return;
|
||
if (localName(element->Name()) == "LN") {
|
||
std::string lnClass = attr(element, "lnClass");
|
||
std::string inst = attr(element, "inst");
|
||
if (!lnClass.empty()) {
|
||
for (const tinyxml2::XMLElement* child = element->FirstChildElement();
|
||
child;
|
||
child = child->NextSiblingElement()) {
|
||
if (localName(child->Name()) != "DOI") continue;
|
||
std::string doiName = attr(child, "name");
|
||
std::string desc = attr(child, "desc");
|
||
if (doiName.empty() || desc.empty()) continue;
|
||
// 这里构造的是 origin 原始路径前缀。后面路径可继续带 $mag$f 等 DA 后缀。
|
||
std::string key = lnClass + inst + "$MX$" + doiName;
|
||
descByDo[key] = desc;
|
||
}
|
||
}
|
||
}
|
||
for (const tinyxml2::XMLElement* child = element->FirstChildElement();
|
||
child;
|
||
child = child->NextSiblingElement()) {
|
||
collectIcdDescriptions(child, descByDo);
|
||
}
|
||
}
|
||
|
||
std::unordered_map<std::string, std::string> parseIcdDescriptions(const fs::path& icdPath) {
|
||
tinyxml2::XMLDocument doc;
|
||
std::string xml = loadXmlText(icdPath);
|
||
tinyxml2::XMLError rc = doc.Parse(xml.c_str(), xml.size());
|
||
if (rc != tinyxml2::XML_SUCCESS) {
|
||
throw std::runtime_error("Failed to parse ICD XML: " + std::to_string((int)rc));
|
||
}
|
||
std::unordered_map<std::string, std::string> descByDo;
|
||
collectIcdDescriptions(doc.RootElement(), descByDo);
|
||
return descByDo;
|
||
}
|
||
|
||
std::string cellText(const tinyxml2::XMLElement* cell) {
|
||
const tinyxml2::XMLElement* data = cell ? cell->FirstChildElement("Data") : nullptr;
|
||
const char* text = data ? data->GetText() : nullptr;
|
||
return text ? text : "";
|
||
}
|
||
|
||
int cellIndex(const tinyxml2::XMLElement* cell, int fallback) {
|
||
const char* idx = cell ? cell->Attribute("ss:Index") : nullptr;
|
||
return idx && *idx ? std::max(1, std::atoi(idx)) : fallback;
|
||
}
|
||
|
||
tinyxml2::XMLElement* cellAt(tinyxml2::XMLElement* row, int col) {
|
||
int current = 1;
|
||
for (tinyxml2::XMLElement* cell = row ? row->FirstChildElement("Cell") : nullptr;
|
||
cell;
|
||
cell = cell->NextSiblingElement("Cell")) {
|
||
current = cellIndex(cell, current);
|
||
if (current == col) return cell;
|
||
++current;
|
||
}
|
||
return nullptr;
|
||
}
|
||
|
||
tinyxml2::XMLElement* ensureCell(tinyxml2::XMLDocument& doc, tinyxml2::XMLElement* row, int col) {
|
||
if (tinyxml2::XMLElement* existing = cellAt(row, col)) return existing;
|
||
// Excel XML 允许通过 ss:Index 跳列；补列时直接设置 ss:Index，避免改动前面已有单元格。
|
||
tinyxml2::XMLElement* cell = doc.NewElement("Cell");
|
||
cell->SetAttribute("ss:Index", col);
|
||
row->InsertEndChild(cell);
|
||
return cell;
|
||
}
|
||
|
||
void setCellText(tinyxml2::XMLDocument& doc, tinyxml2::XMLElement* cell, const std::string& text) {
|
||
tinyxml2::XMLElement* data = cell->FirstChildElement("Data");
|
||
if (!data) {
|
||
data = doc.NewElement("Data");
|
||
data->SetAttribute("ss:Type", "String");
|
||
cell->InsertEndChild(data);
|
||
}
|
||
data->SetText(text.c_str());
|
||
}
|
||
|
||
std::vector<std::string> sortedKeysByLength(const std::unordered_map<std::string, std::string>& values) {
|
||
std::vector<std::string> keys;
|
||
keys.reserve(values.size());
|
||
for (const auto& kv : values) keys.push_back(kv.first);
|
||
// 长路径优先，避免短 key 先命中长 key 的前缀。
|
||
std::sort(keys.begin(), keys.end(), [](const std::string& a, const std::string& b) {
|
||
if (a.size() != b.size()) return a.size() > b.size();
|
||
return a < b;
|
||
});
|
||
return keys;
|
||
}
|
||
|
||
bool pathMatchesDo(const std::string& rawPath, const std::string& doKey) {
|
||
// rawPath 可以正好是 DO，也可以是 DO$DA...；必须在 "$" 边界匹配，避免误命中相似前缀。
|
||
return rawPath == doKey ||
|
||
(rawPath.size() > doKey.size() &&
|
||
rawPath.compare(0, doKey.size(), doKey) == 0 &&
|
||
rawPath[doKey.size()] == '$');
|
||
}
|
||
|
||
std::string descForPath(const std::string& rawPath,
|
||
const std::unordered_map<std::string, std::string>& descByDo,
|
||
const std::vector<std::string>& orderedKeys) {
|
||
for (const auto& key : orderedKeys) {
|
||
if (pathMatchesDo(rawPath, key)) {
|
||
auto it = descByDo.find(key);
|
||
return it == descByDo.end() ? "" : it->second;
|
||
}
|
||
}
|
||
return "";
|
||
}
|
||
|
||
tinyxml2::XMLElement* findWorksheet(tinyxml2::XMLDocument& doc, const char* sheetName) {
|
||
tinyxml2::XMLElement* workbook = doc.FirstChildElement("Workbook");
|
||
for (tinyxml2::XMLElement* ws = workbook ? workbook->FirstChildElement("Worksheet") : nullptr;
|
||
ws;
|
||
ws = ws->NextSiblingElement("Worksheet")) {
|
||
const char* name = ws->Attribute("ss:Name");
|
||
if (name && std::string(name) == sheetName) return ws;
|
||
}
|
||
return nullptr;
|
||
}
|
||
|
||
int headerColumn(tinyxml2::XMLElement* headerRow, const std::string& title) {
|
||
int current = 1;
|
||
for (tinyxml2::XMLElement* cell = headerRow ? headerRow->FirstChildElement("Cell") : nullptr;
|
||
cell;
|
||
cell = cell->NextSiblingElement("Cell")) {
|
||
current = cellIndex(cell, current);
|
||
if (cellText(cell) == title) return current;
|
||
++current;
|
||
}
|
||
return 0;
|
||
}
|
||
|
||
} // namespace
|
||
|
||
bool pqApplyIcdDescriptionsToWorkbook(const std::string& icdPath,
|
||
const std::string& workbookPath,
|
||
std::string* message) {
|
||
try {
|
||
// 第一步：解析 ICD，得到 “DO 路径前缀 -> 中文描述”。
|
||
auto descByDo = parseIcdDescriptions(fs::u8path(icdPath));
|
||
if (descByDo.empty()) {
|
||
if (message) *message = "ICD 中没有解析到 LN/DOI desc";
|
||
return false;
|
||
}
|
||
std::vector<std::string> orderedKeys = sortedKeysByLength(descByDo);
|
||
|
||
// 第二步：读取 main.cpp 输出的工作簿。该 .xls 实际是 XML，可直接用 tinyxml2 修改。
|
||
tinyxml2::XMLDocument doc;
|
||
std::string workbookXml = readAllBinary(fs::u8path(workbookPath));
|
||
tinyxml2::XMLError rc = doc.Parse(workbookXml.c_str(), workbookXml.size());
|
||
if (rc != tinyxml2::XML_SUCCESS) {
|
||
if (message) *message = "表格文件解析失败: " + std::to_string((int)rc);
|
||
return false;
|
||
}
|
||
|
||
// 第三步：只处理“未匹配的原始数据”页，因为已匹配页已经有明确映射关系。
|
||
tinyxml2::XMLElement* sheet = findWorksheet(doc, "未匹配的原始数据");
|
||
if (!sheet) {
|
||
if (message) *message = "没有找到 sheet: 未匹配的原始数据";
|
||
return false;
|
||
}
|
||
tinyxml2::XMLElement* table = sheet->FirstChildElement("Table");
|
||
tinyxml2::XMLElement* header = table ? table->FirstChildElement("Row") : nullptr;
|
||
int pathCol = headerColumn(header, "原始路径");
|
||
int icdCol = headerColumn(header, "icd中代表的字段");
|
||
if (pathCol <= 0 || icdCol <= 0) {
|
||
if (message) *message = "未匹配 sheet 缺少 原始路径 或 icd中代表的字段 列";
|
||
return false;
|
||
}
|
||
|
||
// 第四步：逐行读取原始路径，按 DO 前缀回填 DOI 描述。
|
||
int matched = 0;
|
||
for (tinyxml2::XMLElement* row = header ? header->NextSiblingElement("Row") : nullptr;
|
||
row;
|
||
row = row->NextSiblingElement("Row")) {
|
||
tinyxml2::XMLElement* pathCell = cellAt(row, pathCol);
|
||
std::string rawPath = cellText(pathCell);
|
||
if (rawPath.empty()) continue;
|
||
std::string desc = descForPath(rawPath, descByDo, orderedKeys);
|
||
if (desc.empty()) continue;
|
||
tinyxml2::XMLElement* icdCell = ensureCell(doc, row, icdCol);
|
||
setCellText(doc, icdCell, desc);
|
||
++matched;
|
||
}
|
||
|
||
tinyxml2::XMLPrinter printer;
|
||
doc.Print(&printer);
|
||
writeAllBinary(fs::u8path(workbookPath), printer.CStr());
|
||
if (message) {
|
||
*message = "ICD 导入完成，DOI=" + std::to_string(descByDo.size()) +
|
||
"，填入未匹配原始数据=" + std::to_string(matched);
|
||
}
|
||
return true;
|
||
} catch (const std::exception& e) {
|
||
if (message) *message = e.what();
|
||
return false;
|
||
}
|
||
}
|