testing J2doll/sax reader in c++11

Example.md

@@ -235,4 +235,42 @@ LargeData -r 300000 -c 15 -S 60000 --use-style
   - Reorder sheets (move left/right)
 - Save or Save As to produce a modified workbook
 
+<br />
+
+---
+
+## [sax reader](https://github.com/QtExcel/QXlsx/blob/j2doll/sax_reader/TestExcel/extractdata_sax.cpp)
+
+```cpp
+void dump_all_sheets_sax(QXlsx::Document& doc)
+{
+    QXlsx::sax_options opt;
+    opt.resolve_shared_strings = true;     // If there are many strings, set to false to save more RAM (but sharedString index may be output instead)
+    opt.read_formulas_as_text  = true;     // Output formula as string instead of result value (set false if not needed)
+    opt.stop_on_empty_sheetdata = false;   // Continue even if sheetData is empty
+
+    const QStringList sheets = doc.sheetNames();
+    qInfo() << "sheet count:" << sheets.size();
+
+    for (const QString& sheet_name : sheets) {
+        qInfo().noquote() << "\n=== SHEET:" << sheet_name << "===";
+
+        const bool ok = doc.read_sheet_sax(
+            sheet_name,
+            opt,
+            [&](const QXlsx::sax_cell& cell) -> bool {
+                qDebug().noquote()
+                << QString("%1!R%2C%3 = %4")
+                        .arg(cell.sheet_name)
+                        .arg(cell.row)
+                        .arg(cell.col)
+                        .arg(cell.value.toString());
+                return true; // continue
+            });
+
+        qInfo() << "sheet done:" << sheet_name << "ok=" << ok;
+    }
+}
+```
+
 

HowToSetProject-cmake.md

@@ -1,6 +1,6 @@
 ## Using cmake 
 
-> *Read this in other languages: [English](HowToSetProject-cmake.md), :kr: [한국어](HowToSetProject-cmake.ko.md), :zh-CN: [简体中文](HowToSetProject-cmake.zh-CN.md)*
+> *Read this in other languages: [English](HowToSetProject-cmake.md), :cn: [简体中文](HowToSetProject-cmake.zh-CN.md)*
 
 ### To install QXlsx
 

QXlsx/CMakeLists.txt

@@ -83,6 +83,7 @@ set(SRC_FILES
     source/xlsxdocument.cpp
     source/xlsxrelationships.cpp
     source/xlsxutility.cpp
+    source/xlsxreadsax.cpp
     header/xlsxabstractooxmlfile_p.h
     header/xlsxchartsheet_p.h
     header/xlsxdocpropsapp_p.h
@@ -112,6 +113,7 @@ set(SRC_FILES
     header/xlsxdrawing_p.h
     header/xlsxrichstring_p.h
     header/xlsxutility_p.h
+    header/xlsxreadsax.h
 )
 
 set(QXLSX_PUBLIC_HEADERS

QXlsx/QXlsx.pri

@@ -100,7 +100,8 @@ $${QXLSX_HEADERPATH}xlsxworkbook_p.h \
 $${QXLSX_HEADERPATH}xlsxworksheet.h \
 $${QXLSX_HEADERPATH}xlsxworksheet_p.h \
 $${QXLSX_HEADERPATH}xlsxzipreader_p.h \
-$${QXLSX_HEADERPATH}xlsxzipwriter_p.h
+$${QXLSX_HEADERPATH}xlsxzipwriter_p.h \
+$${QXLSX_HEADERPATH}xlsxreadsax.h
 
 SOURCES += \
 $${QXLSX_SOURCEPATH}xlsxabstractooxmlfile.cpp \
@@ -135,8 +136,8 @@ $${QXLSX_SOURCEPATH}xlsxutility.cpp \
 $${QXLSX_SOURCEPATH}xlsxworkbook.cpp \
 $${QXLSX_SOURCEPATH}xlsxworksheet.cpp \
 $${QXLSX_SOURCEPATH}xlsxzipreader.cpp \
-$${QXLSX_SOURCEPATH}xlsxzipwriter.cpp
-
+$${QXLSX_SOURCEPATH}xlsxzipwriter.cpp \
+$${QXLSX_SOURCEPATH}xlsxreadsax.cpp
 
 ########################################
 # custom setting for compiler & system

QXlsx/QXlsx.pro

@@ -29,4 +29,10 @@ QXLSX_HEADERPATH=$$PWD/header/
 QXLSX_SOURCEPATH=$$PWD/source/
 include($$PWD/QXlsx.pri)
 
+HEADERS += \
+    header/xlsxreadsax.h
+
+SOURCES += \
+    source/xlsxreadsax.cpp
+
 

QXlsx/header/xlsxdocument.h

@@ -6,6 +6,7 @@
 #include "xlsxformat.h"
 #include "xlsxglobal.h"
 #include "xlsxworksheet.h"
+#include "xlsxreadsax.h"
 
 #include <QIODevice>
 #include <QImage>
@@ -132,6 +133,17 @@ class QXLSX_EXPORT Document : public QObject
     bool autosizeColumnWidth(int colFirst, int colLast);
     bool autosizeColumnWidth();
 
+
+    /////////////////////////////////
+    // New feature: Stream sheet cells with callback "without saving"
+    bool read_sheet_sax(const QString& sheet_name,
+                        const sax_options& opt,
+                        const sax_cell_callback& on_cell);
+
+    bool read_sheet_sax(int sheet_index,
+                        const sax_options& opt,
+                        const sax_cell_callback& on_cell);
+
 private:
     QMap<int, int> getMaximalColumnWidth(int firstRow = 1, int lastRow = INT_MAX);
 

QXlsx/header/xlsxdocument_p.h

@@ -10,6 +10,8 @@
 
 #include <QMap>
 
+#include <memory>
+
 QT_BEGIN_NAMESPACE_XLSX
 
 class DocumentPrivate
@@ -35,6 +37,9 @@ class DocumentPrivate
     std::shared_ptr<Workbook> workbook;
     std::shared_ptr<ContentTypes> contentTypes;
     bool isLoad;
+
+    // Store the entire xlsx (zip) bytes so that even when opened with QIODevice, the zip can be reopened in SAX
+    std::shared_ptr<QByteArray> package_bytes;
 };
 
 QT_END_NAMESPACE_XLSX

QXlsx/header/xlsxreadsax.h

@@ -0,0 +1,40 @@
+#ifndef XLSXREADSAX_H
+#define XLSXREADSAX_H
+
+#include <QXmlStreamReader>
+#include <QString>
+#include <QVariant>
+#include <functional>
+
+namespace QXlsx {
+
+struct sax_options
+{
+    bool resolve_shared_strings = true;
+    bool read_formulas_as_text = false;
+    bool stop_on_empty_sheetdata = false;
+};
+
+struct sax_cell
+{
+    QString sheet_name;
+    int row = 0;         // 1-based
+    int col = 0;         // 1-based
+    QVariant value;      // bool/double/QString etc. (minimum required)
+};
+
+using sax_cell_callback = std::function<bool(const sax_cell&)>;
+
+// Load all of sharedStrings.xml (optional) - simple implementation
+class ZipReader;
+QStringList load_shared_strings_all(ZipReader& zip);
+
+// Parse sheet.xml with SAX
+bool read_sheet_xml_sax(const QByteArray& sheet_xml,
+                        const sax_options& opt,
+                        const QStringList* shared_strings, // nullptr 가능
+                        const sax_cell_callback& on_cell);
+
+} // namespace QXlsx
+
+#endif // XLSXREADSAX_H

QXlsx/source/xlsxdocument.cpp

@@ -20,6 +20,8 @@
 #include "xlsxzipreader_p.h"
 #include "xlsxzipwriter_p.h"
 
+#include "xlsxreadsax.h"
+
 #include <QBuffer>
 #include <QDebug>
 #include <QDir>
@@ -1557,4 +1559,67 @@ bool Document::autosizeColumnWidth()
     return erg;
 }
 
+/////////////////////////////////////////////////////////////////////
+// ======================= SAX streaming API =========================
+bool Document::read_sheet_sax(int sheet_index,
+                              const sax_options& opt,
+                              const sax_cell_callback& on_cell)
+{
+    if (!d_ptr || !d_ptr->workbook)
+        return false;
+
+           // Open zip (supports both file path and QIODevice based)
+    std::unique_ptr<QIODevice> owned_device;
+
+    if (!d_ptr->packageName.isEmpty()) {
+        std::unique_ptr<QFile> f(new QFile(d_ptr->packageName));
+        if (!f->open(QIODevice::ReadOnly))
+            return false;
+        owned_device = std::move(f);
+    } else if (d_ptr->package_bytes && !d_ptr->package_bytes->isEmpty()) {
+        std::unique_ptr<QBuffer> b(new QBuffer(d_ptr->package_bytes.get()));
+        if (!b->open(QIODevice::ReadOnly))
+            return false;
+        owned_device = std::move(b);
+    } else {
+        return false;
+    }
+
+    ZipReader zip(owned_device.get());
+
+           // shared strings (optional)
+    QStringList shared_strings;
+    if (opt.resolve_shared_strings) {
+        shared_strings = QXlsx::load_shared_strings_all(zip);
+    }
+
+           // sheet XML path: workbook already has filePath (actual path determined by relationship (rels))
+    AbstractSheet *abs_sheet = d_ptr->workbook->sheet(sheet_index);
+    if (!abs_sheet)
+        return false;
+
+    const QString sheet_path = abs_sheet->filePath();
+    const QByteArray sheet_xml = zip.fileData(sheet_path);
+
+    if (sheet_xml.isEmpty())
+        return false;
+
+    return QXlsx::read_sheet_xml_sax(sheet_xml, opt,
+                                     opt.resolve_shared_strings ? &shared_strings : nullptr,
+                                     on_cell);
+}
+
+bool Document::read_sheet_sax(const QString& sheet_name,
+                              const sax_options& opt,
+                              const sax_cell_callback& on_cell)
+{
+    const QStringList names = d_ptr->workbook->worksheetNames();
+    const int idx = names.indexOf(sheet_name);
+    if (idx < 0)
+        return false;
+    return read_sheet_sax(idx, opt, on_cell);
+}
+//////////////////////////////////////////////////////////////////////
+
+
 QT_END_NAMESPACE_XLSX