qt5-按字节读取并解析含属性的xml文件

参考：
对XML文件读取和编辑2-QXmlStreamReader读取 - 知乎
https://zhuanlan.zhihu.com/p/358862429

本地环境：
win10专业版，64位，Qt 5.12

代码已测试通过。

---

问题描述

需要按字节读取一个文档，解析其中具有xml格式的部分，并存储到一个Hash表中，方便使用。xml标签上可能带有属性信息，如下图红线所示：

解决思路

按字节读取参考：
qt5-入门-QByteArray-CSDN博客
https://blog.csdn.net/pxy7896/article/details/137583591

提取后发现文字内容大概如下：

"<Notes><UUID>71bf0eb6-0477-41e8-8520-f1f5fafac932</UUID><Type>Synthetic</Type><ConfirmedExperimentally>0</ConfirmedExperimentally><CustomMapLabel>Y14837</CustomMapLabel><UseCustomMapLabel>1</UseCustomMapLabel><Description>Cloning vector pUC57, complete sequence.</Description><Created UTC=\"1:41:49\">2020.7.2</Created><LastModified UTC=\"1:8:0\">2021.10.8</LastModified><AccessionNumber>Y14837</AccessionNumber><SequenceClass>UNA</SequenceClass><TransformedInto>unspecified</TransformedInto><References><Reference authors=\"Markausakas A, Dreguniene G.\" journal=\"Unpublished\" title=\"A new cloning vector pUC57\"/><Reference authors=\"Markauskas A.\" journal=\"Submitted (16-SEP-1997) A. Markauskas, Fermentas AB, Graiciuno 8, Vilnius 2028, LITHUANIA\" title=\"Direct Submission\"/></References><Comments>&lt;a href='http://www.informaxinc.com/'>http://www.informaxinc.com/&lt;/a>&lt;br>ORIGDB|GenBank</Comments></Notes>" 

可以看到，<Reference>不仅携带属性信息，还可能重复，所以应该使用QMultiHash。

总的设计思路是：当QXmlStreamReader::TokenType是QXmlStreamReader::StartElement读取标签名称和属性信息，存储到合适的字典里；当是QXmlStreamReader::Characters时读取标签内容；当是QXmlStreamReader::EndElement时，存储到外层字典中，并清空临时值。这样一直读取到这部分结束。

实现

void process(QXmlStreamReader& xml, QMultiHash<QString, QHash<QString, QString>>& hash) {
	// 临时存储
    QString name, value;
    QHash<QString, QString> attrHash;
    while(!xml.atEnd()) {
        QXmlStreamReader::TokenType token = xml.readNext();
        switch ((int)token) {
            case QXmlStreamReader::NoToken:
            //qDebug()<<"没有读到任何东西";
            break;
        case QXmlStreamReader::Invalid:
            //qDebug()<<"发生错误,在error()和errorString()中报告.";
            break;
        case QXmlStreamReader::StartDocument:
            //qDebug()<<"读取文件开始-"<<"版本号:"<<xml.documentVersion()<<"编码格式:"<<xml.documentEncoding();
            break;
        case QXmlStreamReader::EndDocument:
            //qDebug()<<"读取文件结束";
            break;
        case QXmlStreamReader::StartElement:    //开始读取一个元素
        {

            // 如果是元素开始标签
            name = xml.name().toString();
            // 输出标签的属性
            QXmlStreamAttributes attributes = xml.attributes();
            // 此时有属性，需要填充字典
            if (!attributes.isEmpty()) {
                foreach (const QXmlStreamAttribute &attribute, attributes) {
                    attrHash.insert(attribute.name().toString(), attribute.value().toString());
                }
            }

        }
            break;
        case QXmlStreamReader::EndElement:  //读取一个元素结束
        {
            if(name == xml.name().toString()) {
                //attrHash.insert("name", name);
                attrHash.insert("value", value);
                hash.insert(name, attrHash);
            }
            // 清空
            name = "";
            value = "";
            attrHash.clear();
        }
            break;
        case QXmlStreamReader::Characters:  //读取元素中的文本信息
        {
            QString str = xml.text().toString();
            if(!xml.isWhitespace())
            {
                value = str;
            }
        }
            break;
        case QXmlStreamReader::Comment: //文本注释
            break;
        case QXmlStreamReader::ProcessingInstruction:
            //qDebug()<<"ProcessingInstruction: "<< xml.text();
            break;
        }
    } // 读取结束
}

使用：

QString blockContent = byteArray.mid(ptr, blockSize);
// 原始bytes中可能有\n，注意去掉。。。
QXmlStreamReader xml(blockContent);
QMultiHash<QString, QHash<QString, QString> > curHash;
process(xml, curHash);
// 打印一下结果
for (QMultiHash<QString, QHash<QString, QString>>::const_iterator it = curHash.constBegin(); it != curHash.constEnd(); ++it) {
	qDebug() << it.key() << it.value() << endl;
}

/* 解析结果
"Created" QHash(("value", "2020.7.2")("UTC", "1:41:49")) 

"LastModified" QHash(("value", "2021.10.8")("UTC", "1:8:0")) 

"Comments" QHash(("value", "<a href='http://www.informaxinc.com/'>http://www.informaxinc.com/</a><br>ORIGDB|GenBank")) 

"Type" QHash(("value", "Synthetic")) 

"Description" QHash(("value", "Cloning vector pUC57, complete sequence.")) 

"CustomMapLabel" QHash(("value", "Y14837")) 

"UseCustomMapLabel" QHash(("value", "1")) 

"ConfirmedExperimentally" QHash(("value", "0")) 

"SequenceClass" QHash(("value", "UNA")) 

"UUID" QHash(("value", "71bf0eb6-0477-41e8-8520-f1f5fafac932")) 

"TransformedInto" QHash(("value", "unspecified")) 

"Reference" QHash(("value", "")("journal", "Submitted (16-SEP-1997) A. Markauskas, Fermentas AB, Graiciuno 8, Vilnius 2028, LITHUANIA")("authors", "Markauskas A.")("title", "Direct Submission")) 

"Reference" QHash(("value", "")("journal", "Unpublished")("authors", "Markausakas A, Dreguniene G.")("title", "A new cloning vector pUC57")) 

"AccessionNumber" QHash(("value", "Y14837")) 
*/