common_xml_document_parser.h

/*********************************************************************************************************************************************/
/*  DocWire SDK: Award-winning modern data processing in C++20. SourceForge Community Choice & Microsoft support. AI-driven processing.      */
/*  Supports nearly 100 data formats, including email boxes and OCR. Boost efficiency in text extraction, web data extraction, data mining,  */
/*  document analysis. Offline processing possible for security and confidentiality                                                          */
/*                                                                                                                                           */
/*  Copyright (c) SILVERCODERS Ltd, http://silvercoders.com                                                                                  */
/*  Project homepage: https://github.com/docwire/docwire                                                                                     */
/*                                                                                                                                           */
/*  SPDX-License-Identifier: AGPL-3.0-only OR LicenseRef-DocWire-Commercial                                                                  */
/*********************************************************************************************************************************************/
 
#ifndef DOCWIRE_COMMON_XML_PARSER_H
#define DOCWIRE_COMMON_XML_PARSER_H
 
#include "attributes.h"
#include "chain_element.h"
#include "pimpl.h"
#include "xml_children.h"
#include <string>
#include <vector>
#include <map>
 
namespace docwire
{
    class zip_reader;
    class Metadata;
 
enum xml_parse_mode { PARSE_XML, FIX_XML, STRIP_XML };
 
template <safety_policy safety_level = default_safety_level>
class common_xml_document_parser: public chain_element, public with_pimpl<common_xml_document_parser<safety_level>>
{
    private:
        friend pimpl_impl<common_xml_document_parser<safety_level>>;
        using with_pimpl<common_xml_document_parser<safety_level>>::impl;
 
    //public interface for derived classes (and its components)
    public:
        enum odfooxml_list_style
        {
            number,
            bullet
        };
 
        struct comment
        {
            std::string m_author;
            std::string m_time;
            std::string m_text;
            comment() {}
            comment(const std::string& author, const std::string& time, const std::string& text)
                : m_author(author), m_time(time), m_text(text) {}
        };
 
        struct relationship
        {
            std::string m_target;
        };
 
        struct shared_string
        {
            std::string m_text;
        };
 
        typedef std::vector<odfooxml_list_style> ListStyleVector;
        using ListStyleMap = std::map<std::string, common_xml_document_parser<safety_level>::ListStyleVector>;
        using CommentMap = std::map<int, common_xml_document_parser<safety_level>::comment>;
        using RelationshipMap = std::map<std::string, common_xml_document_parser<safety_level>::relationship>;
        using SharedStringVector = std::vector<shared_string>;
 
        typedef std::function<void(xml::node_ref<safety_level>& xml_node, xml_parse_mode mode,
                                 zip_reader* zipfile, std::string& text,
                                 bool& children_processed, std::string& level_suffix, bool first_on_level)> CommandHandler;
 
        void registerODFOOXMLCommandHandler(const std::string& xml_tag, const CommandHandler& handler);
 
        std::string parseXmlData(xml::children_view<safety_level> xml_nodes, xml_parse_mode mode, zip_reader* zipfile);
 
        std::string parseXmlChildren(xml::node_ref<safety_level>& xml_node, xml_parse_mode mode, zip_reader* zipfile);
 
        void extractText(std::string_view xml_contents, xml_parse_mode mode, zip_reader* zipfile, std::string& text);
 
        void parseODFMetadata(std::string_view xml_content, attributes::metadata& metadata) const;
 
        const std::string formatComment(const std::string& author, const std::string& time, const std::string& text);
 
        size_t& getListDepth();
 
        ListStyleMap& getListStyles();
 
        CommentMap& getComments();
 
        RelationshipMap& getRelationships();
 
        SharedStringVector& getSharedStrings();
 
        bool disabledText() const;
 
        xml::reader_blanks blanks() const;
 
        void disableText(bool disable);
 
        void set_blanks(xml::reader_blanks blanks);
 
        void activeEmittingSignals(bool flag);
 
    //public interface
    public:
        common_xml_document_parser();
 
    protected:
        class scoped_context_stack_push
        {
        public:
            scoped_context_stack_push(common_xml_document_parser& parser, const message_callbacks& emit_message);
            ~scoped_context_stack_push();
        private:
            common_xml_document_parser& m_parser;
        };
};
 
} // namespace docwire
 
#endif