epubgrep/src/zip.cpp

/*  This file is part of epubgrep.
 *  Copyright © 2021 tastytea <tastytea@tastytea.de>
 *
 *  This program is free software: you can redistribute it and/or modify
 *  it under the terms of the GNU Affero General Public License as published by
 *  the Free Software Foundation, version 3.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU Affero General Public License for more details.
 *
 *  You should have received zipfile copy of the GNU Affero General Public
 * License along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#include "zip.hpp"

#include "fs-compat.hpp"

#include <archive.h>
#include <archive_entry.h>
#include <boost/locale/message.hpp>
#include <fmt/format.h>
#include <fmt/ostream.h> // For compatibility with fmt 4.
#include <pugixml.hpp>

#include <cstdlib>
#include <cstring>
#include <fstream>
#include <iostream>
#include <stdexcept>
#include <string>
#include <string_view>
#include <vector>

namespace epubgrep::zip
{

using boost::locale::translate;
using fmt::format;
using std::cerr;

std::vector<std::string> list(const fs::path &filepath)
{
    auto *zipfile{open_file(filepath)};

    struct archive_entry *entry{};
    std::vector<std::string> toc;
    while (archive_read_next_header(zipfile, &entry) == ARCHIVE_OK)
    {
        const auto *in_epub_filepath{archive_entry_pathname_utf8(entry)};
        if (in_epub_filepath == nullptr)
        { // If the encoding is broken, we skip the file.
            std::cerr << translate("WARNING: ")
                      << format(translate("File in {0:s} is damaged. "
                                          "Skipping in-EPUB file.\n")
                                    .str()
                                    .data(),
                                filepath);
            continue;
        }
        toc.emplace_back(in_epub_filepath);
        archive_read_data_skip(zipfile);
    }

    close_file(zipfile, filepath);

    return toc;
}

std::string read_file(const fs::path &filepath, std::string_view entry_path)
{
    auto *zipfile{open_file(filepath)};

    struct archive_entry *entry{};
    while (archive_read_next_header(zipfile, &entry) == ARCHIVE_OK)
    {
        const auto *path{archive_entry_pathname_utf8(entry)};
        if (path == nullptr)
        { // If the encoding is broken, we skip the file.
            std::cerr << translate("WARNING: ")
                      << format(translate("File in {0:s} is damaged. "
                                          "Skipping in-EPUB file.\n")
                                    .str()
                                    .data(),
                                filepath);
            continue;
        }
        if (std::strcmp(path, entry_path.data()) == 0)
        {
            const auto length{static_cast<size_t>(archive_entry_size(entry))};
            std::string filecontents;
            filecontents.resize(length);
            auto result_length{static_cast<size_t>(
                archive_read_data(zipfile, &filecontents[0], length))};

            if (result_length != length)
            {
                close_file(zipfile, filepath);

                throw exception{
                    format(translate("Could not read {0:s} in {1:s}.").str(),
                           entry_path, filepath.string())};
            }

            close_file(zipfile, filepath);

            return filecontents;
        }
        archive_read_data_skip(zipfile);
    }

    close_file(zipfile, filepath);

    if (entry_path == "META-INF/container.xml")
    { // File is probably not an EPUB.
        exception e{format(translate("{0:s} not found in {1:s}.").str(),
                           entry_path, filepath.string())};
        e.code = 1;
        throw exception{e};
    }

    cerr << translate("WARNING: ")
         << format(translate("{0:s} not found in {1:s}.").str(), entry_path,
                   filepath.string())
         << '\n';
    return {};
}

struct archive *open_file(const fs::path &filepath)
{
    // Throw exception if we can't open the file.
    std::ifstream file;
    file.exceptions(std::ios::failbit);
    file.open(filepath);
    file.close();

    auto *zipfile{archive_read_new()};
    archive_read_support_filter_all(zipfile);
    archive_read_support_format_zip(zipfile);

    auto result{archive_read_open_filename(zipfile, filepath.c_str(), 10240)};
    if (result != ARCHIVE_OK)
    {
        close_file(zipfile, filepath);

        exception e{format(translate("Could not open {0:s}.").str(),
                           filepath.string())};
        e.code = 1;
        throw exception{e};
    }

    return zipfile;
}

void close_file(struct archive *zipfile, const fs::path &filepath)
{
    auto result{archive_read_free(zipfile)};
    if (result != ARCHIVE_OK)
    {
        throw exception{format(translate("Could not close {0:s}.").str(),
                               filepath.string())};
    }
}

std::vector<std::string> list_spine(const fs::path &filepath)
{
    const fs::path opf_file_path{
        [&filepath]
        {
            pugi::xml_document xml;
            const std::string container{
                read_file(filepath, "META-INF/container.xml")};
            const auto result{xml.load_buffer(&container[0], container.size())};
            if (result)
            {
                return xml.child("container")
                    .child("rootfiles")
                    .first_child()
                    .attribute("full-path")
                    .value();
            }

            return "";
        }()};

    std::vector<std::string> spine_filepaths;
    if (!opf_file_path.empty())
    {
        pugi::xml_document xml;
        const std::string opf_file{read_file(filepath, opf_file_path.string())};
        const auto result{xml.load_buffer(&opf_file[0], opf_file.size())};
        if (result)
        {
            auto manifest{xml.child("package").child("manifest")};
            if (manifest == nullptr)
            {
                manifest = xml.child("opf:package").child("opf:manifest");
            }
            auto spine{xml.child("package").child("spine")};
            if (spine == nullptr)
            {
                spine = xml.child("opf:package").child("opf:spine");
            }

            for (const auto &itemref : spine)
            {
                const auto &idref{itemref.attribute("idref").value()};
                const auto &item{manifest.find_child_by_attribute("id", idref)};
                const std::string href{
                    urldecode(item.attribute("href").value())};
                if (href[0] != '/')
                {
                    spine_filepaths.emplace_back(
                        opf_file_path.parent_path() /= href);
                    continue;
                }
                spine_filepaths.emplace_back(href);
            }
        }
    }

    if (opf_file_path.empty() || spine_filepaths.empty())
    {
        std::cerr << translate("ERROR: ")
                  << format(translate("{0:s} is damaged. Could not read spine. "
                                      "Skipping file.\n")
                                .str()
                                .data(),
                            filepath);
        return {};
    }

    return spine_filepaths;
}

std::string urldecode(const std::string_view url)
{ // RFC 3986, section 2.1.
    size_t pos{0};
    size_t lastpos{0};
    std::string decoded;
    while ((pos = url.find('%', pos)) != std::string_view::npos)
    {
        decoded += url.substr(lastpos, pos - lastpos);
        decoded += static_cast<char>(
            std::stoul(std::string(url.substr(pos + 1, 2)), nullptr, 16));
        pos += 3;
        lastpos = pos;
    }
    decoded += url.substr(lastpos);

    return decoded;
}

} // namespace epubgrep::zip