using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.IO;
using System.IO.Compression;
using System.Linq;
using System.Xml.Serialization;
using API.Archive;
using API.Extensions;
using API.Interfaces.Services;
using API.Services.Tasks;
using Microsoft.Extensions.Logging;
using Microsoft.IO;
using SharpCompress.Archives;
using SharpCompress.Common;
using Image = NetVips.Image;
namespace API.Services
{
///
/// Responsible for manipulating Archive files. Used by and
///
public class ArchiveService : IArchiveService
{
private readonly ILogger _logger;
private const int ThumbnailWidth = 320; // 153w x 230h
private static readonly RecyclableMemoryStreamManager _streamManager = new RecyclableMemoryStreamManager();
public ArchiveService(ILogger logger)
{
_logger = logger;
}
///
/// Checks if a File can be opened. Requires up to 2 opens of the filestream.
///
///
///
public ArchiveLibrary CanOpen(string archivePath)
{
if (!File.Exists(archivePath) || !Parser.Parser.IsArchive(archivePath)) return ArchiveLibrary.NotSupported;
try
{
using var a2 = ZipFile.OpenRead(archivePath);
return ArchiveLibrary.Default;
}
catch (Exception)
{
try
{
using var a1 = ArchiveFactory.Open(archivePath);
return ArchiveLibrary.SharpCompress;
}
catch (Exception)
{
return ArchiveLibrary.NotSupported;
}
}
}
public int GetNumberOfPagesFromArchive(string archivePath)
{
if (!IsValidArchive(archivePath))
{
_logger.LogError("Archive {ArchivePath} could not be found", archivePath);
return 0;
}
try
{
var libraryHandler = CanOpen(archivePath);
switch (libraryHandler)
{
case ArchiveLibrary.Default:
{
_logger.LogDebug("Using default compression handling");
using ZipArchive archive = ZipFile.OpenRead(archivePath);
return archive.Entries.Count(e => !e.FullName.Contains("__MACOSX") && Parser.Parser.IsImage(e.FullName));
}
case ArchiveLibrary.SharpCompress:
{
_logger.LogDebug("Using SharpCompress compression handling");
using var archive = ArchiveFactory.Open(archivePath);
return archive.Entries.Count(entry => !entry.IsDirectory &&
!(Path.GetDirectoryName(entry.Key) ?? string.Empty).Contains("__MACOSX")
&& Parser.Parser.IsImage(entry.Key));
}
case ArchiveLibrary.NotSupported:
_logger.LogError("[GetNumberOfPagesFromArchive] This archive cannot be read: {ArchivePath}. Defaulting to 0 pages", archivePath);
return 0;
default:
_logger.LogError("[GetNumberOfPagesFromArchive] There was an exception when reading archive stream: {ArchivePath}. Defaulting to 0 pages", archivePath);
return 0;
}
}
catch (Exception ex)
{
_logger.LogError(ex, "[GetNumberOfPagesFromArchive] There was an exception when reading archive stream: {ArchivePath}. Defaulting to 0 pages", archivePath);
return 0;
}
}
///
/// Generates byte array of cover image.
/// Given a path to a compressed file (zip, rar, cbz, cbr, etc), will ensure the first image is returned unless
/// a folder.extension exists in the root directory of the compressed file.
///
///
/// Create a smaller variant of file extracted from archive. Archive images are usually 1MB each.
///
public byte[] GetCoverImage(string archivePath, bool createThumbnail = false)
{
if (archivePath == null || !IsValidArchive(archivePath)) return Array.Empty();
try
{
var libraryHandler = CanOpen(archivePath);
switch (libraryHandler)
{
case ArchiveLibrary.Default:
{
_logger.LogDebug("Using default compression handling");
using var archive = ZipFile.OpenRead(archivePath);
var folder = archive.Entries.SingleOrDefault(x => !x.FullName.Contains("__MACOSX") && Path.GetFileNameWithoutExtension(x.Name).ToLower() == "folder");
var entries = archive.Entries.Where(x => Path.HasExtension(x.FullName) && !x.FullName.Contains("__MACOSX") && Parser.Parser.IsImage(x.FullName)).OrderBy(x => x.FullName).ToList();
var entry = folder ?? entries[0];
return createThumbnail ? CreateThumbnail(entry) : ConvertEntryToByteArray(entry);
}
case ArchiveLibrary.SharpCompress:
{
_logger.LogDebug("Using SharpCompress compression handling");
using var archive = ArchiveFactory.Open(archivePath);
return FindCoverImage(archive.Entries.Where(entry => !entry.IsDirectory
&& !(Path.GetDirectoryName(entry.Key) ?? string.Empty).Contains("__MACOSX")
&& Parser.Parser.IsImage(entry.Key)), createThumbnail);
}
case ArchiveLibrary.NotSupported:
_logger.LogError("[GetCoverImage] This archive cannot be read: {ArchivePath}. Defaulting to no cover image", archivePath);
return Array.Empty();
default:
_logger.LogError("[GetCoverImage] There was an exception when reading archive stream: {ArchivePath}. Defaulting to no cover image", archivePath);
return Array.Empty();
}
}
catch (Exception ex)
{
_logger.LogError(ex, "[GetCoverImage] There was an exception when reading archive stream: {ArchivePath}. Defaulting to no cover image", archivePath);
}
return Array.Empty();
}
private byte[] FindCoverImage(IEnumerable entries, bool createThumbnail)
{
var images = entries.ToList();
foreach (var entry in images)
{
if (Path.GetFileNameWithoutExtension(entry.Key).ToLower() == "folder")
{
using var ms = _streamManager.GetStream();
entry.WriteTo(ms);
ms.Position = 0;
var data = ms.ToArray();
return createThumbnail ? CreateThumbnail(data, Path.GetExtension(entry.Key)) : data;
}
}
if (images.Any())
{
var entry = images.OrderBy(e => e.Key).FirstOrDefault();
if (entry == null) return Array.Empty();
using var ms = _streamManager.GetStream();
entry.WriteTo(ms);
ms.Position = 0;
var data = ms.ToArray();
return createThumbnail ? CreateThumbnail(data, Path.GetExtension(entry.Key)) : data;
}
return Array.Empty();
}
private static byte[] ConvertEntryToByteArray(ZipArchiveEntry entry)
{
using var stream = entry.Open();
using var ms = _streamManager.GetStream();
stream.CopyTo(ms);
return ms.ToArray();
}
///
/// Given an archive stream, will assess whether directory needs to be flattened so that the extracted archive files are directly
/// under extract path and not nested in subfolders. See Flatten method.
///
/// An opened archive stream
///
public bool ArchiveNeedsFlattening(ZipArchive archive)
{
// Sometimes ZipArchive will list the directory and others it will just keep it in the FullName
return archive.Entries.Count > 0 &&
!Path.HasExtension(archive.Entries.ElementAt(0).FullName) ||
archive.Entries.Any(e => e.FullName.Contains(Path.AltDirectorySeparatorChar) && !e.FullName.Contains("__MACOSX"));
}
private byte[] CreateThumbnail(byte[] entry, string formatExtension = ".jpg")
{
if (!formatExtension.StartsWith("."))
{
formatExtension = "." + formatExtension;
}
try
{
using var thumbnail = Image.ThumbnailBuffer(entry, ThumbnailWidth);
return thumbnail.WriteToBuffer(formatExtension);
}
catch (Exception ex)
{
_logger.LogError(ex, "[CreateThumbnail] There was a critical error and prevented thumbnail generation. Defaulting to no cover image. Format Extension {Extension}", formatExtension);
}
return Array.Empty();
}
private byte[] CreateThumbnail(ZipArchiveEntry entry, string formatExtension = ".jpg")
{
if (!formatExtension.StartsWith("."))
{
formatExtension = $".{formatExtension}";
}
try
{
using var stream = entry.Open();
using var thumbnail = Image.ThumbnailStream(stream, ThumbnailWidth);
return thumbnail.WriteToBuffer(formatExtension);
}
catch (Exception ex)
{
_logger.LogError(ex, "There was a critical error and prevented thumbnail generation on {EntryName}. Defaulting to no cover image", entry.FullName);
}
return Array.Empty();
}
///
/// Test if the archive path exists and an archive
///
///
///
public bool IsValidArchive(string archivePath)
{
if (!File.Exists(archivePath))
{
_logger.LogError("Archive {ArchivePath} could not be found", archivePath);
return false;
}
if (Parser.Parser.IsArchive(archivePath)) return true;
_logger.LogError("Archive {ArchivePath} is not a valid archive", archivePath);
return false;
}
private static ComicInfo FindComicInfoXml(IEnumerable entries)
{
foreach (var entry in entries)
{
if (Path.GetFileNameWithoutExtension(entry.Key).ToLower().EndsWith("comicinfo") && Parser.Parser.IsXml(entry.Key))
{
using var ms = _streamManager.GetStream();
entry.WriteTo(ms);
ms.Position = 0;
var serializer = new XmlSerializer(typeof(ComicInfo));
var info = (ComicInfo) serializer.Deserialize(ms);
return info;
}
}
return null;
}
public string GetSummaryInfo(string archivePath)
{
var summary = string.Empty;
if (!IsValidArchive(archivePath)) return summary;
ComicInfo info = null;
try
{
if (!File.Exists(archivePath)) return summary;
var libraryHandler = CanOpen(archivePath);
switch (libraryHandler)
{
case ArchiveLibrary.Default:
{
_logger.LogDebug("Using default compression handling");
using var archive = ZipFile.OpenRead(archivePath);
var entry = archive.Entries.SingleOrDefault(x => !x.FullName.Contains("__MACOSX") && Path.GetFileNameWithoutExtension(x.Name).ToLower() == "comicinfo" && Parser.Parser.IsXml(x.FullName));
if (entry != null)
{
using var stream = entry.Open();
var serializer = new XmlSerializer(typeof(ComicInfo));
info = (ComicInfo) serializer.Deserialize(stream);
}
break;
}
case ArchiveLibrary.SharpCompress:
{
_logger.LogDebug("Using SharpCompress compression handling");
using var archive = ArchiveFactory.Open(archivePath);
info = FindComicInfoXml(archive.Entries.Where(entry => !entry.IsDirectory
&& !(Path.GetDirectoryName(entry.Key) ?? string.Empty).Contains("__MACOSX")
&& Parser.Parser.IsXml(entry.Key)));
break;
}
case ArchiveLibrary.NotSupported:
_logger.LogError("[GetSummaryInfo] This archive cannot be read: {ArchivePath}. Defaulting to 0 pages", archivePath);
return summary;
default:
_logger.LogError("[GetSummaryInfo] There was an exception when reading archive stream: {ArchivePath}. Defaulting to 0 pages", archivePath);
return summary;
}
if (info != null)
{
return info.Summary;
}
_logger.LogError("[GetSummaryInfo] Could not parse archive file: {Filepath}", archivePath);
}
catch (Exception ex)
{
_logger.LogError(ex, "[GetSummaryInfo] There was an exception when reading archive stream: {Filepath}", archivePath);
}
return summary;
}
private static void ExtractArchiveEntities(IEnumerable entries, string extractPath)
{
DirectoryService.ExistOrCreate(extractPath);
foreach (var entry in entries)
{
entry.WriteToDirectory(extractPath, new ExtractionOptions()
{
ExtractFullPath = false,
Overwrite = false
});
}
}
private void ExtractArchiveEntries(ZipArchive archive, string extractPath)
{
var needsFlattening = ArchiveNeedsFlattening(archive);
if (!archive.HasFiles() && !needsFlattening) return;
archive.ExtractToDirectory(extractPath, true);
if (needsFlattening)
{
_logger.LogDebug("Extracted archive is nested in root folder, flattening...");
new DirectoryInfo(extractPath).Flatten();
}
}
///
/// Extracts an archive to a temp cache directory. Returns path to new directory. If temp cache directory already exists,
/// will return that without performing an extraction. Returns empty string if there are any invalidations which would
/// prevent operations to perform correctly (missing archivePath file, empty archive, etc).
///
/// A valid file to an archive file.
/// Path to extract to
///
public void ExtractArchive(string archivePath, string extractPath)
{
if (!IsValidArchive(archivePath)) return;
if (Directory.Exists(extractPath)) return;
var sw = Stopwatch.StartNew();
try
{
var libraryHandler = CanOpen(archivePath);
switch (libraryHandler)
{
case ArchiveLibrary.Default:
{
_logger.LogDebug("Using default compression handling");
using var archive = ZipFile.OpenRead(archivePath);
ExtractArchiveEntries(archive, extractPath);
break;
}
case ArchiveLibrary.SharpCompress:
{
_logger.LogDebug("Using SharpCompress compression handling");
using var archive = ArchiveFactory.Open(archivePath);
ExtractArchiveEntities(archive.Entries.Where(entry => !entry.IsDirectory
&& !(Path.GetDirectoryName(entry.Key) ?? string.Empty).Contains("__MACOSX")
&& Parser.Parser.IsImage(entry.Key)), extractPath);
break;
}
case ArchiveLibrary.NotSupported:
_logger.LogError("[GetNumberOfPagesFromArchive] This archive cannot be read: {ArchivePath}. Defaulting to 0 pages", archivePath);
return;
default:
_logger.LogError("[GetNumberOfPagesFromArchive] There was an exception when reading archive stream: {ArchivePath}. Defaulting to 0 pages", archivePath);
return;
}
}
catch (Exception e)
{
_logger.LogError(e, "There was a problem extracting {ArchivePath} to {ExtractPath}",archivePath, extractPath);
return;
}
_logger.LogDebug("Extracted archive to {ExtractPath} in {ElapsedMilliseconds} milliseconds", extractPath, sw.ElapsedMilliseconds);
}
}
}