AZEN-SGG/Kavita
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Migrated up to VersOne 3.3 with epub 3.3 support. (#1999)

Browse source 
This enables collection and reading list support from epubs.
This commit is contained in:
Joe Milazzo 2023-05-15 20:48:23 -05:00 • committed by GitHub
parent 466a0c8028
commit 1b06d98319
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
6 changed files with 71 additions and 179 deletions
Download patch file Download diff file Expand all files Collapse all files

105
API.Benchmark/EpubBenchmark.cs
View file

@ -1,105 +0,0 @@
using System;
using System.Linq;
using System.Text.RegularExpressions;
using System.Threading.Tasks;
using API.Services;
using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Order;
using HtmlAgilityPack;
using VersOne.Epub;
namespace API.Benchmark;
[StopOnFirstError]
[MemoryDiagnoser]
[RankColumn]
[Orderer(SummaryOrderPolicy.FastestToSlowest)]
[SimpleJob(launchCount: 1, warmupCount: 5, invocationCount: 20)]
public class EpubBenchmark
{
private const string FilePath = @"E:\Books\Invaders of the Rokujouma\Invaders of the Rokujouma - Volume 01.epub";
private readonly Regex _wordRegex = new Regex(@"\b\w+\b", RegexOptions.Compiled | RegexOptions.IgnoreCase);
[Benchmark]
public async Task GetWordCount_PassByRef()
{
using var book = await EpubReader.OpenBookAsync(FilePath, BookService.BookReaderOptions);
foreach (var bookFile in book.Content.Html.Values)
{
await GetBookWordCount_PassByRef(bookFile);
}
}
[Benchmark]
public async Task GetBookWordCount_SumEarlier()
{
using var book = await EpubReader.OpenBookAsync(FilePath, BookService.BookReaderOptions);
foreach (var bookFile in book.Content.Html.Values)
{
await GetBookWordCount_SumEarlier(bookFile);
}
}
[Benchmark]
public async Task GetBookWordCount_Regex()
{
using var book = await EpubReader.OpenBookAsync(FilePath, BookService.BookReaderOptions);
foreach (var bookFile in book.Content.Html.Values)
{
await GetBookWordCount_Regex(bookFile);
}
}
private int GetBookWordCount_PassByString(string fileContents)
{
var doc = new HtmlDocument();
doc.LoadHtml(fileContents);
var delimiter = new char[] {' '};
return doc.DocumentNode.SelectNodes("//body//text()[not(parent::script)]")
.Select(node => node.InnerText)
.Select(text => text.Split(delimiter, StringSplitOptions.RemoveEmptyEntries)
.Where(s => char.IsLetter(s[0])))
.Select(words => words.Count())
.Where(wordCount => wordCount > 0)
.Sum();
}
private async Task<int> GetBookWordCount_PassByRef(EpubContentFileRef bookFile)
{
var doc = new HtmlDocument();
doc.LoadHtml(await bookFile.ReadContentAsTextAsync());
var delimiter = new char[] {' '};
var textNodes = doc.DocumentNode.SelectNodes("//body//text()[not(parent::script)]");
if (textNodes == null) return 0;
return textNodes.Select(node => node.InnerText)
.Select(text => text.Split(delimiter, StringSplitOptions.RemoveEmptyEntries)
.Where(s => char.IsLetter(s[0])))
.Select(words => words.Count())
.Where(wordCount => wordCount > 0)
.Sum();
}
private async Task<int> GetBookWordCount_SumEarlier(EpubContentFileRef bookFile)
{
var doc = new HtmlDocument();
doc.LoadHtml(await bookFile.ReadContentAsTextAsync());
return doc.DocumentNode.SelectNodes("//body//text()[not(parent::script)]")
.DefaultIfEmpty()
.Select(node => node.InnerText.Split(' ', StringSplitOptions.RemoveEmptyEntries)
.Where(s => char.IsLetter(s[0])))
.Sum(words => words.Count());
}
private async Task<int> GetBookWordCount_Regex(EpubContentFileRef bookFile)
{
var doc = new HtmlDocument();
doc.LoadHtml(await bookFile.ReadContentAsTextAsync());
return doc.DocumentNode.SelectNodes("//body//text()[not(parent::script)]")
.Sum(node => _wordRegex.Matches(node.InnerText).Count);
}
}