Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feat/#12 #14

Merged
merged 6 commits into from
Mar 10, 2024
Merged

Feat/#12 #14

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f3652d0
wip
TakenPt Mar 5, 2024
93f8df6
Merge branch 'main' into feat/#12
TakenPt Mar 5, 2024
16a7f93
会話文の開始、終了記号に『及び』を認めるよう変更
TakenPt Mar 6, 2024
718b8fd
著者名がaタグでなくても著者名を取得できるよう修正
TakenPt Mar 6, 2024
f91fbf6
不要な行(処理)を削除
TakenPt Mar 7, 2024
f7135a4
フォーマット
TakenPt Mar 8, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 27 additions & 16 deletions Epub/KoeBook.Epub/Services/ScrapingNaroService.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,12 +25,23 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
var doc = await context.OpenAsync(url, ct).ConfigureAwait(false);

// title の取得
var bookTitle = doc.QuerySelector(".novel_title")
var bookTitleElement = doc.QuerySelector(".novel_title")
?? throw new EpubDocumentException($"Failed to get title properly.\nUrl may be not collect");
var bookTitle = bookTitleElement.InnerHtml;

// auther の取得
var bookAuther = doc.QuerySelector(".novel_writername a")
var bookAutherElement = doc.QuerySelector(".novel_writername")
?? throw new EpubDocumentException($"Failed to get auther properly.\nUrl may be not collect");
var bookAuther = string.Empty;
if (bookAutherElement.QuerySelector("a") is IHtmlAnchorElement bookAutherAnchorElement)
{
bookAuther = bookAutherAnchorElement.InnerHtml;
}
else
{
bookAuther = bookAutherElement.InnerHtml.Replace("作者：", "");
}

bool isRensai = true;
int allNum = 0;

Expand Down Expand Up @@ -65,14 +76,13 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
throw new EpubDocumentException("faild to get data by Narou API");
}

var document = new EpubDocument(bookTitle.InnerHtml, bookAuther.InnerHtml, coverFilePath, id);
if (isRensai)
var document = new EpubDocument(bookTitle, bookAuther, coverFilePath, id);
if (isRensai) // 連載の時
{
List<SectionWithChapterTitle> SectionWithChapterTitleList = new List<SectionWithChapterTitle>();
for (int i = 1; i <= allNum; i++)
{
Console.WriteLine(i);
await Task.Delay(500, ct);
await Task.Delay(1500, ct);
var pageUrl = Path.Combine(url, i.ToString());
var load = await ReadPageAsync(pageUrl, isRensai, imageDirectory, ct).ConfigureAwait(false);
SectionWithChapterTitleList.Add(load);
Expand Down Expand Up @@ -106,7 +116,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
}
}
}
else
else // 短編の時
{
var load = await ReadPageAsync(url, isRensai, imageDirectory, ct).ConfigureAwait(false);
if (load != null)
Expand All @@ -122,19 +132,22 @@ public record BookInfo(int? allcount, int? noveltype, int? general_all_no);

private record SectionWithChapterTitle(string? title, Section section);

private async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRensai, string imageDirectory, CancellationToken ct)
private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRensai, string imageDirectory, CancellationToken ct)
{
var config = Configuration.Default.WithDefaultLoader();
using var context = BrowsingContext.New(config);
var doc = await context.OpenAsync(url, ct).ConfigureAwait(false);

var chapterTitleElement = doc.QuerySelector(".chapter_title");
string? chapterTitle = null;
if (chapterTitleElement != null)
if (!isRensai)
{
if (chapterTitleElement.InnerHtml != null)
var chapterTitleElement = doc.QuerySelector(".chapter_title");
if (chapterTitleElement != null)
{
chapterTitle = chapterTitleElement.InnerHtml;
if (chapterTitleElement.InnerHtml != null)
{
chapterTitle = chapterTitleElement.InnerHtml;
}
}
}

Expand All @@ -148,11 +161,10 @@ private async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRen
sectionTitleElement = doc.QuerySelector(".novel_title");
}

string sectionTitle = "";
if (sectionTitleElement == null)
throw new EpubDocumentException("Can not find title of page");

sectionTitle = sectionTitleElement.InnerHtml;
var sectionTitle = sectionTitleElement.InnerHtml;

var section = new Section(sectionTitleElement.InnerHtml);

Expand Down Expand Up @@ -223,6 +235,7 @@ private async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRen
if (tags.TagName != "RUBY")
{
isAllRuby = false;
break;
}
}

Expand All @@ -241,8 +254,6 @@ private async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRen
return new SectionWithChapterTitle(chapterTitle, section);
}



[System.Text.RegularExpressions.GeneratedRegex(@"https://.{5,7}.syosetu.com/(.{7}).?")]
private static partial System.Text.RegularExpressions.Regex UrlToNcode();

Expand Down
10 changes: 5 additions & 5 deletions Epub/KoeBook.Epub/Utility/ScrapingHelper.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,16 +4,16 @@ public static class ScrapingHelper
{
public static List<string> SplitBrace(string text)
{
if (text.Length == 1 && (text == "「" || text == "」"))
if (text.Length == 1 && (text == "「" || text == "『" || text == "」" || text == "』"))
return [text];

var bracket = 0;
var brackets = new int[text.Length];
for (var i = 0; i < text.Length; i++)
{
var c = text[i];
if (c == '「') bracket++;
else if (c == '」') bracket--;
if (c == '「' || c == '『') bracket++;
else if (c == '」' || c == '』') bracket--;
brackets[i] = bracket;
}

Expand All @@ -23,12 +23,12 @@ public static List<string> SplitBrace(string text)
for (var i = 0; i < brackets.Length; i++)
{
brackets[i] -= mn;
if (text[i] == '「' && brackets[i] == 1 && i != 0 && startIdx != i)
if ((text[i] == '「' || text[i] == '『') && brackets[i] == 1 && i != 0 && startIdx != i)
{
result.Add(text[startIdx..i]);
startIdx = i;
}
if (text[i] == '」' && brackets[i] == 0)
if ((text[i] == '」' || text[i] == '』') && brackets[i] == 0)
{
result.Add(text[startIdx..(i + 1)]);
startIdx = i + 1;
Expand Down
47 changes: 46 additions & 1 deletion KoeBook.Test/Epub/ScrapingHelperTest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@ public class ScrapingHelperTest
public static object[][] TestCases()
{
(string, List<string>)[] cases = [
// '「''」'のみの場合のケース
("「", ["「"]),
("」", ["」"]),
("a", ["a"]),
Expand All @@ -29,7 +30,51 @@ public static object[][] TestCases()
("abc「abc「abc」abc", ["abc", "「abc「abc」abc"]),
("abc「abc」abc」abc", ["abc「abc」abc」", "abc"]),
("abc「abc「abc", ["abc", "「abc「abc"]),
("abc」abc」abc", ["abc」abc」", "abc"])
("abc」abc」abc", ["abc」abc」", "abc"]),
// '『''』'のみの場合のケース
("『", ["『"]),
("』", ["』"]),
("a", ["a"]),
("abc『abc』abc", ["abc", "『abc』", "abc"]),
("abc『abc』", ["abc", "『abc』"]),
("『abc』abc", ["『abc』", "abc",]),
("abc『abc』", ["abc", "『abc』"]),
("『abc』", ["『abc』",]),
("abc『abc", ["abc", "『abc"]),
("abc『", ["abc", "『"]),
("『abc", ["『abc"]),
("abc』abc", ["abc』", "abc"]),
("abc』", ["abc』"]),
("』abc", ["』", "abc"]),
("abc『abc』abc『abc』abc", ["abc", "『abc』", "abc", "『abc』", "abc"]),
("『abc』abc『abc』abc", ["『abc』", "abc", "『abc』", "abc"]),
("abc『abc』『abc』abc", ["abc", "『abc』", "『abc』", "abc"]),
("abc『abc』abc『abc』", ["abc", "『abc』", "abc", "『abc』"]),
("abc『abc『abc』abc』abc", ["abc", "『abc『abc』abc』", "abc"]),
("abc『abc『abc』abc", ["abc", "『abc『abc』abc"]),
("abc『abc』abc』abc", ["abc『abc』abc』", "abc"]),
("abc『abc『abc", ["abc", "『abc『abc"]),
("abc』abc』abc", ["abc』abc』", "abc"]),
// '「''」''『''』'が混在するパターン
("abc「abc」abc『abc』abc", ["abc", "「abc」", "abc", "『abc』", "abc"]),
("abc『abc』abc「abc」abc", ["abc", "『abc』", "abc", "「abc」", "abc"]),
("「abc」abc『abc』abc", ["「abc」", "abc", "『abc』", "abc"]),
("『abc』abc「abc」abc", ["『abc』", "abc", "「abc」", "abc"]),
("abc「abc」『abc』abc", ["abc", "「abc」", "『abc』", "abc"]),
("abc『abc』「abc」abc", ["abc", "『abc』", "「abc」", "abc"]),
("abc「abc」abc『abc』", ["abc", "「abc」", "abc", "『abc』"]),
("abc『abc』abc「abc」", ["abc", "『abc』", "abc", "「abc」"]),
("abc「abc『abc』abc」abc", ["abc", "「abc『abc』abc」", "abc"]),
("abc『abc「abc」abc』abc", ["abc", "『abc「abc」abc』", "abc"]),
("abc「abc『abc』abc", ["abc", "「abc『abc』abc"]),
("abc『abc「abc」abc", ["abc", "『abc「abc」abc"]),
("abc「abc」abc』abc", ["abc「abc」abc』", "abc"]),
("abc『abc』abc」abc", ["abc『abc』abc」", "abc"]),
("abc「abc『abc", ["abc", "「abc『abc"]),
("abc『abc「abc", ["abc", "『abc「abc"]),
("abc」abc』abc", ["abc」abc』", "abc"]),
("abc』abc」abc", ["abc』abc」", "abc"]),
("abc』abc』abc", ["abc』abc』", "abc"])
];
return cases.Select(c => new object[] { c.Item1, c.Item2 }).ToArray();
}
Expand Down
Loading