Mirrors/SubtitleEdit
1
0
Fork 0
mirror of https://github.com/SubtitleEdit/subtitleedit.git synced 2024-11-22 03:02:35 +01:00
Code Issues Projects Releases Wiki Activity

Fix #8184 - thx Codling/ivandrofly :)

Browse Source
This commit is contained in:
Nikolaj Olsson 2024-07-02 06:52:30 +02:00
parent 9dad06922b
commit dc37f04338
2 changed files with 51 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
src/Test/FixCommonErrors/FixCommonErrorsTest.cs
View File

@ -563,6 +563,28 @@ namespace Test.FixCommonErrors
}
}
[TestMethod]
public void FixCommonOcrErrorsFrenchHardCodedRuleNoChange2()
{
using (var target = GetFixCommonErrorsLib())
{
InitializeFixCommonErrorsLine(target, "Je connaîtrai la peur." + Environment.NewLine + "La peur tue l'esprit.");
target.FixOcrErrorsViaReplaceList("fra");
Assert.AreEqual("Je connaîtrai la peur." + Environment.NewLine + "La peur tue l'esprit.", target.Subtitle.Paragraphs[0].Text);
}
}
[TestMethod]
public void FixCommonOcrErrorsFrenchHardCodedRuleNoChange3()
{
using (var target = GetFixCommonErrorsLib())
{
InitializeFixCommonErrorsLine(target, "Je connaîtrai la peur." + Environment.NewLine + "La peur tue l'Esprit tue et.");
target.FixOcrErrorsViaReplaceList("fra");
Assert.AreEqual("Je connaîtrai la peur." + Environment.NewLine + "La peur tue l'Esprit tue et.", target.Subtitle.Paragraphs[0].Text);
}
}
#endregion Fix OCR errors
#region Fix missing spaces

33
src/ui/Logic/Ocr/OcrFixEngine.cs
View File

@ -588,7 +588,7 @@ namespace Nikse.SubtitleEdit.Logic.Ocr
var affixes = new[] { " I'", " L'", " l'", " I’", " L’", " l’" };
foreach (var affix in affixes)
{
text = FixFrenchLApostrophe(text, affix, prevLine);
text = FixFrenchLApostrophe(text, affix, prevLine);
}
}
@ -666,7 +666,7 @@ namespace Nikse.SubtitleEdit.Logic.Ocr
private static bool IsToKeepCasing(string sentence)
{
// related to https://github.com/SubtitleEdit/subtitleedit/issues/8052
if (sentence.Length > 2)
if (sentence.Length > 2 && IsCurrentWordUpperCase(sentence))
{
// do not change 'L' to lowercase in text like "L'ASSASSIN"
return char.IsUpper(sentence[2]) && (sentence[1] == '\'' || sentence[1] == '’');
@ -674,12 +674,37 @@ namespace Nikse.SubtitleEdit.Logic.Ocr
return false;
}
private static bool IsCurrentWordUpperCase(string sentence)
{
var maxMinusOne = sentence.Length - 1;
for (var i = 2; i <= maxMinusOne; i++)
{
var ch = sentence[i];
if (!char.IsLetter(ch))
{
return true;
}
if (!char.IsUpper(ch))
{
return false;
}
if (i == maxMinusOne)
{
return true;
}
}
return false;
}
public static string FixFrenchLApostrophe(string input, string affix, string prevLine)
{
var text = input;
var isPreviousLineClose = prevLine.HasSentenceEnding();
if (text.StartsWith(affix.TrimStart(), StringComparison.Ordinal) && text.Length > 3)
{
if (isPreviousLineClose || char.IsUpper(text[2]))