From fa4a8204a4987fc38c056d023e6071c52c8ee133 Mon Sep 17 00:00:00 2001
From: Robin Appelman <icewind@owncloud.com>
Date: Tue, 9 Feb 2016 20:43:10 +0100
Subject: [PATCH] prefer numbers without anything appended when parsing chapter
 numbers

---
 .../tachiyomi/util/ChapterRecognition.java    | 35 ++++++++++++++-----
 .../tachiyomi/ChapterRecognitionTest.java     |  7 ++++
 2 files changed, 33 insertions(+), 9 deletions(-)

diff --git a/app/src/main/java/eu/kanade/tachiyomi/util/ChapterRecognition.java b/app/src/main/java/eu/kanade/tachiyomi/util/ChapterRecognition.java
index 26d505f699..69c3b98ac5 100644
--- a/app/src/main/java/eu/kanade/tachiyomi/util/ChapterRecognition.java
+++ b/app/src/main/java/eu/kanade/tachiyomi/util/ChapterRecognition.java
@@ -10,9 +10,11 @@ import eu.kanade.tachiyomi.data.database.models.Manga;
 
 public class ChapterRecognition {
 
-    private static final Pattern p1 = Pattern.compile("ch[^0-9]?\\s*(\\d+[\\.,]?\\d*)");
-    private static final Pattern p2 = Pattern.compile("(\\d+[\\.,]?\\d*)");
-    private static final Pattern p3 = Pattern.compile("(\\d+[\\.,]?\\d*\\s*:)");
+    private static final Pattern cleanWithToken = Pattern.compile("ch[^0-9]?\\s*(\\d+[\\.,]?\\d+)($|\\b)");
+    private static final Pattern uncleanWithToken = Pattern.compile("ch[^0-9]?\\s*(\\d+[\\.,]?\\d*)");
+    private static final Pattern cleanNumber = Pattern.compile("(\\d+[\\.,]?\\d+)($|\\b)");
+    private static final Pattern uncleanNumber = Pattern.compile("(\\d+[\\.,]?\\d*)");
+    private static final Pattern withColon = Pattern.compile("(\\d+[\\.,]?\\d*\\s*:)");
 
     private static final Pattern pUnwanted =
             Pattern.compile("(\\b|\\d)(v|ver|vol|version|volume)\\.?\\s*\\d+\\b");
@@ -24,8 +26,15 @@ public class ChapterRecognition {
         String name = chapter.name.toLowerCase();
         Matcher matcher;
 
-        // Safest option, the chapter has a token prepended
-        matcher = p1.matcher(name);
+        // Safest option, the chapter has a token prepended and nothing at the end of the number
+        matcher = cleanWithToken.matcher(name);
+        if (matcher.find()) {
+            chapter.chapter_number = Float.parseFloat(matcher.group(1));
+            return;
+        }
+
+        // the chapter has a token prepended and something at the end of the number
+        matcher = uncleanWithToken.matcher(name);
         if (matcher.find()) {
             chapter.chapter_number = Float.parseFloat(matcher.group(1));
             return;
@@ -37,7 +46,7 @@ public class ChapterRecognition {
         List<Float> occurrences;
 
         // If there's only one number, use it
-        matcher = p2.matcher(name);
+        matcher = uncleanNumber.matcher(name);
         occurrences = getAllOccurrences(matcher);
         if (occurrences.size() == 1) {
             chapter.chapter_number =  occurrences.get(0);
@@ -45,7 +54,15 @@ public class ChapterRecognition {
         }
 
         // If it has a colon, the chapter number should be that one
-        matcher = p3.matcher(name);
+        matcher = withColon.matcher(name);
+        occurrences = getAllOccurrences(matcher);
+        if (occurrences.size() == 1) {
+            chapter.chapter_number =  occurrences.get(0);
+            return;
+        }
+
+        // Prefer numbers without anything appended
+        matcher = cleanNumber.matcher(name);
         occurrences = getAllOccurrences(matcher);
         if (occurrences.size() == 1) {
             chapter.chapter_number =  occurrences.get(0);
@@ -59,7 +76,7 @@ public class ChapterRecognition {
         String mangaName = replaceIrrelevantCharacters(manga.title);
         String nameWithoutManga = difference(mangaName, name);
         if (!nameWithoutManga.isEmpty()) {
-            matcher = p2.matcher(nameWithoutManga);
+            matcher = uncleanNumber.matcher(nameWithoutManga);
             occurrences = getAllOccurrences(matcher);
             if (occurrences.size() == 1) {
                 chapter.chapter_number =  occurrences.get(0);
@@ -76,7 +93,7 @@ public class ChapterRecognition {
         while (matcher.find()) {
             // Match again to get only numbers from the captured text
             String text = matcher.group();
-            Matcher m = p2.matcher(text);
+            Matcher m = uncleanNumber.matcher(text);
             if (m.find()) {
                 try {
                     Float value = Float.parseFloat(m.group(1));
diff --git a/app/src/test/java/eu/kanade/tachiyomi/ChapterRecognitionTest.java b/app/src/test/java/eu/kanade/tachiyomi/ChapterRecognitionTest.java
index 8626d8bb60..e8e8a7eac1 100644
--- a/app/src/test/java/eu/kanade/tachiyomi/ChapterRecognitionTest.java
+++ b/app/src/test/java/eu/kanade/tachiyomi/ChapterRecognitionTest.java
@@ -141,4 +141,11 @@ public class ChapterRecognitionTest {
         ChapterRecognition.parseChapterNumber(c, randomManga);
         assertThat(c.chapter_number).isEqualTo(11f);
     }
+
+    @Test
+    public void testWithNumberInChapterTitle() {
+        Chapter c = createChapter("Ansatsu Kyoushitsu 099 Present Time - 2nd Hour");
+        ChapterRecognition.parseChapterNumber(c, randomManga);
+        assertThat(c.chapter_number).isEqualTo(99f);
+    }
 }