From 7b07aa842e3cae09d0f6fb70672352c21663032d Mon Sep 17 00:00:00 2001
From: Aryan Vanaik <u7475698@anu.edu.au>
Date: Fri, 27 Oct 2023 13:09:36 +1100
Subject: [PATCH] Added Boyer-Moore string deletion algorithm as a fix for
 issue #286

---
 CHANGELOG.md                                  |  5 +-
 .../jabref/logic/importer/fetcher/IEEE.java   |  4 +-
 .../logic/importer/fetcher/IEEEcleanup.java   | 54 +++++++++++++++++++
 .../logic/importer/fetcher/IEEETest.java      | 10 ++++
 4 files changed, 69 insertions(+), 4 deletions(-)
 create mode 100644 src/main/java/org/jabref/logic/importer/fetcher/IEEEcleanup.java

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 91fc3245e5c..af1b29a374e 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,9 +10,10 @@ Note that this project **does not** adhere to [Semantic Versioning](https://semv
 ## [Unreleased]
 
 ### Added
-
+    - New Class called IEEE cleanup
+    - Added a new testcase called 
 ### Changed
-
+    added cleanup method before setting title field
 ### Fixed
 
 ### Removed
diff --git a/src/main/java/org/jabref/logic/importer/fetcher/IEEE.java b/src/main/java/org/jabref/logic/importer/fetcher/IEEE.java
index 20242693755..d47fa4883d2 100644
--- a/src/main/java/org/jabref/logic/importer/fetcher/IEEE.java
+++ b/src/main/java/org/jabref/logic/importer/fetcher/IEEE.java
@@ -77,7 +77,7 @@ public IEEE(ImportFormatPreferences importFormatPreferences, ImporterPreferences
      */
     private static BibEntry parseJsonResponse(JSONObject jsonEntry, Character keywordSeparator) {
         BibEntry entry = new BibEntry();
-
+        IEEEcleanup ieeEcleanup = new IEEEcleanup();
         switch (jsonEntry.optString("content_type")) {
             case "Books" -> entry.setType(StandardEntryType.Book);
             case "Conferences" -> entry.setType(StandardEntryType.InProceedings);
@@ -129,7 +129,7 @@ private static BibEntry parseJsonResponse(JSONObject jsonEntry, Character keywor
         entry.setField(StandardField.EVENTTITLEADDON, jsonEntry.optString("conference_location"));
         entry.setField(StandardField.EVENTDATE, jsonEntry.optString("conference_dates"));
         entry.setField(StandardField.PUBLISHER, jsonEntry.optString("publisher"));
-        entry.setField(StandardField.TITLE, jsonEntry.optString("title"));
+        entry.setField(StandardField.TITLE, ieeEcleanup.clean(jsonEntry.optString("title"),"{&}{#}x2014$\\mathsemicolon$"));
         entry.setField(StandardField.VOLUME, jsonEntry.optString("volume"));
 
         return entry;
diff --git a/src/main/java/org/jabref/logic/importer/fetcher/IEEEcleanup.java b/src/main/java/org/jabref/logic/importer/fetcher/IEEEcleanup.java
new file mode 100644
index 00000000000..c75068da710
--- /dev/null
+++ b/src/main/java/org/jabref/logic/importer/fetcher/IEEEcleanup.java
@@ -0,0 +1,54 @@
+package org.jabref.logic.importer.fetcher;
+
+public class IEEEcleanup {
+
+    private  void badCharHeuristic( char []str, int size,int badchar[])
+    {
+
+        // Initialize all occurrences as -1
+        for (int i = 0; i < 256; i++){
+            badchar[i] = -1;
+        }
+
+
+        // Fill the actual value of last occurrence
+        // of a character (indices of table are ascii and values are index of occurrence)
+        for (int i = 0; i < size; i++){
+            badchar[(int) str[i]] = i;
+        }
+    }
+
+    /* A pattern searching function that uses Bad
+    Character Heuristic of Boyer Moore Algorithm */
+    public String clean(String str, String t) {
+        int m = t.length();
+        int n = str.length();
+        StringBuilder s = new StringBuilder(str);
+        int[] badChar = new int[256];
+
+        badCharHeuristic(t.toCharArray(), m, badChar);
+
+        int i = 0;
+        while (i <= n - m) {
+            int j = m - 1;
+
+            // Keep reducing the index j of pattern while characters of pattern
+            // and string are matching at this shift s
+            while (j >= 0 && t.charAt(j) == str.charAt(i + j)) {
+                j--;
+            }
+
+            // If the pattern is present at current shift, then remove it
+            if (j < 0) {
+                s.delete(i, i + m);
+                n = s.length();
+                i += m;
+            }
+            else {
+                // Shift the pattern so that the bad character in text aligns with the last occurrence of it in pattern.
+                i += Math.max(1, j - badChar[s.charAt(i + j)]);
+            }
+        }
+        return s.toString();
+    }
+}
diff --git a/src/test/java/org/jabref/logic/importer/fetcher/IEEETest.java b/src/test/java/org/jabref/logic/importer/fetcher/IEEETest.java
index 965fde89a1a..afafa81d019 100644
--- a/src/test/java/org/jabref/logic/importer/fetcher/IEEETest.java
+++ b/src/test/java/org/jabref/logic/importer/fetcher/IEEETest.java
@@ -145,6 +145,16 @@ void searchByQuotedQueryFindsEntry() throws Exception {
         assertEquals(Collections.singletonList(IGOR_NEWCOMERS), fetchedEntries);
     }
 
+    @Test
+    void testEmDashCleanUp() throws Exception{
+        List<BibEntry> entry = fetcher.performSearch("10.1109/PERCOMW.2015.7133989");
+        if (entry.get(0).getTitle().isPresent()){
+            assertEquals("Towards situation-aware adaptive workflows: SitOPT A general purpose situation-aware workflow management system"
+                    ,entry.get(0).getTitle().get());
+        }
+    }
+
+
     @Override
     public SearchBasedFetcher getFetcher() {
         return fetcher;