Merge remote-tracking branch 'upstream/main' into downgradeRunnerImage

* upstream/main:
  Refactor the isUnwantedText (JabRef#12369)
  Searching for entries with empty field (JabRef#12376)
  Downgrade Ubuntu (JabRef#12375)

src/main/java/org/jabref/logic/importer/fileformat/PdfContentImporter.java

@@ -252,16 +252,24 @@ private boolean isFarAway(TextPosition previous, TextPosition current) {
             return Math.abs(Xgap) > XspaceThreshold && Math.abs(Ygap) > YspaceThreshold;
         }
 
-        private boolean isUnwantedText(TextPosition previousTextPosition, TextPosition textPosition) {
+        private boolean isUnwantedText(TextPosition previousTextPosition, TextPosition textPosition,
+                                       Map<Float, TextPosition> lastPositionMap, float fontSize) {
+            // This indicates that the text is at the start of the line, so it is needed.
             if (textPosition == null || previousTextPosition == null) {
                 return false;
             }
+            // We use the font size to identify titles. Blank characters don't have a font size, so we discard them.
+            // The space will be added back in the final result, but not in this method.
             if (StringUtil.isBlank(textPosition.getUnicode())) {
                 return true;
             }
-            // The title usually don't in the bottom 10% of a page.
-            return (textPosition.getPageHeight() - textPosition.getYDirAdj())
-                    < (textPosition.getPageHeight() * 0.1);
+            // Titles are generally not located in the bottom 10% of a page.
+            if ((textPosition.getPageHeight() - textPosition.getYDirAdj()) < (textPosition.getPageHeight() * 0.1)) {
+                return true;
+            }
+            // Characters in a title typically remain close together,
+            // so a distant character is unlikely to be part of the title.
+            return lastPositionMap.containsKey(fontSize) && isFarAway(lastPositionMap.get(fontSize), textPosition);
         }
 
         private Optional<String> findLargestFontText(List<TextPosition> textPositions) {
@@ -271,8 +279,7 @@ private Optional<String> findLargestFontText(List<TextPosition> textPositions) {
             for (TextPosition textPosition : textPositions) {
                 float fontSize = textPosition.getFontSizeInPt();
                 // Exclude unwanted text based on heuristics
-                if (isUnwantedText(previousTextPosition, textPosition) ||
-                    (lastPositionMap.containsKey(fontSize) && isFarAway(lastPositionMap.get(fontSize), textPosition))) {
+                if (isUnwantedText(previousTextPosition, textPosition, lastPositionMap, fontSize)) {
                     continue;
                 }
                 fontSizeTextMap.putIfAbsent(fontSize, new StringBuilder());

src/main/java/org/jabref/logic/search/query/SearchToSqlVisitor.java

@@ -207,6 +207,16 @@ public SqlQueryNode visitComparison(SearchParser.ComparisonContext ctx) {
             setFlags(searchFlags, REGULAR_EXPRESSION, true, true);
         }
 
+        // field = "" -> should find entries where the field is empty
+        // field != "" -> should find entries where the field is not empty
+        if (term.isEmpty()) {
+            if (searchFlags.contains(NEGATION)) {
+                searchFlags.remove(NEGATION);
+            } else {
+                searchFlags.add(NEGATION);
+            }
+        }
+
         return getFieldQueryNode(field.toLowerCase(Locale.ROOT), term, searchFlags);
     }
 

src/test/java/org/jabref/logic/search/query/SearchQuerySQLConversionTest.java

@@ -695,6 +695,38 @@ cte0 AS (
                             )
                         )
                         SELECT * FROM cte0 GROUP BY entryid"""
+                ),
+
+                Arguments.of(
+                        "file = \"\"",
+                        """
+                        WITH
+                        cte0 AS (
+                            SELECT main_table.entryid
+                            FROM bib_fields."tableName" AS main_table
+                            WHERE main_table.entryid NOT IN (
+                                SELECT inner_table.entryid
+                                FROM bib_fields."tableName" AS inner_table
+                                WHERE (
+                                    (inner_table.field_name = 'file') AND ((inner_table.field_value_literal ILIKE ('%%')) OR (inner_table.field_value_transformed ILIKE ('%%')))
+                                )
+                            )
+                        )
+                        SELECT * FROM cte0 GROUP BY entryid"""
+                ),
+
+                Arguments.of(
+                        "file != \"\"",
+                        """
+                        WITH
+                        cte0 AS (
+                            SELECT main_table.entryid
+                            FROM bib_fields."tableName" AS main_table
+                            WHERE (
+                                (main_table.field_name = 'file') AND ((main_table.field_value_literal ILIKE ('%%')) OR (main_table.field_value_transformed ILIKE ('%%')))
+                            )
+                        )
+                        SELECT * FROM cte0 GROUP BY entryid"""
                 )
         );
     }