TIKA-4278: colon delimiter detection is unreliable, use the next one …

…if same confidence
apache · Jul 16, 2024 · 6bc7389 · 6bc7389
1 parent ca0ceab
commit 6bc7389
Showing 1 changed file with 5 additions and 0 deletions.
diff --git a/...-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/csv/CSVSniffer.java b/...-modules/tika-parser-text-module/src/main/java/org/apache/tika/parser/csv/CSVSniffer.java
@@ -93,6 +93,11 @@ CSVResult getBest(Reader reader, Metadata metadata) throws IOException {
         if (bestResult.getConfidence() < minConfidence) {
             return CSVResult.TEXT;
         }
+        // TIKA-4278: colon isn't reliable, e.g. govdocs1/242/242970.txt
+        if (results.size() > 1 && bestResult.getDelimiter().equals(':') &&
+                results.get(1).getConfidence() == bestResult.getConfidence()) {
+            return results.get(1);
+        }
         return bestResult;
     }