From f2b9bc8bbbfcf8d64249d852c076a0c511700fe0 Mon Sep 17 00:00:00 2001 From: Greg DiCristofaro Date: Fri, 18 Jun 2021 14:26:09 -0400 Subject: [PATCH] changes to business logic --- .../keywordsearch/KeywordSearchIngestModule.java | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/KeywordSearchIngestModule.java b/KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/KeywordSearchIngestModule.java index 7345c1e774..4fc6f90ff3 100644 --- a/KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/KeywordSearchIngestModule.java +++ b/KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/KeywordSearchIngestModule.java @@ -460,9 +460,16 @@ public final class KeywordSearchIngestModule implements FileIngestModule { * is on. */ private boolean isLimitedOCRFile(AbstractFile aFile, String mimeType) { - return aFile.getSize() > LIMITED_OCR_SIZE_MIN - || aFile.getType() == TskData.TSK_DB_FILES_TYPE_ENUM.DERIVED - || PDF_MIME_TYPE.equals(mimeType); + if (PDF_MIME_TYPE.equals(mimeType)) { + return true; + } + + if (mimeType.startsWith(IMAGE_MIME_TYPE_PREFIX)) { + return aFile.getSize() > LIMITED_OCR_SIZE_MIN + || aFile.getType() == TskData.TSK_DB_FILES_TYPE_ENUM.DERIVED; + } + + return false; } /**