From 6acc85266c6253940396188e7328af9a1a8634a9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Dominique=20Eifl=C3=A4nder?= Date: Mon, 16 Sep 2024 12:16:11 +0200 Subject: [PATCH] RED-9974: Ignore enoughChars when section identifierer regex matches for documine old --- .../classification/DocuMineClassificationService.java | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java index e73dc0c..4d18626 100644 --- a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java +++ b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java @@ -172,8 +172,7 @@ public class DocuMineClassificationService { && isAtLeast3Characters && !headlineWithSlashesMatches && !isAmount - && !isTocItem - && enoughChars) { + && !isTocItem) { setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes); } else if (!isTocItem @@ -182,8 +181,7 @@ public class DocuMineClassificationService { && PositionUtils.getApproxLineCount(textBlock) < 2.9 && (tableOrFigureMatcher.reset().find() || headlineWithSingleIdentifierMatcher.reset().find()) && !isAmount - && !headlineWithSlashesMatches - && enoughChars) { + && !headlineWithSlashesMatches) { setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes); } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) @@ -248,7 +246,8 @@ public class DocuMineClassificationService { return surroundingBlocks.stream() .mapToDouble(surroundingBlock -> calculateSeparation(textBlock, surroundingBlock)) - .min().orElse(Double.MAX_VALUE); + .min() + .orElse(Double.MAX_VALUE); }