Marmelator/layout-parser
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 289 Wiki Activity

RED-9974: Ignore enoughChars when section identifierer regex matches for documine old

Browse Source
This commit is contained in:
Dominique Eifländer 2024-09-16 12:16:11 +02:00
parent a4d6d2326e
commit 6acc85266c
1 changed files with 4 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java
View File

@ -172,8 +172,7 @@ public class DocuMineClassificationService {
&& isAtLeast3Characters && isAtLeast3Characters
&& !headlineWithSlashesMatches && !headlineWithSlashesMatches
&& !isAmount && !isAmount
&& !isTocItem && !isTocItem) {
&& enoughChars) {
setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes); setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
} else if (!isTocItem } else if (!isTocItem
@ -182,8 +181,7 @@ public class DocuMineClassificationService {
&& PositionUtils.getApproxLineCount(textBlock) < 2.9 && PositionUtils.getApproxLineCount(textBlock) < 2.9
&& (tableOrFigureMatcher.reset().find() || headlineWithSingleIdentifierMatcher.reset().find()) && (tableOrFigureMatcher.reset().find() || headlineWithSingleIdentifierMatcher.reset().find())
&& !isAmount && !isAmount
&& !headlineWithSlashesMatches && !headlineWithSlashesMatches) {
&& enoughChars) {
setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes); setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
} else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)
@ -248,7 +246,8 @@ public class DocuMineClassificationService {
return surroundingBlocks.stream() return surroundingBlocks.stream()
.mapToDouble(surroundingBlock -> calculateSeparation(textBlock, surroundingBlock)) .mapToDouble(surroundingBlock -> calculateSeparation(textBlock, surroundingBlock))
.min().orElse(Double.MAX_VALUE); .min()
.orElse(Double.MAX_VALUE);
} }