From 541a55798c075baf0b75aa7da538fcfddf4dd3a4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Dominique=20Eifl=C3=A4nder?= Date: Wed, 2 Jun 2021 09:41:22 +0200 Subject: [PATCH] Fixed text parsing orientation problem --- .../classification/service/BlockificationService.java | 11 ++++++++++- .../src/test/resources/dictionaries/CBI_address.txt | 1 + 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/classification/service/BlockificationService.java b/redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/classification/service/BlockificationService.java index d394bbe6..abe13409 100644 --- a/redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/classification/service/BlockificationService.java +++ b/redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/classification/service/BlockificationService.java @@ -49,6 +49,11 @@ public class BlockificationService { if (prev != null && (lineSeparation || startFromTop || splitByX || newLineAfterSplit || splittedByRuling)) { + Orientation prevOrientation = null; + if(!chunkBlockList1.isEmpty()) { + prevOrientation = chunkBlockList1.get(chunkBlockList1.size() - 1).getOrientation(); + } + TextBlock cb1 = buildTextBlock(chunkWords); chunkBlockList1.add(cb1); chunkWords = new ArrayList<>(); @@ -57,12 +62,16 @@ public class BlockificationService { wasSplitted = true; cb1.setOrientation(Orientation.LEFT); splitX1 = word.getX1(); - } + } else if (newLineAfterSplit && !splittedByRuling) { wasSplitted = false; cb1.setOrientation(Orientation.RIGHT); splitX1 = null; + } else + + if(prevOrientation != null && prevOrientation.equals(Orientation.RIGHT) && (lineSeparation || !startFromTop || !splitByX || !newLineAfterSplit || !splittedByRuling)){ + cb1.setOrientation(Orientation.LEFT); } minX = 1000; diff --git a/redaction-service-v1/redaction-service-server-v1/src/test/resources/dictionaries/CBI_address.txt b/redaction-service-v1/redaction-service-server-v1/src/test/resources/dictionaries/CBI_address.txt index 8de2feb7..b6ad4398 100644 --- a/redaction-service-v1/redaction-service-server-v1/src/test/resources/dictionaries/CBI_address.txt +++ b/redaction-service-v1/redaction-service-server-v1/src/test/resources/dictionaries/CBI_address.txt @@ -1653,3 +1653,4 @@ Zoecon Corp., Palo Alto, USA Zyma SA Zyma SA, Nyon, Switzerland Mambo-Tox Ltd. Biomedical Sciences Building Bassett Crescent East Southampton SO16 7PX UK +Syngenta Environmental Sciences Jealott’s Hill International Research Centre Bracknell, Berkshire RG42 6EY UK