Marmelator/redaction-service
1
0
Fork 0
Code Issues Pull Requests 14 Actions Packages Projects Releases 622 Wiki Activity

Fixed text parsing orientation problem

Browse Source
This commit is contained in:
Dominique Eifländer 2021-06-02 09:41:22 +02:00
parent 5a9b86ba76
commit 541a55798c
2 changed files with 11 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/classification/service/BlockificationService.java
View File

@ -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;

1
redaction-service-v1/redaction-service-server-v1/src/test/resources/dictionaries/CBI_address.txt
View File

@ -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