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

Fixed RegEx for et al. recommendations

Browse Source
This commit is contained in:
Dominique Eifländer 2021-01-07 13:07:21 +01:00
parent e58b4ff6c1
commit 633fb403e0
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/redaction/model/Section.java
View File

@ -144,7 +144,7 @@ public class Section {
public void redactByRegEx(String pattern, boolean patternCaseInsensitive, int group, String asType, int ruleNumber, String reason, String legalBasis) { public void redactByRegEx(String pattern, boolean patternCaseInsensitive, int group, String asType, int ruleNumber, String reason, String legalBasis) {
Pattern compiledPattern = Patterns.getCompiledPattern(pattern, patternCaseInsensitive); Pattern compiledPattern = Patterns.getCompiledPattern(pattern, patternCaseInsensitive);
Matcher matcher = compiledPattern.matcher(text); Matcher matcher = compiledPattern.matcher(searchText);
while (matcher.find()) { while (matcher.find()) {
String match = matcher.group(group); String match = matcher.group(group);

2
redaction-service-v1/redaction-service-server-v1/src/test/resources/drools/rules.drl
View File

@ -138,7 +138,7 @@ rule "12: Add recommendation for et al. author"
when when
Section(searchText.contains("et al.")) Section(searchText.contains("et al."))
then then
section.addRecommendationByRegEx("([^\\s(]*?( \\w\\.?)?) et al\\.?", false, 1, "CBI_author"); section.addRecommendationByRegEx("\\b([A-ZÄÖÜ][^\\s\\.,]+( [A-ZÄÖÜ]\\.?)?( [A-ZÄÖÜ]\\.?)?) et al\\.?", false, 1, "CBI_author");
end end
// --------------------------------------- PII rules ------------------------------------------------------------------- // --------------------------------------- PII rules -------------------------------------------------------------------