Fix wrong Xpath expression

kermitt2 · May 4, 2024 · 39892ff · 39892ff
1 parent 83c7a10
commit 39892ff
Show file tree

Hide file tree

Showing 3 changed files with 30 additions and 4 deletions.
diff --git a/grobid-core/src/main/java/org/grobid/core/engines/FundingAcknowledgementParser.java b/grobid-core/src/main/java/org/grobid/core/engines/FundingAcknowledgementParser.java
@@ -215,15 +215,15 @@ public MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affili
 //                        offsetPositionList.add(new OffsetPosition(pos, pos + sentenceLayoutToken.size()));
 //                        pos += sentenceLayoutToken.size();
 //                    }
-                    Nodes sentences = paragraph.query("//s");
+                    Nodes sentences = paragraph.query(".//s");
 
                     if(sentences.size() == 0) {
                         // Overly careful - we should never end up here.
                         LOGGER.warn("While the configuration claim that paragraphs must be segmented, we did not find any sentence. ");
                         updateParagraphNodeWithAnnotations(paragraph, annotations);
                     }
 
-                    updateNodes(sentences, annotations);
+                    updateSentencesNodes(sentences, annotations);
                 } else {
                     updateParagraphNodeWithAnnotations(paragraph, annotations);
                 }
@@ -285,7 +285,7 @@ private static void updateParagraphNodeWithAnnotations(Node paragraph, List<Pair
         }
     }
 
-    private static void updateNodes(Nodes sentences, List<Pair<OffsetPosition, Element>> annotations) {
+    private static void updateSentencesNodes(Nodes sentences, List<Pair<OffsetPosition, Element>> annotations) {
         int pos = 0;
         int sentenceStartOffset = 0;
         for (Node sentence : sentences) {

diff --git a/grobid-core/src/main/java/org/grobid/core/utilities/TextUtilities.java b/grobid-core/src/main/java/org/grobid/core/utilities/TextUtilities.java
@@ -1586,7 +1586,7 @@ public static List<OffsetPosition> matchTokenAndString(List<LayoutToken> layoutT
                     if (StringUtils.isNotEmpty(accumulator)) {
                         int start = text.indexOf(accumulator.toString(), pos);
                         newPositions.add(new OffsetPosition(start, start + accumulator.toString().length()));
-                        pos = textPositionOfToken + 1;
+                        pos = textPositionOfToken;
                         break;
                     }
                     pos = textPositionOfToken;

diff --git a/...re/src/test/kotlin/org/grobid/core/engines/FundingAcknowledgementParserIntegrationTest.kt b/...re/src/test/kotlin/org/grobid/core/engines/FundingAcknowledgementParserIntegrationTest.kt
@@ -132,6 +132,32 @@ class FundingAcknowledgementParserIntegrationTest {
         assertThat(element.toXML(), CompareMatcher.isIdenticalTo(output))
     }
 
+    @Test
+    fun testXmlFragmentProcessing_ErrorCase2_withSentenceSegmentation_shouldWork() {
+        val input ="""
+			<div type="acknowledgement">
+<div><head>Acknowledgements</head><p><s>The authors would like to acknowledge Lucy Popplewell in the preparation of EMR notes for this study.</s></p></div>
+<div><head>The authors would like to acknowledge Keele University's Prognosis and Consultation Epidemiology</head><p><s>Research Group who have given us permission to utilise the morbidity definitions (©2014).</s><s>The copyright of the morbidity definitions/categorization lists (©2014) used in this publication is owned by Keele University, the development of which was supported by the Primary Care Research Consortium; For access/details relating to the morbidity definitions/categorisation lists (©2014) please go to www.keele.ac.uk/mrr.</s></p></div>
+			</div>
+
+"""
+
+        val output ="""
+			<div type="acknowledgement">
+<div><head>Acknowledgements</head><p><s>The authors would like to acknowledge <rs xmlns="http://www.tei-c.org/ns/1.0" type="person">Lucy Popplewell</rs> in the preparation of EMR notes for this study.</s></p></div>
+<div><head>The authors would like to acknowledge Keele University's Prognosis and Consultation Epidemiology</head><p><s>Research Group who have given us permission to utilise the morbidity definitions (<rs xmlns="http://www.tei-c.org/ns/1.0" type="grantNumber">©2014</rs>).</s><s>The copyright of the morbidity definitions/categorization lists (<rs xmlns="http://www.tei-c.org/ns/1.0" type="grantNumber">©2014</rs>) used in this publication is owned by <rs xmlns="http://www.tei-c.org/ns/1.0" type="funder">Keele University</rs>, the development of which was supported by the <rs xmlns="http://www.tei-c.org/ns/1.0" type="funder">Primary Care Research Consortium</rs>; For access/details relating to the morbidity definitions/categorisation lists (<rs xmlns="http://www.tei-c.org/ns/1.0" type="grantNumber">©2014</rs>) please go to www.keele.ac.uk/mrr.</s></p></div>
+			</div>
+
+"""
+        val config = GrobidAnalysisConfig.GrobidAnalysisConfigBuilder()
+            .withSentenceSegmentation(true)
+            .build()
+
+        val (element, mutableTriple) = target.processingXmlFragment(input, config)
+
+        assertThat(element.toXML(), CompareMatcher.isIdenticalTo(output))
+    }
+
     companion object {
         @JvmStatic
         @BeforeClass