commit 98469201ca2816f319f7614014915577c625e0a7
author Marc Kupietz <kupietz@ids-mannheim.de> Sun Mar 10 19:26:51 2024 +0100
committer Marc Kupietz <kupietz@ids-mannheim.de> Mon Mar 11 15:54:06 2024 +0100
tree e1c919eb6784597bcf6198e427e3a00dfb62c696
parent 46aaae8dd0c073b13e50744a68cdf5c93d7c178e

On parse exceptions log warning and skip text

Change-Id: I4721aaa704c779a5bb295dc8f5085907342f8d57

app/src/main/kotlin/de/ids_mannheim/korapxmltools/KorapXml2Conllu.kt

diff --git a/app/src/main/kotlin/de/ids_mannheim/korapxmltools/KorapXml2Conllu.kt b/app/src/main/kotlin/de/ids_mannheim/korapxmltools/KorapXml2Conllu.kt
index 9888f2a..de32dcb 100644
--- a/app/src/main/kotlin/de/ids_mannheim/korapxmltools/KorapXml2Conllu.kt
+++ b/app/src/main/kotlin/de/ids_mannheim/korapxmltools/KorapXml2Conllu.kt
@@ -4,6 +4,7 @@
 import org.w3c.dom.Element
 import org.w3c.dom.NodeList
 import org.xml.sax.InputSource
+import org.xml.sax.SAXParseException
 import picocli.CommandLine
 import picocli.CommandLine.*
 import java.io.File
@@ -224,7 +225,12 @@
                             val inputStream: InputStream = zipFile.getInputStream(zipEntry)
                             val dbFactory: DocumentBuilderFactory = DocumentBuilderFactory.newInstance()
                             val dBuilder: DocumentBuilder = dbFactory.newDocumentBuilder()
-                            val doc: Document = dBuilder.parse(InputSource(InputStreamReader(inputStream, "UTF-8")))
+                            val doc: Document = try {
+                                dBuilder.parse(InputSource(InputStreamReader(inputStream, "UTF-8")))
+                            } catch (e: SAXParseException) {
+                                LOGGER.warning("Error parsing file: " + zipEntry.name + " " + e.message)
+                                return@forEach
+                            }
 
                             doc.documentElement.normalize()
                             val docId: String = doc.documentElement.getAttribute("docid")