From 3cd5590f3b64d0b847864e5df23460fbca67aaea Mon Sep 17 00:00:00 2001
From: Giambattista Bloisi <giambattista.bloisi@openaire.eu>
Date: Wed, 28 Feb 2024 15:14:18 +0100
Subject: [PATCH] When converting json to XML, remove characters that are not
 allowed in the XML 1.0 specs, as they will cause xpath failures even if
 escaped

---
 .../dhp/collection/plugin/utils/JsonUtils.java  | 17 +++++++++++++++++
 .../collection/plugin/utils/JsonUtilsTest.java  |  7 +++++++
 2 files changed, 24 insertions(+)

diff --git a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtils.java b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtils.java
index 962c55cfe..1bdc8b138 100644
--- a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtils.java
+++ b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtils.java
@@ -49,6 +49,23 @@ public class JsonUtils {
 				res.put(i, cleanValue(array.opt(i)));
 			}
 			return res;
+		} else if (object instanceof String) {
+			String value = (String) object;
+
+			// XML 1.0 Allowed characters
+			// Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD] | [#x10000-#x10FFFF]
+
+			return value
+				.codePoints()
+				.filter(
+					cp -> cp == 0x9 || cp == 0xA || cp == 0xD || (cp >= 0x20 && cp <= 0xD7FF)
+						|| (cp >= 0xE000 && cp <= 0xFFFD)
+						|| (cp >= 0x10000 && cp <= 0x10FFFF))
+				.collect(
+					StringBuilder::new,
+					StringBuilder::appendCodePoint,
+					StringBuilder::append)
+				.toString();
 		}
 
 		return object;
diff --git a/dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtilsTest.java b/dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtilsTest.java
index 5f340d6d8..59d24fdc2 100644
--- a/dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtilsTest.java
+++ b/dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/utils/JsonUtilsTest.java
@@ -38,4 +38,11 @@ class JsonUtilsTest {
 			wrapped("<parent><id>1</id></parent><parent><id>2</id></parent>"),
 			JsonUtils.convertToXML("{\"parent\": [{\"id\": 1}, {\"id\": 2}]}"));
 	}
+
+	@Test
+	void removeControlCharacters() {
+		assertEquals(
+			wrapped("<m_100><n_200v>Test</n_200v></m_100>"),
+			JsonUtils.convertToXML("{\"100\" : {\"200v\" : \"\\u0000\\u000cTest\"}}"));
+	}
 }