From 4d6735f7c7d0e9ed674403f311721fdcfa3b0bb5 Mon Sep 17 00:00:00 2001
From: Leslie Bellony <leslie.bellony-ext@bnf.fr>
Date: Tue, 9 Dec 2025 17:33:18 +0100
Subject: [PATCH] Add ExtractorJson from BL (#698)

---
 modules/pom.xml                               |  6 ++
 .../modules/extractor/ExtractorJson.java      | 76 +++++++++++++++++++
 2 files changed, 82 insertions(+)
 create mode 100644 modules/src/main/java/org/archive/modules/extractor/ExtractorJson.java

diff --git a/modules/pom.xml b/modules/pom.xml
index 69ef4e84..62f0fa72 100644
--- a/modules/pom.xml
+++ b/modules/pom.xml
@@ -111,6 +111,12 @@
 			<artifactId>junit-jupiter</artifactId>
 			<optional>true</optional>
 		</dependency>
+		<dependency>
+			<!-- https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-databind -->
+			<groupId>com.fasterxml.jackson.core</groupId>
+			<artifactId>jackson-databind</artifactId>
+			<version>2.20.0</version>
+		</dependency>
 	</dependencies>
 	<build>
 		<plugins>
diff --git a/modules/src/main/java/org/archive/modules/extractor/ExtractorJson.java b/modules/src/main/java/org/archive/modules/extractor/ExtractorJson.java
new file mode 100644
index 00000000..e40a4737
--- /dev/null
+++ b/modules/src/main/java/org/archive/modules/extractor/ExtractorJson.java
@@ -0,0 +1,76 @@
+package org.archive.modules.extractor;
+
+import com.fasterxml.jackson.core.JsonFactory;
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import org.archive.modules.CrawlURI;
+import org.archive.util.UriUtils;
+
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Map;
+import java.util.logging.Level;
+import java.util.logging.Logger;
+
+/**
+ * Extracts URIs from JSON resources.
+ * <p>
+ * n.b. chokes on JSONP, e.g.
+ * <p>
+ * breakingNews({"pollPeriod":30000,"isError":false,"html":""})
+ *
+ * @author rcoram
+ */
+public class ExtractorJson extends ContentExtractor {
+    public final static String JSON_URI = "^https?://[^/]+/.+\\.json\\b.*$";
+    private static final Logger LOGGER = Logger.getLogger(ExtractorJson.class.getName());
+    private final JsonFactory factory = new JsonFactory();
+    private final ObjectMapper mapper = new ObjectMapper(factory);
+
+    @Override
+    protected boolean innerExtract(CrawlURI curi) {
+        try {
+            List<String> links = new ArrayList<>();
+            JsonNode rootNode = mapper.readTree(curi.getRecorder().getContentReplayInputStream());
+            parse(rootNode, links);
+            for (String link : links) {
+                try {
+                    int max = getExtractorParameters().getMaxOutlinks();
+                    addRelativeToBase(curi, max, link, LinkContext.INFERRED_MISC, Hop.INFERRED);
+                    numberOfLinksExtracted.incrementAndGet();
+                } catch (org.archive.url.URIException e) {
+                    logUriError(e, curi.getUURI(), link);
+                }
+            }
+        } catch (Exception e) {
+            // Only record this as INFO, as malformed JSON is fairly common.
+            LOGGER.log(Level.INFO, curi.getURI() + " : " + e.getMessage());
+        }
+        return false;
+    }
+
+    @Override
+    protected boolean shouldExtract(CrawlURI curi) {
+        String contentType = curi.getContentType();
+        if (contentType != null && contentType.contains("json")) {
+            return true;
+        }
+        return curi.isSuccess() && curi.toString().matches(JSON_URI);
+    }
+
+    protected List<String> parse(JsonNode rootNode, List<String> links) {
+        for (Map.Entry<String, JsonNode> field : rootNode.properties()) {
+            if (field.getValue().textValue() != null
+                    && UriUtils.isVeryLikelyUri(field.getValue().textValue())) {
+                links.add(field.getValue().textValue());
+            } else if (field.getValue().isObject()) {
+                parse(field.getValue(), links);
+            } else if (field.getValue().isArray()) {
+                field.getValue()
+                        .propertyStream()
+                        .forEach(fieldValue -> parse(fieldValue.getValue(), links));
+            }
+        }
+        return links;
+    }
+}
\ No newline at end of file