Merge pull request #428 from r0ann3l/NUTCH-2686
NUTCH-2686 New property: "moreIndexingFilter.mapMimeTypes.field"
diff --git a/conf/host-protocol-mapping.txt.template b/conf/host-protocol-mapping.txt.template
new file mode 100644
index 0000000..a09bca6
--- /dev/null
+++ b/conf/host-protocol-mapping.txt.template
@@ -0,0 +1,16 @@
+# This file defines a hostname to protocol plugin mapping. Each line takes a
+# host name followed by a tab, followed by the ID of the protocol plugin. You
+# can find the ID in the protocol plugin's plugin.xml file.
+#
+# <hostname>\t<plugin_id>\n
+# nutch.apache.org org.apache.nutch.protocol.httpclient.Http
+# tika.apache.org org.apache.nutch.protocol.http.Http
+#
+# If the requested host is not mapped, Nutch can choose any of the enabled
+# plugins so you can force defaults using:
+#
+# protocol:<protocol>\t<plugin_id>\n
+#
+# This example forces httpclient for all protocol in case the host is not mapped:
+# protocol:http org.apache.nutch.protocol.httpclient.Http
+# protocol:https org.apache.nutch.protocol.httpclient.Http
diff --git a/conf/nutch-default.xml b/conf/nutch-default.xml
index 3e5cede..a42e6a9 100644
--- a/conf/nutch-default.xml
+++ b/conf/nutch-default.xml
@@ -1369,11 +1369,11 @@
<value>protocol-http|urlfilter-(regex|validator)|parse-(html|tika)|index-(basic|anchor)|indexer-solr|scoring-opic|urlnormalizer-(pass|regex|basic)</value>
<description>Regular expression naming plugin directory names to
include. Any plugin not matching this expression is excluded.
- In any case you need at least include the nutch-extensionpoints plugin. By
- default Nutch includes crawling just HTML and plain text via HTTP,
- and basic indexing and search plugins. In order to use HTTPS please enable
- protocol-httpclient, but be aware of possible intermittent problems with the
- underlying commons-httpclient library. Set parsefilter-naivebayes for classification based focused crawler.
+ By default Nutch includes plugins to crawl HTML and various other
+ document formats via HTTP/HTTPS and indexing the crawled content
+ into Solr. More plugins are available to support more indexing
+ backends, to fetch ftp:// and file:// URLs, for focused crawling,
+ and many other use cases.
</description>
</property>
diff --git a/src/java/org/apache/nutch/plugin/Extension.java b/src/java/org/apache/nutch/plugin/Extension.java
index e73b850..be737cb 100644
--- a/src/java/org/apache/nutch/plugin/Extension.java
+++ b/src/java/org/apache/nutch/plugin/Extension.java
@@ -197,4 +197,8 @@
public void setDescriptor(PluginDescriptor pDescriptor) {
fDescriptor = pDescriptor;
}
+
+ public String toString() {
+ return getId() + ", " + getClazz() + ", " + getTargetPoint();
+ }
}
diff --git a/src/java/org/apache/nutch/protocol/ProtocolFactory.java b/src/java/org/apache/nutch/protocol/ProtocolFactory.java
index 2d20ecd..7f900b2 100644
--- a/src/java/org/apache/nutch/protocol/ProtocolFactory.java
+++ b/src/java/org/apache/nutch/protocol/ProtocolFactory.java
@@ -17,8 +17,13 @@
package org.apache.nutch.protocol;
+import java.io.BufferedReader;
+import java.io.IOException;
+import java.lang.invoke.MethodHandles;
import java.net.URL;
import java.net.MalformedURLException;
+import java.util.HashMap;
+import java.util.Map;
import org.apache.nutch.plugin.Extension;
import org.apache.nutch.plugin.ExtensionPoint;
@@ -26,8 +31,13 @@
import org.apache.nutch.plugin.PluginRuntimeException;
import org.apache.nutch.util.ObjectCache;
+import org.apache.commons.lang.StringUtils;
+
import org.apache.hadoop.conf.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
/**
* Creates and caches {@link Protocol} plugins. Protocol plugins should define
* the attribute "protocolName" with the name of the protocol that they
@@ -37,10 +47,16 @@
*/
public class ProtocolFactory {
+ private static final Logger LOG = LoggerFactory
+ .getLogger(MethodHandles.lookup().lookupClass());
+
private ExtensionPoint extensionPoint;
private Configuration conf;
+ protected Map<String, String> defaultProtocolImplMapping = new HashMap<>();
+ protected Map<String, String> hostProtocolMapping = new HashMap<>();
+
public ProtocolFactory(Configuration conf) {
this.conf = conf;
this.extensionPoint = PluginRepository.get(conf).getExtensionPoint(
@@ -49,8 +65,35 @@
throw new RuntimeException("x-point " + Protocol.X_POINT_ID
+ " not found.");
}
- }
+ try {
+ BufferedReader reader = new BufferedReader(conf.getConfResourceAsReader("host-protocol-mapping.txt"));
+ String line;
+ String parts[];
+ while ((line = reader.readLine()) != null) {
+ if (StringUtils.isNotBlank(line) && !line.startsWith("#")) {
+ line = line.trim();
+ parts = line.split("\t");
+
+ // Must be at least two parts
+ if (parts.length == 2) {
+ // Is this a host to plugin mapping, or a default?
+ if (parts[0].indexOf(":") == -1) {
+ hostProtocolMapping.put(parts[0].trim(), parts[1].trim());
+ } else {
+ String[] moreParts = parts[0].split(":");
+ defaultProtocolImplMapping.put(moreParts[1].trim(), parts[1].trim());
+ }
+ } else {
+ LOG.warn("Wrong format of line: {}", line);
+ LOG.warn("Expected format: <hostname> <tab> <plugin_id> or protocol:<protocol> <tab> <plugin_id>");
+ }
+ }
+ }
+ } catch (IOException e) {
+ LOG.error("Unable to read host-protocol-mapping.txt", e);
+ }
+ }
/**
* Returns the appropriate {@link Protocol} implementation for a url.
*
@@ -83,52 +126,92 @@
*/
public Protocol getProtocol(URL url)
throws ProtocolNotFound {
- ObjectCache objectCache = ObjectCache.get(conf);
try {
- String protocolName = url.getProtocol();
- if (protocolName == null) {
- throw new ProtocolNotFound(url.toString());
+ Protocol protocol = null;
+
+ // First attempt to resolve a protocol implementation by hostname
+ String host = url.getHost();
+ if (hostProtocolMapping.containsKey(host)) {
+ Extension extension = getExtensionById(hostProtocolMapping.get(host));
+ if (extension != null) {
+ protocol = getProtocolInstanceByExtension(extension);
+ }
}
- String cacheId = Protocol.X_POINT_ID + protocolName;
- synchronized (objectCache) {
- Protocol protocol = (Protocol) objectCache.getObject(cacheId);
- if (protocol != null) {
- return protocol;
+ // Nothing, see if we have defaults configured
+ if (protocol == null) {
+ // Protocol listed in default map?
+ if (defaultProtocolImplMapping.containsKey(url.getProtocol())) {
+ Extension extension = getExtensionById(defaultProtocolImplMapping.get(url.getProtocol()));
+ if (extension != null) {
+ protocol = getProtocolInstanceByExtension(extension);
+ }
}
+ }
- Extension extension = findExtension(protocolName);
- if (extension == null) {
- throw new ProtocolNotFound(protocolName);
+ // Still couldn't find a protocol? Attempt by protocol
+ if (protocol == null) {
+ Extension extension = findExtension(url.getProtocol(), "protocolName");
+ if (extension != null) {
+ protocol = getProtocolInstanceByExtension(extension);
}
+ }
- protocol = (Protocol) extension.getExtensionInstance();
- objectCache.setObject(cacheId, protocol);
+ // Got anything?
+ if (protocol != null) {
return protocol;
}
+
+ // Nothing!
+ throw new ProtocolNotFound(url.toString());
} catch (PluginRuntimeException e) {
throw new ProtocolNotFound(url.toString(), e.toString());
}
}
- private Extension findExtension(String name) throws PluginRuntimeException {
+ private Protocol getProtocolInstanceByExtension(Extension extension) throws PluginRuntimeException {
+ Protocol protocol = null;
+ String cacheId = extension.getId();
+ ObjectCache objectCache = ObjectCache.get(conf);
+ synchronized (objectCache) {
+ if (!objectCache.hasObject(cacheId)) {
+ protocol = (Protocol) extension.getExtensionInstance();
+ objectCache.setObject(cacheId, protocol);
+ }
+ protocol = (Protocol) objectCache.getObject(cacheId);
+ }
+ return protocol;
+ }
+
+ private Extension getExtensionById(String id) {
Extension[] extensions = this.extensionPoint.getExtensions();
-
for (int i = 0; i < extensions.length; i++) {
- Extension extension = extensions[i];
+ if (id.equals(extensions[i].getId())) {
+ return extensions[i];
+ }
+ }
- if (contains(name, extension.getAttribute("protocolName")))
+ return null;
+ }
+
+ private Extension findExtension(String name, String attribute) throws PluginRuntimeException {
+ for (int i = 0; i < this.extensionPoint.getExtensions().length; i++) {
+ Extension extension = this.extensionPoint.getExtensions()[i];
+
+ if (contains(name, extension.getAttribute(attribute)))
return extension;
}
return null;
}
boolean contains(String what, String where) {
- String parts[] = where.split("[, ]");
- for (int i = 0; i < parts.length; i++) {
- if (parts[i].equals(what))
- return true;
+ if (where != null) {
+ String parts[] = where.split("[, ]");
+ for (int i = 0; i < parts.length; i++) {
+ if (parts[i].equals(what))
+ return true;
+ }
}
return false;
}
diff --git a/src/java/org/apache/nutch/util/JexlUtil.java b/src/java/org/apache/nutch/util/JexlUtil.java
index f5e67cc..b480033 100644
--- a/src/java/org/apache/nutch/util/JexlUtil.java
+++ b/src/java/org/apache/nutch/util/JexlUtil.java
@@ -29,45 +29,47 @@
import org.slf4j.LoggerFactory;
/**
- * A collection of Jexl utilit(y|ies).
+ * Utility methods for handling JEXL expressions
*/
public class JexlUtil {
private static final Logger LOG = LoggerFactory
.getLogger(MethodHandles.lookup().lookupClass());
- /**
- *
- */
- public static Pattern datePattern = Pattern.compile("\\d{4}-\\d{2}-\\d{2}T\\d{2}:\\d{2}:\\d{2}Z");
+ /** Supported format for date parsing yyyy-MM-ddTHH:mm:ssZ */
+ private static final Pattern DATE_PATTERN = Pattern.compile("\\d{4}-\\d{2}-\\d{2}T\\d{2}:\\d{2}:\\d{2}Z");
/**
- * Parses the given experssion to a Jexl expression. This supports
+ * Parses the given expression to a JEXL expression. This supports
* date parsing.
*
- * @param expr the Jexl expression
- * @return parsed Jexl expression or null in case of parse error
+ * @param expr string JEXL expression
+ * @return parsed JEXL expression or null in case of parse error
*/
public static Expression parseExpression(String expr) {
if (expr == null) return null;
try {
- // Translate any date object into a long, dates must be specified as 20-03-2016T00:00:00Z
- Matcher matcher = datePattern.matcher(expr);
+ // Translate any date object into a long. Dates must be in the DATE_PATTERN
+ // format. For example: 2016-03-20T00:00:00Z
+ Matcher matcher = DATE_PATTERN.matcher(expr);
+
if (matcher.find()) {
String date = matcher.group();
- // Parse the thing and get epoch!
+ // parse the matched substring and get the epoch
Date parsedDate = DateUtils.parseDateStrictly(date, new String[] {"yyyy-MM-dd'T'HH:mm:ss'Z'"});
long time = parsedDate.getTime();
- // Replace in the original expression
+ // replace the original string date with the numeric value
expr = expr.replace(date, Long.toString(time));
}
-
+
JexlEngine jexl = new JexlEngine();
+
jexl.setSilent(true);
jexl.setStrict(true);
+
return jexl.createExpression(expr);
} catch (Exception e) {
LOG.error(e.getMessage());
diff --git a/src/java/org/apache/nutch/util/ObjectCache.java b/src/java/org/apache/nutch/util/ObjectCache.java
index 4ed3fd0..f1b14c8 100644
--- a/src/java/org/apache/nutch/util/ObjectCache.java
+++ b/src/java/org/apache/nutch/util/ObjectCache.java
@@ -52,6 +52,10 @@
return objectMap.get(key);
}
+ public boolean hasObject(String key) {
+ return objectMap.containsKey(key);
+ }
+
public synchronized void setObject(String key, Object value) {
objectMap.put(key, value);
}
diff --git a/src/plugin/index-jexl-filter/src/java/org/apache/nutch/indexer/jexl/JexlIndexingFilter.java b/src/plugin/index-jexl-filter/src/java/org/apache/nutch/indexer/jexl/JexlIndexingFilter.java
index 3fa2294..890020a 100644
--- a/src/plugin/index-jexl-filter/src/java/org/apache/nutch/indexer/jexl/JexlIndexingFilter.java
+++ b/src/plugin/index-jexl-filter/src/java/org/apache/nutch/indexer/jexl/JexlIndexingFilter.java
@@ -18,6 +18,7 @@
package org.apache.nutch.indexer.jexl;
import java.lang.invoke.MethodHandles;
+import java.util.List;
import java.util.Map.Entry;
import org.apache.commons.jexl2.Expression;
@@ -41,7 +42,6 @@
/**
* An {@link org.apache.nutch.indexer.IndexingFilter} that allows filtering of
* documents based on a JEXL expression.
- *
*/
public class JexlIndexingFilter implements IndexingFilter {
@@ -84,9 +84,12 @@
metadataToContext(parse.getData().getParseMeta()));
JexlContext context = new MapContext();
+
for (Entry<String, NutchField> entry : doc) {
- context.set(entry.getKey(), entry.getValue().getValues());
+ List<Object> values = entry.getValue().getValues();
+ context.set(entry.getKey(), values.size() > 1 ? values : values.get(0));
}
+
jcontext.set("doc", context);
try {
@@ -102,16 +105,21 @@
@Override
public void setConf(Configuration conf) {
this.conf = conf;
- String str = conf.get("index.jexl.filter");
- if (str == null) {
- LOG.warn(
+ String strExpr = conf.get("index.jexl.filter");
+
+ if (strExpr == null) {
+ LOG.error(
"The property index.jexl.filter must have a value when index-jexl-filter is used. You can use 'true' or 'false' to index all/none");
+
throw new RuntimeException(
"The property index.jexl.filter must have a value when index-jexl-filter is used. You can use 'true' or 'false' to index all/none");
}
- expr = JexlUtil.parseExpression(str);
+
+ expr = JexlUtil.parseExpression(strExpr);
+
if (expr == null) {
- LOG.warn("Failed parsing JEXL from index.jexl.filter: {}", str);
+ LOG.error("Failed parsing JEXL from index.jexl.filter: {}", strExpr);
+
throw new RuntimeException("Failed parsing JEXL from index.jexl.filter");
}
}
@@ -123,9 +131,12 @@
private JexlContext metadataToContext(Metadata metadata) {
JexlContext context = new MapContext();
+
for (String name : metadata.names()) {
- context.set(name, metadata.getValues(name));
+ String[] values = metadata.getValues(name);
+ context.set(name, values.length > 1 ? values : values[0]);
}
+
return context;
}
}
diff --git a/src/plugin/index-jexl-filter/src/test/org/apache/nutch/indexer/jexl/TestJexlIndexingFilter.java b/src/plugin/index-jexl-filter/src/test/org/apache/nutch/indexer/jexl/TestJexlIndexingFilter.java
index 0427ad4..f3cc655 100644
--- a/src/plugin/index-jexl-filter/src/test/org/apache/nutch/indexer/jexl/TestJexlIndexingFilter.java
+++ b/src/plugin/index-jexl-filter/src/test/org/apache/nutch/indexer/jexl/TestJexlIndexingFilter.java
@@ -39,7 +39,7 @@
@Test
public void testAllowMatchingDocument() throws Exception {
Configuration conf = NutchConfiguration.create();
- conf.set("index.jexl.filter", "doc.lang[0]=='en'");
+ conf.set("index.jexl.filter", "doc.lang=='en'");
JexlIndexingFilter filter = new JexlIndexingFilter();
filter.setConf(conf);
@@ -73,7 +73,7 @@
@Test
public void testBlockNotMatchingDocuments() throws Exception {
Configuration conf = NutchConfiguration.create();
- conf.set("index.jexl.filter", "doc.lang[0]=='en'");
+ conf.set("index.jexl.filter", "doc.lang=='en'");
JexlIndexingFilter filter = new JexlIndexingFilter();
filter.setConf(conf);
@@ -115,7 +115,7 @@
@Test
public void testInvalidExpression() throws Exception {
Configuration conf = NutchConfiguration.create();
- conf.set("index.jexl.filter", "doc.lang[0]=<>:='en'");
+ conf.set("index.jexl.filter", "doc.lang=<>:='en'");
JexlIndexingFilter filter = new JexlIndexingFilter();
thrown.expect(RuntimeException.class);
diff --git a/src/plugin/index-more/src/java/org/apache/nutch/indexer/more/MoreIndexingFilter.java b/src/plugin/index-more/src/java/org/apache/nutch/indexer/more/MoreIndexingFilter.java
index 81fbb22..05c215f 100644
--- a/src/plugin/index-more/src/java/org/apache/nutch/indexer/more/MoreIndexingFilter.java
+++ b/src/plugin/index-more/src/java/org/apache/nutch/indexer/more/MoreIndexingFilter.java
@@ -277,7 +277,7 @@
static {
try {
// order here is important
- patterns[0] = Pattern.compile("\\bfilename=['\"](.+)['\"]");
+ patterns[0] = Pattern.compile("\\bfilename=['\"]([^\"]+)");
patterns[1] = Pattern.compile("\\bfilename=(\\S+)\\b");
} catch (PatternSyntaxException e) {
// just ignore
diff --git a/src/test/org/apache/nutch/protocol/TestProtocolFactory.java b/src/test/org/apache/nutch/protocol/TestProtocolFactory.java
index 394c303..7cab623 100644
--- a/src/test/org/apache/nutch/protocol/TestProtocolFactory.java
+++ b/src/test/org/apache/nutch/protocol/TestProtocolFactory.java
@@ -59,12 +59,6 @@
Assert.fail("Must not throw any other exception");
}
- // cache key
- Object protocol = ObjectCache.get(conf).getObject(
- Protocol.X_POINT_ID + "http");
- Assert.assertNotNull(protocol);
- Assert.assertEquals(httpProtocol, protocol);
-
// test same object instance
try {
Assert.assertTrue(httpProtocol == factory.getProtocol("http://somehost"));