add smartnames mode for orcid, see issue #8

README.md

@@ -82,6 +82,12 @@ instructions.
   For example, if you have a column containing Scopus EIDs, you can select the "Include?" checkbox
   for it and enter "eid" in the "As Property" box on the reconciliation screen.
 
+* By default, queries are keyword searches on the entire ORCID bios,
+  which can return odd results sometimes. The "smartnames" mode (see
+  the instructions below) splits up names and searches on the
+  given-names and family-name fields specifically; if there are no
+  results, it falls back to a keyword search.
+
 ### Open Library
 
 * Open Library has rate limits on its API, so requests are not run in a
@@ -165,6 +171,12 @@ A docker image created by [tobinski](https://github.com/tobinski) is available h
     http://localhost:8080/reconcile/orcid
     ```
 
+    To use ORCID with "smartnames" mode when reconciliing names:
+
+    ```
+    http://localhost:8080/reconcile/orcid/smartnames
+    ```
+
     To use Open Library: (On the reconciliation screen, under the
     "Also use relevant details from other columns" panel, you can
     check the "Include?" box for columns to include in the query. Give

pom.xml

@@ -4,7 +4,7 @@
 
     <groupId>com.codefork.conciliator</groupId>
     <artifactId>conciliator</artifactId>
-    <version>2.4.0</version>
+    <version>2.5.0</version>
     <packaging>jar</packaging>
 
     <name>conciliator</name>

src/main/java/com/codefork/refine/orcid/Orcid.java

@@ -9,25 +9,47 @@
 import org.apache.commons.logging.LogFactory;
 import org.springframework.web.util.UriUtils;
 
+import javax.servlet.http.HttpServletRequest;
 import javax.xml.parsers.SAXParser;
 import javax.xml.parsers.SAXParserFactory;
 import java.io.IOException;
 import java.io.InputStream;
 import java.net.HttpURLConnection;
 import java.net.URL;
+import java.util.Collections;
+import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
 
 public class Orcid extends WebServiceDataSource {
+    public static String EXTRA_PARAM_MODE_FROM_PATH = "mode";
+    public static String MODE_SMART_NAMES = "smartnames";
+
     Log log = LogFactory.getLog(Orcid.class);
 
     private SAXParserFactory spf = SAXParserFactory.newInstance();
 
     private static String createQueryString(SearchQuery query) {
         StringBuffer buf = new StringBuffer();
         buf.append(query.getQuery());
-        for(Map.Entry<String, PropertyValue> prop : query.getProperties().entrySet()) {
+        String fields = createSearchFieldsQueryString(query);
+        if(fields.length() > 0) {
             buf.append(" ");
+            buf.append(fields);
+        }
+        return buf.toString();
+    }
+
+    /**
+     * creates Solr-style "field:value" query string from properties in SearchQuery
+     */
+    private static String createSearchFieldsQueryString(SearchQuery query) {
+        StringBuffer buf = new StringBuffer();
+        boolean first = true;
+        for(Map.Entry<String, PropertyValue> prop : query.getProperties().entrySet()) {
+            if(!first) {
+                buf.append(" ");
+            }
             buf.append(prop.getKey());
             buf.append(":");
             String val = prop.getValue().asString();
@@ -38,22 +60,69 @@ private static String createQueryString(SearchQuery query) {
             } else {
                 buf.append(val);
             }
+            first = false;
         }
         return buf.toString();
     }
 
     @Override
     public ServiceMetaDataResponse createServiceMetaDataResponse(Map<String, String> extraParams) {
+        if(MODE_SMART_NAMES.equals(extraParams.get(EXTRA_PARAM_MODE_FROM_PATH))) {
+            return new OrcidMetaDataResponse(getName() + " - Smart Names Mode");
+        }
         return new OrcidMetaDataResponse(getName());
     }
 
+    @Override
+    public Map<String, String> parseRequestToExtraParams(HttpServletRequest request) {
+        String[] parts = request.getServletPath().split("/");
+        String dataSourceStr = parts[2];
+        String mode = null;
+        if (parts.length > 3) {
+            mode = parts[3];
+        }
+
+        Map<String, String> extraParams = new HashMap<String, String>();
+        extraParams.put(EXTRA_PARAM_MODE_FROM_PATH, mode);
+        return extraParams;
+    }
+
     @Override
     public List<Result> search(SearchQuery query) throws Exception {
+        List<Result> results = Collections.emptyList();
 
-        String q = createQueryString(query);
+        if(MODE_SMART_NAMES.equals(query.getExtraParams().get(EXTRA_PARAM_MODE_FROM_PATH))) {
+            String name = query.getQuery();
+            String[] namePieces = name.split("\\s+");
+            if(namePieces.length == 2) {
+                results = searchSmartNames(query, namePieces[0], namePieces[1]);
+            }
+        }
+        if(results.isEmpty()) {
+            results = searchKeyword(query);
+        }
+        return results;
+    }
+
+    private List<Result> searchSmartNames(SearchQuery query, String givenName, String familyName) throws Exception {
+        String q = String.format("given-names:%s AND family-name:%s", givenName, familyName);
+        String fields = createSearchFieldsQueryString(query);
+        if(fields.length() > 0) {
+            q += " " + fields;
+        }
+        String url = String.format("http://pub.orcid.org/v1.2/search/orcid-bio/?rows=%d&q=", query.getLimit()) +
+                UriUtils.encodeQueryParam(q, "UTF-8");
+        return doSearch(query, url);
+    }
 
+    private List<Result> searchKeyword(SearchQuery query) throws Exception {
+        String q = createQueryString(query);
         String url = String.format("http://pub.orcid.org/v1.2/search/orcid-bio/?rows=%d&q=", query.getLimit()) +
                 UriUtils.encodeQueryParam(q, "UTF-8");
+        return doSearch(query, url);
+    }
+
+    private List<Result> doSearch(SearchQuery query, String url) throws Exception {
         log.debug("Making request to " + url);
         HttpURLConnection conn = getConnectionFactory().createConnection(url);