Add shorten DOI field formatter (koppor#343) (#5276)

JabRef · Sep 8, 2019 · 164843b · 164843b
1 parent dde32bd
commit 164843b
Show file tree

Hide file tree

Showing 10 changed files with 372 additions and 25 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -13,6 +13,8 @@ We refer to [GitHub issues](https://github.com/JabRef/jabref/issues) by using `#
 
 ### Changed
 
+- We added a short DOI field formatter which shortens DOI to more human readable form. [koppor#343](https://github.com/koppor/jabref/issues/343)
+
 ### Fixed
 
 ### Removed

diff --git a/src/main/java/org/jabref/logic/formatter/Formatters.java b/src/main/java/org/jabref/logic/formatter/Formatters.java
@@ -19,6 +19,7 @@
 import org.jabref.logic.formatter.bibtexfields.OrdinalsToSuperscriptFormatter;
 import org.jabref.logic.formatter.bibtexfields.RegexFormatter;
 import org.jabref.logic.formatter.bibtexfields.RemoveBracesFormatter;
+import org.jabref.logic.formatter.bibtexfields.ShortenDOIFormatter;
 import org.jabref.logic.formatter.bibtexfields.UnicodeToLatexFormatter;
 import org.jabref.logic.formatter.bibtexfields.UnitsToLatexFormatter;
 import org.jabref.logic.formatter.casechanger.CapitalizeFormatter;
@@ -67,7 +68,8 @@ public static List<Formatter> getOthers() {
                 new OrdinalsToSuperscriptFormatter(),
                 new RemoveBracesFormatter(),
                 new UnitsToLatexFormatter(),
-                new EscapeUnderscoresFormatter()
+                new EscapeUnderscoresFormatter(),
+                new ShortenDOIFormatter()
         );
     }
 
@@ -102,5 +104,4 @@ public static Optional<Formatter> getFormatterForModifier(String modifier) {
             return getAll().stream().filter(f -> f.getKey().equals(modifier)).findAny();
         }
     }
-
 }
diff --git a/src/main/java/org/jabref/logic/formatter/bibtexfields/ShortenDOIFormatter.java b/src/main/java/org/jabref/logic/formatter/bibtexfields/ShortenDOIFormatter.java
@@ -0,0 +1,59 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import java.util.Objects;
+import java.util.Optional;
+
+import org.jabref.logic.importer.util.ShortDOIService;
+import org.jabref.logic.importer.util.ShortDOIServiceException;
+import org.jabref.logic.l10n.Localization;
+import org.jabref.model.cleanup.Formatter;
+import org.jabref.model.entry.identifier.DOI;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+public class ShortenDOIFormatter extends Formatter {
+
+    private static final Logger LOGGER = LoggerFactory.getLogger(ShortenDOIFormatter.class);
+
+    @Override
+    public String getName() {
+        return Localization.lang("Shorten DOI");
+    }
+
+    @Override
+    public String getKey() {
+        return "short_doi";
+    }
+
+    @Override
+    public String format(String value) {
+        Objects.requireNonNull(value);
+
+        ShortDOIService shortDOIService = new ShortDOIService();
+
+        Optional<DOI> doi = Optional.empty();
+
+        try {
+            doi = DOI.parse(value);
+
+            if (doi.isPresent()) {
+                return shortDOIService.getShortDOI(doi.get()).getDOI();
+            }
+        } catch (ShortDOIServiceException e) {
+            LOGGER.error(e.getMessage(), e);
+        }
+
+        return value;
+    }
+
+    @Override
+    public String getDescription() {
+        return Localization.lang("Shortens DOI to more human readable form.");
+    }
+
+    @Override
+    public String getExampleInput() {
+        return "10.1006/jmbi.1998.2354";
+    }
+}
diff --git a/src/main/java/org/jabref/logic/importer/util/ShortDOIService.java b/src/main/java/org/jabref/logic/importer/util/ShortDOIService.java
@@ -0,0 +1,63 @@
+package org.jabref.logic.importer.util;
+
+import java.io.IOException;
+import java.net.MalformedURLException;
+import java.net.URI;
+import java.net.URISyntaxException;
+import java.net.URL;
+
+import org.jabref.logic.importer.ParseException;
+import org.jabref.logic.net.URLDownload;
+import org.jabref.model.entry.identifier.DOI;
+
+import org.apache.http.client.utils.URIBuilder;
+import org.json.JSONException;
+import org.json.JSONObject;
+
+/**
+ * Class for obtaining shortened DOI names.
+ *
+ * @see http://shortdoi.org
+ */
+public class ShortDOIService {
+
+    private static final String BASIC_URL = "http://shortdoi.org/";
+
+    /**
+     * Obtains shortened DOI name for given DOI
+     *
+     * @param doi DOI
+     * @return A shortened DOI name
+     */
+    public DOI getShortDOI(DOI doi) throws ShortDOIServiceException {
+        JSONObject responseJSON = makeRequest(doi);
+        String shortDoi = responseJSON.getString("ShortDOI");
+
+        return new DOI(shortDoi);
+    }
+
+    private JSONObject makeRequest(DOI doi) throws ShortDOIServiceException {
+
+        URIBuilder uriBuilder = null;
+        URL url = null;
+
+        try {
+            uriBuilder = new URIBuilder(BASIC_URL);
+            uriBuilder.setPath(uriBuilder.getPath() + doi.getDOI());
+            uriBuilder.addParameter("format", "json");
+
+            URI uri = uriBuilder.build();
+            url = uri.toURL();
+        } catch (URISyntaxException | MalformedURLException e) {
+            throw new ShortDOIServiceException("Cannot get short DOI", e);
+        }
+
+        URLDownload urlDownload = new URLDownload(url);
+
+        try {
+            return JsonReader.toJsonObject(urlDownload.asInputStream());
+        } catch (ParseException | IOException | JSONException e) {
+            throw new ShortDOIServiceException("Cannot get short DOI", e);
+        }
+    }
+}
diff --git a/src/main/java/org/jabref/logic/importer/util/ShortDOIServiceException.java b/src/main/java/org/jabref/logic/importer/util/ShortDOIServiceException.java
@@ -0,0 +1,25 @@
+package org.jabref.logic.importer.util;
+
+import org.jabref.JabRefException;
+
+public class ShortDOIServiceException extends JabRefException {
+    public ShortDOIServiceException(String message) {
+        super(message);
+    }
+
+    public ShortDOIServiceException(String message, Throwable cause) {
+        super(message, cause);
+    }
+
+    public ShortDOIServiceException(String message, String localizedMessage) {
+        super(message, localizedMessage);
+    }
+
+    public ShortDOIServiceException(String message, String localizedMessage, Throwable cause) {
+        super(message, localizedMessage, cause);
+    }
+
+    public ShortDOIServiceException(Throwable cause) {
+        super(cause);
+    }
+}
diff --git a/src/main/java/org/jabref/model/entry/identifier/DOI.java b/src/main/java/org/jabref/model/entry/identifier/DOI.java
@@ -14,14 +14,15 @@
 import org.slf4j.LoggerFactory;
 
 /**
- * Class for working with Digital object identifiers (DOIs)
+ * Class for working with Digital object identifiers (DOIs) and Short DOIs
  *
  * @see https://en.wikipedia.org/wiki/Digital_object_identifier
+ * @see http://shortdoi.org
  */
 public class DOI implements Identifier {
     private static final Logger LOGGER = LoggerFactory.getLogger(DOI.class);
 
-    // DOI resolver
+    // DOI/Short DOI resolver
     private static final URI RESOLVER = URI.create("https://doi.org");
     // Regex
     // (see http://www.doi.org/doi_handbook/2_Numbering.html)
@@ -43,20 +44,46 @@ public class DOI implements Identifier {
             + "[/:]"                            // divider
             + "(?:[^\\s]+)"                     // suffix alphanumeric without space
             + ")";                              // end group \1
+
+    // Regex (Short DOI)
+    private static final String SHORT_DOI_EXP = ""
+            + "(?:urn:)?"                       // optional urn
+            + "(?:doi:)?"                       // optional doi
+            + "("                               // begin group \1
+            + "10"                              // directory indicator
+            + "[/:%]"                            // divider
+            + "[a-zA-Z0-9]+"
+            + ")";                              // end group \1
+    private static final String FIND_SHORT_DOI_EXP = ""
+            + "(?:urn:)?"                       // optional urn
+            + "(?:doi:)?"                       // optional doi
+            + "("                               // begin group \1
+            + "10"                              // directory indicator
+            + "[/:]"                            // divider
+            + "[a-zA-Z0-9]+"
+            + "(?:[^\\s]+)"                     // suffix alphanumeric without space
+            + ")";                              // end group \1
+
     private static final String HTTP_EXP = "https?://[^\\s]+?" + DOI_EXP;
+    private static final String SHORT_DOI_HTTP_EXP = "https?://[^\\s]+?" + SHORT_DOI_EXP;
     // Pattern
     private static final Pattern EXACT_DOI_PATT = Pattern.compile("^(?:https?://[^\\s]+?)?" + DOI_EXP + "$", Pattern.CASE_INSENSITIVE);
     private static final Pattern DOI_PATT = Pattern.compile("(?:https?://[^\\s]+?)?" + FIND_DOI_EXP, Pattern.CASE_INSENSITIVE);
+    // Pattern (short DOI)
+    private static final Pattern EXACT_SHORT_DOI_PATT = Pattern.compile("^(?:https?://[^\\s]+?)?" + SHORT_DOI_EXP, Pattern.CASE_INSENSITIVE);
+    private static final Pattern SHORT_DOI_PATT = Pattern.compile("(?:https?://[^\\s]+?)?" + FIND_SHORT_DOI_EXP, Pattern.CASE_INSENSITIVE);
     // DOI
     private final String doi;
+    // Short DOI
+    private boolean isShortDoi;
 
     /**
-     * Creates a DOI from various schemes including URL, URN, and plain DOIs.
+     * Creates a DOI from various schemes including URL, URN, and plain DOIs/Short DOIs.
      *
-     * @param doi the DOI string
-     * @throws NullPointerException if DOI is null
-     * @throws IllegalArgumentException if doi does not include a valid DOI
+     * @param doi the DOI/Short DOI string
      * @return an instance of the DOI class
+     * @throws NullPointerException     if DOI/Short DOI is null
+     * @throws IllegalArgumentException if doi does not include a valid DOI/Short DOI
      */
     public DOI(String doi) {
         Objects.requireNonNull(doi);
@@ -65,33 +92,40 @@ public DOI(String doi) {
         String trimmedDoi = doi.trim();
 
         // HTTP URL decoding
-        if (doi.matches(HTTP_EXP)) {
+        if (doi.matches(HTTP_EXP) || doi.matches(SHORT_DOI_HTTP_EXP)) {
             try {
                 // decodes path segment
                 URI url = new URI(trimmedDoi);
                 trimmedDoi = url.getScheme() + "://" + url.getHost() + url.getPath();
             } catch (URISyntaxException e) {
-                throw new IllegalArgumentException(doi + " is not a valid HTTP DOI.");
+                throw new IllegalArgumentException(doi + " is not a valid HTTP DOI/Short DOI.");
             }
         }
 
-        // Extract DOI
+        // Extract DOI/Short DOI
         Matcher matcher = EXACT_DOI_PATT.matcher(trimmedDoi);
         if (matcher.find()) {
             // match only group \1
             this.doi = matcher.group(1);
         } else {
-            throw new IllegalArgumentException(trimmedDoi + " is not a valid DOI.");
+            // Short DOI
+            Matcher shortDoiMatcher = EXACT_SHORT_DOI_PATT.matcher(trimmedDoi);
+            if (shortDoiMatcher.find()) {
+                this.doi = shortDoiMatcher.group(1);
+                isShortDoi = true;
+            } else {
+                throw new IllegalArgumentException(trimmedDoi + " is not a valid DOI/Short DOI.");
+            }
         }
     }
 
     /**
      * Creates an Optional<DOI> from various schemes including URL, URN, and plain DOIs.
      *
-     * Useful for suppressing the <c>IllegalArgumentException</c> of the Constructor
-     * and checking for Optional.isPresent() instead.
+     * Useful for suppressing the <c>IllegalArgumentException</c> of the Constructor and checking for
+     * Optional.isPresent() instead.
      *
-     * @param doi the DOI string
+     * @param doi the DOI/Short DOI string
      * @return an Optional containing the DOI or an empty Optional
      */
     public static Optional<DOI> parse(String doi) {
@@ -105,19 +139,19 @@ public static Optional<DOI> parse(String doi) {
     }
 
     /**
-     * Determines whether a DOI is valid or not
+     * Determines whether a DOI/Short DOI is valid or not
      *
-     * @param doi the DOI string
+     * @param doi the DOI/Short DOI string
      * @return true if DOI is valid, false otherwise
      */
     public static boolean isValid(String doi) {
         return parse(doi).isPresent();
     }
 
     /**
-     * Tries to find a DOI inside the given text.
+     * Tries to find a DOI/Short DOI inside the given text.
      *
-     * @param text the Text which might contain a DOI
+     * @param text the Text which might contain a DOI/Short DOI
      * @return an Optional containing the DOI or an empty Optional
      */
     public static Optional<DOI> findInText(String text) {
@@ -128,6 +162,12 @@ public static Optional<DOI> findInText(String text) {
             // match only group \1
             result = Optional.of(new DOI(matcher.group(1)));
         }
+
+        matcher = SHORT_DOI_PATT.matcher(text);
+        if (matcher.find()) {
+            result = Optional.of(new DOI(matcher.group(1)));
+        }
+
         return result;
     }
 
@@ -139,18 +179,27 @@ public String toString() {
     }
 
     /**
-     * Return the plain DOI
+     * Return the plain DOI/Short DOI
      *
-     * @return the plain DOI value.
+     * @return the plain DOI/Short DOI value.
      */
     public String getDOI() {
         return doi;
     }
 
     /**
-     * Return a URI presentation for the DOI
+     * Determines whether DOI is short DOI or not
+     *
+     * @return true if DOI is short DOI, false otherwise
+     */
+    public boolean isShortDoi() {
+        return isShortDoi;
+    }
+
+    /**
+     * Return a URI presentation for the DOI/Short DOI
      *
-     * @return an encoded URI representation of the DOI
+     * @return an encoded URI representation of the DOI/Short DOI
      */
     @Override
     public Optional<URI> getExternalURI() {
@@ -165,9 +214,9 @@ public Optional<URI> getExternalURI() {
     }
 
     /**
-     * Return an ASCII URL presentation for the DOI
+     * Return an ASCII URL presentation for the DOI/Short DOI
      *
-     * @return an encoded URL representation of the DOI
+     * @return an encoded URL representation of the DOI/Short DOI
      */
     public String getURIAsASCIIString() {
         return getExternalURI().map(URI::toASCIIString).orElse("");

diff --git a/src/main/resources/l10n/JabRef_en.properties b/src/main/resources/l10n/JabRef_en.properties
@@ -1555,6 +1555,8 @@ Add\ enclosing\ braces=Add enclosing braces
 Add\ braces\ encapsulating\ the\ complete\ field\ content.=Add braces encapsulating the complete field content.
 Remove\ enclosing\ braces=Remove enclosing braces
 Removes\ braces\ encapsulating\ the\ complete\ field\ content.=Removes braces encapsulating the complete field content.
+Shorten\ DOI=Shorten DOI
+Shortens\ DOI\ to\ more\ human\ readable\ form.=Shortens DOI to more human readable form.
 Sentence\ case=Sentence case
 Shortens\ lists\ of\ persons\ if\ there\ are\ more\ than\ 2\ persons\ to\ "et\ al.".=Shortens lists of persons if there are more than 2 persons to "et al.".
 Title\ case=Title case