diff options
Diffstat (limited to 'pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/tools/Normalizer.java')
| -rw-r--r-- | pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/tools/Normalizer.java | 280 |
1 files changed, 280 insertions, 0 deletions
diff --git a/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/tools/Normalizer.java b/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/tools/Normalizer.java new file mode 100644 index 0000000..a08c12e --- /dev/null +++ b/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/tools/Normalizer.java @@ -0,0 +1,280 @@ +/**
+ * <copyright> Copyright 2006 by Know-Center, Graz, Austria </copyright>
+ * PDF-AS has been contracted by the E-Government Innovation Center EGIZ, a
+ * joint initiative of the Federal Chancellery Austria and Graz University of
+ * Technology.
+ *
+ * Licensed under the EUPL, Version 1.1 or - as soon they will be approved by
+ * the European Commission - subsequent versions of the EUPL (the "Licence");
+ * You may not use this work except in compliance with the Licence.
+ * You may obtain a copy of the Licence at:
+ * http://www.osor.eu/eupl/
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the Licence is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the Licence for the specific language governing permissions and
+ * limitations under the Licence.
+ *
+ * This product combines work with different licenses. See the "NOTICE" text
+ * file for details on the various modules and licenses.
+ * The "NOTICE" text file is part of the distribution. Any derivative works
+ * that you distribute must include a readable copy of the "NOTICE" text file.
+ *
+ * $Id: Normalizer.java,v 1.5 2006/10/31 08:20:56 wprinz Exp $
+ */
+package at.knowcenter.wag.egov.egiz.tools;
+
+import java.io.Serializable;
+
+import org.apache.log4j.Level;
+import org.apache.log4j.Logger;
+
+import at.knowcenter.wag.egov.egiz.cfg.ConfigLogger;
+import at.knowcenter.wag.egov.egiz.cfg.SettingsReader;
+import at.knowcenter.wag.egov.egiz.exceptions.NormalizeException;
+import at.knowcenter.wag.egov.egiz.exceptions.SettingsException;
+
+/**
+ * This class provides wrapper methods to get an access to different normalizer implementations.
+ * <br>
+ * This class is to load the corresponding implementation of a normalizer class. Therefor it seams
+ * to be a factory. The factory settings are read from the configuration file calling the
+ * SettingsReader.
+ *
+ * @author wlackner
+ * @see at.knowcenter.wag.egov.egiz.tools.Normalizer
+ * @see at.knowcenter.wag.egov.egiz.tools.NormalizeV01
+ * @see at.knowcenter.wag.egov.egiz.cfg.SettingsReader
+ */
+public class Normalizer implements Serializable {
+// 04.11.2010 changed by exthex - normalize methods use and propagate the keepMultipleNewlines parameter
+
+ /**
+ * SVUID.
+ */
+ private static final long serialVersionUID = 4201772508393848555L;
+
+ /**
+ * The current raw string to normalize
+ */
+ private String rawString_ = null;
+ /**
+ * The current normalisation version string
+ */
+ private String normVersion_ = null;
+ /**
+ * The normalized string cache
+ */
+ private String normString_ = null;
+ /**
+ * The reference to the normalizer implementation
+ */
+ private Normalize normalize_ = null;
+// /**
+// * A given Encoding, not used now
+// */
+// private String encoding_ = null;
+ /**
+ * The SettingsReader instance
+ */
+ private SettingsReader settings_ = null;
+ /**
+ * The factory class prefix
+ */
+ private final static String CLASS_PREFIX = ".Normalize";
+ /**
+ * The default version string
+ */
+ protected final static String DEFAULT_VERSION = "V01";
+ /**
+ * The settings key defined in the settings file
+ *
+ * @see SettingsReader
+ */
+ protected final static String SETTINGS_VERSION_KEY = "normalizer.version";
+ /**
+ * The logger definition.
+ */
+ private static final Logger logger_ = ConfigLogger.getLogger(Normalizer.class);
+
+ /**
+ * New Normalizer init by the raw string and a normalizer version.
+ *
+ * @param rawString the raw string to normalize
+ * @param normVersion the nomalizer version that should be used
+ * @throws NormalizeException ErrorCode:400
+ */
+ public Normalizer(String rawString, String normVersion) throws NormalizeException {
+ rawString_ = rawString;
+ normVersion_ = normVersion;
+ init();
+ }
+
+ /**
+ * New Normalizer init by the raw string.
+ *
+ * @param rawString the raw string to normalize
+ * @throws NormalizeException ErrorCode:400
+ */
+ public Normalizer(String rawString) throws NormalizeException {
+ rawString_ = rawString;
+ init();
+ }
+
+ /**
+ * The empty constructor.
+ *
+ * @throws NormalizeException ErrorCode:400
+ */
+ public Normalizer() throws NormalizeException {
+ init();
+ }
+
+ /**
+ * Load the factory implementation. This method trys to load the configured normalizer library.
+ *
+ * @throws NormalizeException
+ */
+ public void init() throws NormalizeException {
+ loadSettings();
+ String class_name = this.getClass().getPackage().getName() + getClassName();
+ Class normalize_class = null;
+ try {
+ normalize_class = Class.forName(class_name);
+ } catch (ClassNotFoundException e) {
+ if (logger_.isEnabledFor(Level.FATAL)) {
+ logger_.fatal("Class not found:" + class_name);
+ }
+ throw new NormalizeException("Can not load normalizer library", e);
+ }
+ try {
+ normalize_ = (Normalize) normalize_class.newInstance();
+ } catch (InstantiationException e) {
+ if (logger_.isEnabledFor(Level.FATAL)) {
+ logger_.fatal("Can not instantiate:" + class_name);
+ }
+ throw new NormalizeException("Can not load normalizer library", e);
+ } catch (IllegalAccessException e) {
+ if (logger_.isEnabledFor(Level.FATAL)) {
+ logger_.fatal("Can not access:" + class_name);
+ }
+ throw new NormalizeException("Can not load normalizer library", e);
+ }
+ }
+
+ /**
+ * Returns the underlying normalizer instance.
+ * @author tknall
+ */
+ public Normalize getInstance() {
+ return this.normalize_;
+ }
+
+ /**
+ * Read the class postfix from the configuration file
+ *
+ * @return the full qualified class name
+ */
+ private String getClassName() {
+ if (normVersion_ == null) {
+ normVersion_ = settings_.getSetting(SETTINGS_VERSION_KEY, DEFAULT_VERSION);
+ }
+ return CLASS_PREFIX + normVersion_;
+ }
+
+ /*
+ * public void setEncoding(String encoding) { encoding_ = encoding; }
+ */
+
+ /**
+ * Set the raw string to normalize
+ */
+ public void setRawString(String rawString) {
+ rawString_ = rawString;
+ }
+
+// /**
+// * Return the normalized string. If the chached value does not exist the normalize method from the
+// * current normalizer implementation is called.
+// *
+// * @return the normalized string
+// */
+// public String getNormalizedString() {
+// if (normString_ == null) {
+// normalize();
+// }
+// return normString_;
+// }
+
+ /**
+ * Set a normalizer version. This activity load the new requested normalizer implementation.
+ *
+ * @param normVersion the normalizer version to be use
+ * @throws NormalizeException ErrorCode:400
+ */
+ public void setVersion(String normVersion) throws NormalizeException {
+ normVersion_ = normVersion;
+ init();
+ }
+
+ /**
+ * Return the current version string.
+ *
+ * @return the normaliser version string
+ */
+ public String getVersion() {
+ return normVersion_;
+ }
+
+ /**
+ * Wrapper method. Call the normalizer implementation method.
+ *
+ * @param rawString the raw string to normalize
+ * @param keepMultipleNewlines
+ * @return the normalized string
+ * @see NormalizeV01
+ */
+ public String normalize(String rawString, boolean keepMultipleNewlines) {
+ return normalize_.normalize(rawString, keepMultipleNewlines);
+ }
+
+ /**
+ * Wrapper method. Call the normalizer implementation method. Normalize the current raw string.
+ *
+ * @return the normalized string
+ * @see NormalizeV01
+ */
+ public String normalize(boolean keepMultipleNewlines) {
+ if (normString_ == null) {
+ normString_ = normalize(rawString_, keepMultipleNewlines);
+ }
+ return normString_;
+ }
+
+ /**
+ * Returns the normalizer line separator string.
+ * @return the line separator string
+ */
+ public String getNormCR() {
+ return normalize_.getNormCR();
+ }
+
+ /**
+ * load the class settings
+ *
+ * @throws NormalizeException
+ * @see SettingsReader
+ */
+ private void loadSettings() throws NormalizeException {
+ if (settings_ == null) {
+ try {
+ settings_ = SettingsReader.getInstance();
+ } catch (SettingsException e) {
+ String log_message = "Can not load normalizer settings. Cause:\n" + e.getMessage();
+ logger_.error(log_message, e);
+ throw new NormalizeException(log_message, e);
+ }
+ }
+ }
+}
\ No newline at end of file |