[WIP] Use CERMINE as PDF parser

build.gradle

@@ -60,6 +60,9 @@ repositories {
  maven {
  url 'https://oss.sonatype.org/content/groups/public'
  }
+ maven {
+ url 'http://maven.icm.edu.pl/artifactory/repo'
+ }
 }
 
 configurations {
@@ -81,6 +84,8 @@ dependencies {
  compile 'org.apache.pdfbox:fontbox:1.8.13'
  compile 'org.apache.pdfbox:jempbox:1.8.13'
 
+ compile 'pl.edu.icm.cermine:cermine-impl:1.11'
+
  // required for reading write-protected PDFs - see https://github.com/JabRef/jabref/pull/942#issuecomment-209252635
  compile 'org.bouncycastle:bcprov-jdk15on:1.55'
 

src/main/java/net/sf/jabref/logic/bibtexkeypattern/BibtexKeyPatternPreferences.java

@@ -1,5 +1,7 @@
 package net.sf.jabref.logic.bibtexkeypattern;
 
+import java.util.Collections;
+
 import net.sf.jabref.model.bibtexkeypattern.GlobalBibtexKeyPattern;
 
 public class BibtexKeyPatternPreferences {
@@ -12,6 +14,16 @@ public class BibtexKeyPatternPreferences {
  private final GlobalBibtexKeyPattern keyPattern;
  private Character keywordDelimiter;
 
+ public BibtexKeyPatternPreferences() {
+ this.keyPatternRegex = "";
+ this.keyPatternReplacement = "";
+ this.alwaysAddLetter = false;
+ this.firstLetterA = false;
+ this.enforceLegalKey = false;
+ this.keyPattern = new GlobalBibtexKeyPattern(Collections.emptyList());
+ this.keywordDelimiter = ',';
+ }
+
  public BibtexKeyPatternPreferences(String keyPatternRegex, String keyPatternReplacement, boolean alwaysAddLetter,
  boolean firstLetterA, boolean enforceLegalKey, GlobalBibtexKeyPattern keyPattern,
  Character keywordDelimiter) {

src/main/java/net/sf/jabref/logic/formatter/bibtexfields/NormalizeDateFormatter.java

@@ -1,14 +1,10 @@
 package net.sf.jabref.logic.formatter.bibtexfields;
 
-import java.time.format.DateTimeFormatter;
-import java.time.format.DateTimeParseException;
-import java.time.temporal.TemporalAccessor;
-import java.util.Arrays;
-import java.util.List;
 import java.util.Optional;
 
 import net.sf.jabref.logic.l10n.Localization;
 import net.sf.jabref.model.cleanup.Formatter;
+import net.sf.jabref.model.entry.Date;
 
 /**
  * This class transforms date to the format yyyy-mm-dd or yyyy-mm..
@@ -34,13 +30,8 @@ public String getKey() {
  */
  @Override
  public String format(String value) {
- Optional<TemporalAccessor> parsedDate = tryParseDate(value);
- if (!parsedDate.isPresent()) {
- return value;
- }
-
- DateTimeFormatter dateFormatter = DateTimeFormatter.ofPattern("uuuu-MM[-dd]");
- return dateFormatter.format(parsedDate.get());
+ Optional<Date> parsedDate = Date.parse(value);
+ return parsedDate.map(Date::getNormalized).orElse(value);
  }
 
  @Override
@@ -53,29 +44,6 @@ public String getExampleInput() {
  return "29.11.2003";
  }
 
- /*
- * Try to parse the following formats
- * "M/y" (covers 9/15, 9/2015, and 09/2015)
- * "MMMM (dd), yyyy" (covers September 1, 2015 and September, 2015)
- * "yyyy-MM-dd" (covers 2009-1-15)
- * "d.M.uuuu" (covers 15.1.2015)
- * "uuuu.M.d" (covers 2015.1.15)
- * The code is essentially taken from http://stackoverflow.com/questions/4024544/how-to-parse-dates-in-multiple-formats-using-simpledateformat.
- */
- private Optional<TemporalAccessor> tryParseDate(String dateString) {
- List<String> formatStrings = Arrays.asList("uuuu-M-d", "uuuu-M", "M/uu", "M/uuuu", "MMMM d, uuuu", "MMMM, uuuu",
- "d.M.uuuu", "uuuu.M.d");
- for (String formatString : formatStrings) {
- try {
- return Optional.of(DateTimeFormatter.ofPattern(formatString).parse(dateString));
- } catch (DateTimeParseException ignored) {
- // Ignored
- }
- }
-
- return Optional.empty();
- }
-
  @Override
  public int hashCode() {
  return defaultHashCode();

src/main/java/net/sf/jabref/logic/importer/ImportFormatPreferences.java

@@ -1,6 +1,8 @@
 package net.sf.jabref.logic.importer;
 
 import java.nio.charset.Charset;
+import java.nio.charset.StandardCharsets;
+import java.util.Collections;
 import java.util.Set;
 
 import net.sf.jabref.logic.bibtex.FieldContentParserPreferences;
@@ -16,6 +18,15 @@ public class ImportFormatPreferences {
  private final FieldContentParserPreferences fieldContentParserPreferences;
  private final boolean keywordSyncEnabled;
 
+ public ImportFormatPreferences() {
+ this.customImportList = Collections.emptySet();
+ this.encoding = StandardCharsets.UTF_8;
+ this.keywordSeparator = ',';
+ this.bibtexKeyPatternPreferences = new BibtexKeyPatternPreferences();
+ this.fieldContentParserPreferences = new FieldContentParserPreferences();
+ this.keywordSyncEnabled = false;
+ }
+
  public ImportFormatPreferences(Set<CustomImporter> customImportList, Charset encoding, Character keywordSeparator,
  BibtexKeyPatternPreferences bibtexKeyPatternPreferences,
  FieldContentParserPreferences fieldContentParserPreferences, boolean keywordSyncEnabled) {

src/main/java/net/sf/jabref/logic/importer/ParserResult.java

@@ -21,14 +21,12 @@ public class ParserResult {
 
  private static final ParserResult NULL_RESULT = new ParserResult(null, null, null);
  private final BibDatabase base;
- private MetaData metaData;
  private final Map<String, EntryType> entryTypes;
- private BibDatabaseContext bibDatabaseContext;
-
- private File file;
  private final List<String> warnings = new ArrayList<>();
  private final List<String> duplicateKeys = new ArrayList<>();
-
+ private MetaData metaData;
+ private BibDatabaseContext bibDatabaseContext;
+ private File file;
  private String errorMessage;
 
  private boolean invalid;
@@ -61,6 +59,18 @@ public static ParserResult fromErrorMessage(String message) {
  return parserResult;
  }
 
+ public static ParserResult getNullResult() {
+ return NULL_RESULT;
+ }
+
+ public static ParserResult fromError(Exception exception) {
+ return fromErrorMessage(exception.getLocalizedMessage());
+ }
+
+ public static ParserResult fromEntry(BibEntry entry) {
+ return new ParserResult(Collections.singleton(entry));
+ }
+
  /**
  * Check if this base is marked to be added to the currently open tab. Default is false.
  *
@@ -181,8 +191,4 @@ public boolean hasDatabaseContext() {
  public boolean isNullResult() {
  return this == NULL_RESULT;
  }
-
- public static ParserResult getNullResult() {
- return NULL_RESULT;
- }
 }