Add Collection of Comp Sci Bibliographies fetcher

src/main/java/org/jabref/logic/formatter/bibtexfields/RemoveDigitsFormatter.java

@@ -0,0 +1,39 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import java.util.Objects;
+import java.util.regex.Pattern;
+
+import org.jabref.logic.l10n.Localization;
+import org.jabref.model.cleanup.Formatter;
+
+public class RemoveDigitsFormatter extends Formatter {
+
+ private static final Pattern DIGITS = Pattern.compile("[ ]\\d+");
+
+ @Override
+ public String getName() {
+ return Localization.lang("Remove digits");
+ }
+
+ @Override
+ public String getKey() {
+ return "remove_digits";
+ }
+
+ @Override
+ public String format(String value) {
+ Objects.requireNonNull(value);
+
+ return DIGITS.matcher(value).replaceAll("");
+ }
+
+ @Override
+ public String getDescription() {
+ return Localization.lang("Removes digits.");
+ }
+
+ @Override
+ public String getExampleInput() {
+ return "In 012 CDMA";
+ }
+}

src/main/java/org/jabref/logic/formatter/bibtexfields/RemoveRedundantSpacesFormatter.java

@@ -0,0 +1,42 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import java.util.Objects;
+import java.util.regex.Pattern;
+
+import org.jabref.logic.l10n.Localization;
+import org.jabref.model.cleanup.Formatter;
+
+/**
+ * Finds any occurrence of consecutive spaces and replaces it with a single space
+ */
+public class RemoveRedundantSpacesFormatter extends Formatter {
+
+ private static final Pattern MULTIPLE_SPACES = Pattern.compile(" {2,}");
+
+ @Override
+ public String getName() {
+ return Localization.lang("Remove redundant spaces");
+ }
+
+ @Override
+ public String getKey() {
+ return "remove_redundant_spaces";
+ }
+
+ @Override
+ public String format(String value) {
+ Objects.requireNonNull(value);
+
+ return MULTIPLE_SPACES.matcher(value).replaceAll(" ");
+ }
+
+ @Override
+ public String getDescription() {
+ return Localization.lang("Replaces consecutive spaces with a single space in the field content.");
+ }
+
+ @Override
+ public String getExampleInput() {
+ return "In CDMA";
+ }
+}

src/main/java/org/jabref/logic/formatter/bibtexfields/RemoveTabsFormatter.java

@@ -0,0 +1,42 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import java.util.Objects;
+import java.util.regex.Pattern;
+
+import org.jabref.logic.l10n.Localization;
+import org.jabref.model.cleanup.Formatter;
+
+/**
+ * Replaces any tab with a space
+ */
+public class RemoveTabsFormatter extends Formatter {
+
+ private static final Pattern TAB = Pattern.compile("\t+");
+
+ @Override
+ public String getName() {
+ return Localization.lang("Remove tabs");
+ }
+
+ @Override
+ public String getKey() {
+ return "remove_tabs";
+ }
+
+ @Override
+ public String format(String value) {
+ Objects.requireNonNull(value);
+
+ return TAB.matcher(value).replaceAll(" ");
+ }
+
+ @Override
+ public String getDescription() {
+ return Localization.lang("Removes tabs in the field content.");
+ }
+
+ @Override
+ public String getExampleInput() {
+ return "In \t\t CDMA";
+ }
+}

src/main/java/org/jabref/logic/importer/fetcher/CollectionOfComputerScienceBibliographiesFetcher.java

@@ -4,10 +4,17 @@
 import java.net.URISyntaxException;
 import java.net.URL;
 
+import org.jabref.logic.formatter.bibtexfields.RemoveDigitsFormatter;
+import org.jabref.logic.formatter.bibtexfields.RemoveNewlinesFormatter;
+import org.jabref.logic.formatter.bibtexfields.RemoveRedundantSpacesFormatter;
+import org.jabref.logic.formatter.bibtexfields.RemoveTabsFormatter;
 import org.jabref.logic.importer.FetcherException;
 import org.jabref.logic.importer.ImportFormatPreferences;
 import org.jabref.logic.importer.Parser;
 import org.jabref.logic.importer.SearchBasedParserFetcher;
+import org.jabref.model.cleanup.FieldFormatterCleanup;
+import org.jabref.model.entry.BibEntry;
+import org.jabref.model.entry.field.StandardField;
 
 import org.apache.http.client.utils.URIBuilder;
 
@@ -39,4 +46,12 @@ public Parser getParser() {
  public String getName() {
  return "Collection of Computer Science Bibliographies";
  }
+
+ @Override
+ public void doPostCleanup(BibEntry entry) {
+ new FieldFormatterCleanup(StandardField.ABSTRACT, new RemoveNewlinesFormatter()).cleanup(entry);
+ new FieldFormatterCleanup(StandardField.ABSTRACT, new RemoveTabsFormatter()).cleanup(entry);
+ new FieldFormatterCleanup(StandardField.ABSTRACT, new RemoveRedundantSpacesFormatter()).cleanup(entry);
+ new FieldFormatterCleanup(StandardField.EDITOR, new RemoveDigitsFormatter()).cleanup(entry);
+ }
 }

src/test/java/org/jabref/logic/formatter/bibtexfields/RemoveDigitsFormatterTest.java

@@ -0,0 +1,31 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+public class RemoveDigitsFormatterTest {
+
+ private RemoveDigitsFormatter formatter;
+
+ @BeforeEach
+ public void setUp() {
+ formatter = new RemoveDigitsFormatter();
+ }
+
+ @Test
+ public void doNothingIfSingleSpace() {
+ assertEquals("one digit", formatter.format("one 1 digit"));
+ }
+
+ @Test
+ public void doNothingIfNoSpace() {
+ assertEquals("two digits", formatter.format("two 01 digits"));
+ }
+
+ @Test
+ public void removeAllButOneSpacesIfTwo() {
+ assertEquals("no digits", formatter.format("no digits"));
+ }
+}

src/test/java/org/jabref/logic/formatter/bibtexfields/RemoveRedundantSpacesFormatterTest.java

@@ -0,0 +1,36 @@
+package org.jabref.logic.formatter.bibtexfields;
+
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+public class RemoveRedundantSpacesFormatterTest {
+
+ private RemoveRedundantSpacesFormatter formatter;
+
+ @BeforeEach
+ public void setUp() {
+ formatter = new RemoveRedundantSpacesFormatter();
+ }
+
+ @Test
+ public void doNothingIfSingleSpace() {
+ assertEquals("single space", formatter.format("single space"));
+ }
+
+ @Test
+ public void doNothingIfNoSpace() {
+ assertEquals("nospace", formatter.format("nospace"));
+ }
+
+ @Test
+ public void removeAllButOneSpacesIfTwo() {
+ assertEquals("two spaces", formatter.format("two spaces"));
+ }
+
+ @Test
+ public void removeAllButOneSpacesIfThree() {
+ assertEquals("three spaces", formatter.format("three spaces"));
+ }
+}

src/test/java/org/jabref/logic/formatter/bibtexfields/RemoveTabsFormatterTest.java

@@ -0,0 +1,32 @@
+
+package org.jabref.logic.formatter.bibtexfields;
+
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+public class RemoveTabsFormatterTest {
+
+ private RemoveTabsFormatter formatter;
+
+ @BeforeEach
+ public void setUp() {
+ formatter = new RemoveTabsFormatter();
+ }
+
+ @Test
+ public void removeSingleTab() {
+ assertEquals("single tab", formatter.format("single\ttab"));
+ }
+
+ @Test
+ public void removeMultipleTabs() {
+ assertEquals("multiple tabs", formatter.format("multiple\t\ttabs"));
+ }
+
+ @Test
+ public void doNothingIfNoTab() {
+ assertEquals("notab", formatter.format("notab"));
+ }
+}

src/test/java/org/jabref/logic/importer/fetcher/CollectionOfComputerScienceBibliographiesFetcherTest.java

@@ -69,28 +69,28 @@ public void performSearchReturnsMatchingMultipleEntries() throws FetcherExceptio
  .withField(new UnknownField("identifier"), "urn:isbn:978-1-4503-5217-8; doi:10.1145/3129790.3129810; ISI:000426556400034")
  .withField(new UnknownField("subject"), "Software Architecture; Code Churn; Open Source; Architecrual Erosion; Technical Debt; Software Engineering; Programvaruteknik")
  .withField(new UnknownField("relation"), "ACM International Conference Proceeding Series; ECSA '17~Proceedings of the 11th European Conference on Software Architecture : Companion Proceedings, p. 152-158")
- .withField(StandardField.ABSTRACT, "The open source application JabRef has existed since\r\n" +
- "\t\t 2003. In 2015, the developers decided to make an\r\n" +
- "\t\t architectural refactoring as continued development was\r\n" +
- "\t\t deemed too demanding. The developers also introduced\r\n" +
- "\t\t Static Architecture Conformance Checking (SACC) to\r\n" +
- "\t\t prevent violations to the intended architecture.\r\n" +
- "\t\t Measurements mined from source code repositories such\r\n" +
- "\t\t as code churn and code ownership has been linked to\r\n" +
- "\t\t several problems, for example fault proneness, security\r\n" +
- "\t\t vulnerabilities, code smells, and degraded\r\n" +
- "\t\t maintainability. The root cause of such problems can be\r\n" +
- "\t\t architectural. To determine the impact of the\r\n" +
- "\t\t refactoring of JabRef, we measure the code churn and\r\n" +
- "\t\t code ownership before and after the refactoring and\r\n" +
- "\t\t find that large files with violations had a\r\n" +
- "\t\t significantly higher code churn than large files\r\n" +
- "\t\t without violations before the refactoring. After the\r\n" +
- "\t\t refactoring, the files that had violations show a more\r\n" +
- "\t\t normal code churn. We find no such effect on code\r\n" +
- "\t\t ownership. We conclude that files that contain\r\n" +
- "\t\t violations detectable by SACC methods are connected to\r\n" +
- "\t\t higher than normal code churn.")
+ .withField(StandardField.ABSTRACT, "The open source application JabRef has existed since" +
+ " 2003. In 2015, the developers decided to make an" +
+ " architectural refactoring as continued development was" +
+ " deemed too demanding. The developers also introduced" +
+ " Static Architecture Conformance Checking (SACC) to" +
+ " prevent violations to the intended architecture." +
+ " Measurements mined from source code repositories such" +
+ " as code churn and code ownership has been linked to" +
+ " several problems, for example fault proneness, security" +
+ " vulnerabilities, code smells, and degraded" +
+ " maintainability. The root cause of such problems can be" +
+ " architectural. To determine the impact of the" +
+ " refactoring of JabRef, we measure the code churn and" +
+ " code ownership before and after the refactoring and" +
+ " find that large files with violations had a" +
+ " significantly higher code churn than large files" +
+ " without violations before the refactoring. After the" +
+ " refactoring, the files that had violations show a more" +
+ " normal code churn. We find no such effect on code" +
+ " ownership. We conclude that files that contain" +
+ " violations detectable by SACC methods are connected to" +
+ " higher than normal code churn.")
  .withField(StandardField.TYPE, "info:eu-repo/semantics/conferenceObject")
  .withField(new UnknownField("description"), "Information and Software Qualtiy")
  .withField(StandardField.PAGES, "152--158")
@@ -104,6 +104,7 @@ public void performSearchReturnsMatchingMultipleEntries() throws FetcherExceptio
  .withField(StandardField.AUTHOR, "Tobias Olsson and Morgan Ericsson and Anna Wingkvist")
  .withField(StandardField.YEAR, "2017");
 
+ // Checking entries in the set as the query is generic and returns a changing result set
  assertTrue(searchResult.contains(firstBibEntry));
  assertTrue(searchResult.contains(secondBibEntry));
  }