[CSV-233] Add predefined CSVFormats for printing MongoDB CSV and TSV.

Author: garydgregory

src/changes/changes.xml

@@ -50,6 +50,7 @@
       <action issue="CSV-209" type="fix" dev="ggregory" due-to="Gary Gregory">Create CSVFormat.ORACLE preset.</action>
       <action issue="CSV-224" type="fix" dev="ggregory" due-to="David Warshaw">Some multi-iterator parsing peek sequences incorrectly consume elements.</action>
       <action issue="CSV-225" type="fix" dev="ggregory" due-to="Anson Schwabecher">Parse method should avoid creating a redundant BufferedReader.</action>
+      <action issue="CSV-233" type="fix" dev="ggregory" due-to="Gary Gregory">Add predefined CSVFormats for printing MongoDB CSV and TSV.</action>
     </release>
     <release version="1.5" date="2017-09-03" description="Feature and bug fix release">
       <action issue="CSV-203" type="fix" dev="ggregory" due-to="Richard Wheeldon, Kai Paroth">withNullString value is printed without quotes when QuoteMode.ALL is specified; add QuoteMode.ALL_NON_NULL. PR #17.</action>

src/main/java/org/apache/commons/csv/CSVFormat.java

@@ -189,6 +189,18 @@ public enum Predefined {
          */
         InformixUnloadCsv(CSVFormat.INFORMIX_UNLOAD_CSV),
 
+        /**
+         * @see CSVFormat#MONGODB_CSV
+         * @since 1.7
+         */
+        MongoDBCsv(CSVFormat.MONGODB_CSV),
+
+        /**
+         * @see CSVFormat#MONGODB_TSV
+         * @since 1.7
+         */
+        MongoDBTsv(CSVFormat.MONGODB_TSV),
+
         /**
          * @see CSVFormat#MYSQL
          */
@@ -351,6 +363,78 @@ public CSVFormat getFormat() {
             .withRecordSeparator(LF);
     // @formatter:on
 
+    /**
+     * Default MongoDB CSV format used by the {@code mongoexport} operation.
+     * <p>
+     * <b>Parsing is not supported yet.</b>
+     * </p>
+     *
+     * <p>
+     * This is a comma-delimited format. Values are double quoted only if needed and special characters are escaped with
+     * {@code '"'}. A header line with field names is expected.
+     * </p>
+     *
+     * <p>
+     * Settings are:
+     * </p>
+     * <ul>
+     * <li>{@code withDelimiter(',')}</li>
+     * <li>{@code withEscape('"')}</li>
+     * <li>{@code withQuote('"')}</li>
+     * <li>{@code withQuoteMode(QuoteMode.ALL_NON_NULL)}</li>
+     * <li>{@code withSkipHeaderRecord(false)}</li>
+     * </ul>
+     *
+     * @see Predefined#MongoDBCsv
+     * @see <a href="https://docs.mongodb.com/manual/reference/program/mongoexport/">MongoDB mongoexport command
+     *      documentation</a>
+     * @since 1.7
+     */
+    // @formatter:off
+    public static final CSVFormat MONGODB_CSV = DEFAULT
+            .withDelimiter(COMMA)
+            .withEscape(DOUBLE_QUOTE_CHAR)
+            .withQuote(DOUBLE_QUOTE_CHAR)
+            .withQuoteMode(QuoteMode.MINIMAL)
+            .withSkipHeaderRecord(false);
+    // @formatter:off
+    
+    /**
+     * Default MongoDB TSV format used by the {@code mongoexport} operation.
+     * <p>
+     * <b>Parsing is not supported yet.</b>
+     * </p>
+     *
+     * <p>
+     * This is a tab-delimited format. Values are double quoted only if needed and special
+     * characters are escaped with {@code '"'}. A header line with field names is expected.
+     * </p>
+     *
+     * <p>
+     * Settings are:
+     * </p>
+     * <ul>
+     * <li>{@code withDelimiter('\t')}</li>
+     * <li>{@code withEscape('"')}</li>
+     * <li>{@code withQuote('"')}</li>
+     * <li>{@code withQuoteMode(QuoteMode.ALL_NON_NULL)}</li>
+     * <li>{@code withSkipHeaderRecord(false)}</li>
+     * </ul>
+     *
+     * @see Predefined#MongoDBCsv
+     * @see <a href="https://docs.mongodb.com/manual/reference/program/mongoexport/">MongoDB mongoexport command
+     *          documentation</a>
+     * @since 1.7
+     */
+    // @formatter:off
+    public static final CSVFormat MONGODB_TSV = DEFAULT
+            .withDelimiter(TAB)
+            .withEscape(DOUBLE_QUOTE_CHAR)
+            .withQuote(DOUBLE_QUOTE_CHAR)
+            .withQuoteMode(QuoteMode.MINIMAL)
+            .withSkipHeaderRecord(false);
+    // @formatter:off
+    
     /**
      * Default MySQL format used by the {@code SELECT INTO OUTFILE} and {@code LOAD DATA INFILE} operations.
      *

src/test/java/org/apache/commons/csv/CSVFormatPredefinedTest.java

@@ -55,6 +55,16 @@ public void testPostgreSqlCsv() {
         test(CSVFormat.POSTGRESQL_CSV, "PostgreSQLCsv");
     }
 
+    @Test
+    public void testMongoDbCsv() {
+        test(CSVFormat.MONGODB_CSV, "MongoDBCsv");
+    }
+
+    @Test
+    public void testMongoDbTsv() {
+        test(CSVFormat.MONGODB_TSV, "MongoDBTsv");
+    }
+
     @Test
     public void testPostgreSqlText() {
         test(CSVFormat.POSTGRESQL_TEXT, "PostgreSQLText");

src/test/java/org/apache/commons/csv/CSVParserTest.java

@@ -773,19 +773,37 @@ record = records.next();
         assertFalse(records.hasNext());
     }
 
+    @Test
+    @Ignore
+    public void testMongoDbCsv() throws Exception {
+        try (final CSVParser parser = CSVParser.parse("\"a a\",b,c" + LF + "d,e,f", CSVFormat.MONGODB_CSV)) {
+            final Iterator<CSVRecord> itr1 = parser.iterator();
+            final Iterator<CSVRecord> itr2 = parser.iterator();
+
+            final CSVRecord first = itr1.next();
+            assertEquals("a a", first.get(0));
+            assertEquals("b", first.get(1));
+            assertEquals("c", first.get(2));
+
+            final CSVRecord second = itr2.next();
+            assertEquals("d", second.get(0));
+            assertEquals("e", second.get(1));
+            assertEquals("f", second.get(2));
+        }
+    }
+    
     @Test
     // TODO this may lead to strange behavior, throw an exception if iterator() has already been called?
     public void testMultipleIterators() throws Exception {
-        try (final CSVParser parser = CSVParser.parse("a,b,c" + CR + "d,e,f", CSVFormat.DEFAULT)) {
+        try (final CSVParser parser = CSVParser.parse("a,b,c" + CRLF + "d,e,f", CSVFormat.DEFAULT)) {
             final Iterator<CSVRecord> itr1 = parser.iterator();
-            final Iterator<CSVRecord> itr2 = parser.iterator();
 
             final CSVRecord first = itr1.next();
             assertEquals("a", first.get(0));
             assertEquals("b", first.get(1));
             assertEquals("c", first.get(2));
 
-            final CSVRecord second = itr2.next();
+            final CSVRecord second = itr1.next();
             assertEquals("d", second.get(0));
             assertEquals("e", second.get(1));
             assertEquals("f", second.get(2));