Merge pull request apache#412 from ikfid/mongo_db_csv_empty_first_column

MongoDB CSV empty first column parsing fix

Author: garydgregory

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

@@ -334,18 +334,7 @@ private Token parseEncapsulatedToken(final Token token) throws IOException {
         while (true) {
             c = reader.read();
 
-            if (isEscape(c)) {
-                if (isEscapeDelimiter()) {
-                    token.content.append(delimiter);
-                } else {
-                    final int unescaped = readEscape();
-                    if (unescaped == EOF) { // unexpected char after escape
-                        token.content.append((char) c).append((char) reader.getLastChar());
-                    } else {
-                        token.content.append((char) unescaped);
-                    }
-                }
-            } else if (isQuoteChar(c)) {
+            if (isQuoteChar(c)) {
                 if (isQuoteChar(reader.lookAhead())) {
                     // double or escaped encapsulator -> add single encapsulator to token
                     c = reader.read();
@@ -376,6 +365,17 @@ private Token parseEncapsulatedToken(final Token token) throws IOException {
                         }
                     }
                 }
+            } else if (isEscape(c)) {
+                if (isEscapeDelimiter()) {
+                    token.content.append(delimiter);
+                } else {
+                    final int unescaped = readEscape();
+                    if (unescaped == EOF) { // unexpected char after escape
+                        token.content.append((char) c).append((char) reader.getLastChar());
+                    } else {
+                        token.content.append((char) unescaped);
+                    }
+                }
             } else if (isEndOfFile(c)) {
                 if (lenientEof) {
                     token.type = Token.Type.EOF;

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

@@ -1562,6 +1562,41 @@ public void testTrim() throws Exception {
             assertEquals(3, record.size());
         }}
 
+    @Test
+    public void testParsingPrintedEmptyFirstColumn() throws Exception {
+        String[][] lines = new String[][] {
+                {"a", "b"},
+                {"", "x"}
+        };
+        Exception firstException = null;
+        for (CSVFormat.Predefined format : CSVFormat.Predefined.values()) {
+            try {
+                StringWriter buf = new StringWriter();
+                try (CSVPrinter printer = new CSVPrinter(buf, format.getFormat())) {
+                    for (String[] line : lines) {
+                        printer.printRecord((Object[]) line);
+                    }
+                }
+                try (CSVParser csvRecords = new CSVParser(new StringReader(buf.toString()), format.getFormat())) {
+                    for (String[] line : lines) {
+                        assertArrayEquals(line, csvRecords.nextRecord().values());
+                    }
+                    assertNull(csvRecords.nextRecord());
+                }
+            } catch (Exception | Error e) {
+                Exception detailedException = new RuntimeException("format: " + format, e);
+                if (firstException == null) {
+                    firstException = detailedException;
+                } else {
+                    firstException.addSuppressed(detailedException);
+                }
+            }
+        }
+
+        if (firstException != null)
+            throw firstException;
+    }
+
     private void validateLineNumbers(final String lineSeparator) throws IOException {
         try (final CSVParser parser = CSVParser.parse("a" + lineSeparator + "b" + lineSeparator + "c", CSVFormat.DEFAULT.withRecordSeparator(lineSeparator))) {
             assertEquals(0, parser.getCurrentLineNumber());