Removed unescaping of double quotes in CSV export & added tests in pr…

…inter to ensure proper escape of multiple adjacent double quotes
Bit-Quill · Nov 18, 2022 · 32677c4 · 32677c4
1 parent 3960457
commit 32677c4
Show file tree

Hide file tree

Showing 3 changed files with 69 additions and 4 deletions.
diff --git a/...ne-export/src/main/java/com/amazonaws/services/neptune/propertygraph/schema/DataType.java b/...ne-export/src/main/java/com/amazonaws/services/neptune/propertygraph/schema/DataType.java
@@ -378,8 +378,7 @@ public static String escapeSeparators(Object value, String separator) {
     }
 
     public static String escapeDoubleQuotes(Object value) {
-        String temp = value.toString().replace("\"\"", "\"");
-        return temp.replace("\"", "\"\"");
+        return value.toString().replace("\"", "\"\"");
     }
 
     public String typeDescription() {

diff --git a/...est/java/com/amazonaws/services/neptune/propertygraph/io/CsvPropertyGraphPrinterTest.java b/...est/java/com/amazonaws/services/neptune/propertygraph/io/CsvPropertyGraphPrinterTest.java
@@ -17,12 +17,18 @@
 import com.amazonaws.services.neptune.propertygraph.schema.DataType;
 import com.amazonaws.services.neptune.propertygraph.schema.LabelSchema;
 import com.amazonaws.services.neptune.propertygraph.schema.PropertySchema;
+import org.apache.commons.csv.CSVFormat;
+import org.apache.commons.csv.CSVRecord;
 import org.junit.Test;
 
+import java.io.Reader;
+import java.io.StringReader;
 import java.io.StringWriter;
 import java.util.Arrays;
+import java.util.Collections;
 import java.util.HashMap;
 import java.util.List;
+import java.util.stream.Collectors;
 
 import static org.junit.Assert.assertEquals;
 
@@ -112,4 +118,58 @@ public void shouldUseEmptySeparatorToSeparateMultipleValues() throws Exception {
                 stringWriter.toString());
     }
 
+    @Test
+    public void shouldEscapeDoubleQuotesProperlyAfterPrintPropertiesToCSVAndRewrite() throws Exception {
+
+        String original2Quotes = "{\"hobby\" : \"watching \"Flash\"\"}";
+        String original3Quotes = "{\"hobby\" : \"watching \"The \"Flash\"\"\"}";
+
+        StringWriter stringWriter = new StringWriter();
+
+        PropertySchema propertySchema1 = new PropertySchema("property1", false, DataType.String, false);
+        PropertySchema propertySchema2 = new PropertySchema("property2", false, DataType.String, false);
+
+        LabelSchema labelSchema = new LabelSchema(new Label("Entity"));
+        labelSchema.put("property1", propertySchema1);
+        labelSchema.put("property2", propertySchema2);
+
+        HashMap<String, List<String>> props = new HashMap<String, List<String>>() {{
+            put("property1", Collections.singletonList(original2Quotes));
+            put("property2", Collections.singletonList(original3Quotes));
+        }};
+
+        CsvPropertyGraphPrinter printer = new CsvPropertyGraphPrinter(
+                new PrintOutputWriter("outputId", stringWriter),
+                labelSchema,
+                new PrinterOptions(CsvPrinterOptions.builder().build()));
+
+        printer.printProperties(props);
+
+        // all double quotes should be escaped when printer prints
+        assertEquals("\"{\"\"hobby\"\" : \"\"watching \"\"Flash\"\"\"\"}\"," +
+                        "\"{\"\"hobby\"\" : \"\"watching \"\"The \"\"Flash\"\"\"\"\"\"}\"",
+                stringWriter.toString());
+
+        // using CSVFormat to read in printed items (same library used by RewriteCSV)
+        String[] filePropertyHeaders = labelSchema.propertySchemas().stream()
+                .map(p -> p.property().toString())
+                .collect(Collectors.toList())
+                .toArray(new String[]{});
+        CSVFormat format = CSVFormat.RFC4180.withHeader(filePropertyHeaders);
+        Reader in = new StringReader(stringWriter.toString());
+        Iterable<CSVRecord> records = format.parse(in);
+
+        for (CSVRecord record : records) {
+            // what CSVFormat read in from printed CSV should be the original value
+            assertEquals(original2Quotes, record.get("property1"));
+            assertEquals(original3Quotes, record.get("property2"));
+
+            // double quotes should all be properly escaped again when we format for rewrite
+            assertEquals("\"{\"\"hobby\"\" : \"\"watching \"\"Flash\"\"\"\"}\"",
+                    DataType.String.format(record.get("property1")));
+            assertEquals("\"{\"\"hobby\"\" : \"\"watching \"\"The \"\"Flash\"\"\"\"\"\"}\"",
+                    DataType.String.format(record.get("property2")));
+        }
+    }
+
 }
diff --git a/...xport/src/test/java/com/amazonaws/services/neptune/propertygraph/schema/DataTypeTest.java b/...xport/src/test/java/com/amazonaws/services/neptune/propertygraph/schema/DataTypeTest.java
@@ -39,9 +39,15 @@ public void shouldEscapeDoubleQuotes() {
     }
 
     @Test
-    public void shouldNotDoubleEscapeDoubleQuotesThatHaveAlreadyBeenEscaped() {
+    public void shouldEscapeTwoDoubleQuotes() {
         String result = DataType.String.format("One \"\"two\"\" three");
-        assertEquals("\"One \"\"two\"\" three\"", result);
+        assertEquals("\"One \"\"\"\"two\"\"\"\" three\"", result);
+    }
+
+    @Test
+    public void shouldEscapeThreeDoubleQuotes() {
+        String result = DataType.String.format("One \"\"\"two\"\"\" three");
+        assertEquals("\"One \"\"\"\"\"\"two\"\"\"\"\"\" three\"", result);
     }
 
     @Test