[CSV-239] Cannot get headers in column order from CSVRecord.

Author: Gary Gregory

src/changes/changes.xml

@@ -45,6 +45,7 @@
       <action issue="CSV-234" type="add" dev="ggregory" due-to="Roberto Benedetti, Gary Gregory">Add support for java.sql.Clob.</action>
       <action issue="CSV-237" type="update" dev="ggregory" due-to="Gary Gregory">Update to Java 8.</action>
       <action issue="CSV-238" type="fix" dev="ggregory" due-to="Stephen Olander-Waters">Escape quotes in CLOBs #39.</action>
+      <action issue="CSV-239" type="add" dev="ggregory" due-to="Gary Gregory, Dave Moten">Cannot get headers in column order from CSVRecord.</action>
       <action                 type="update" dev="ggregory" due-to="Gary Gregory">Update tests from H2 1.4.198 to 1.4.199.</action>
     </release>
     <release version="1.6" date="2018-09-22" description="Feature and bug fix release (Java 7)">

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

@@ -39,6 +39,7 @@
 import java.util.Map;
 import java.util.NoSuchElementException;
 import java.util.TreeMap;
+import java.util.stream.Collectors;
 
 /**
  * Parses CSV files according to the specified format.
@@ -133,6 +134,61 @@
  */
 public final class CSVParser implements Iterable<CSVRecord>, Closeable {
 
+    class CSVRecordIterator implements Iterator<CSVRecord> {
+        private CSVRecord current;
+
+        private CSVRecord getNextRecord() {
+            try {
+                return CSVParser.this.nextRecord();
+            } catch (final IOException e) {
+                throw new IllegalStateException(
+                        e.getClass().getSimpleName() + " reading next record: " + e.toString(), e);
+            }
+        }
+
+        @Override
+        public boolean hasNext() {
+            if (CSVParser.this.isClosed()) {
+                return false;
+            }
+            if (this.current == null) {
+                this.current = this.getNextRecord();
+            }
+
+            return this.current != null;
+        }
+
+        @Override
+        public CSVRecord next() {
+            if (CSVParser.this.isClosed()) {
+                throw new NoSuchElementException("CSVParser has been closed");
+            }
+            CSVRecord next = this.current;
+            this.current = null;
+
+            if (next == null) {
+                // hasNext() wasn't called before
+                next = this.getNextRecord();
+                if (next == null) {
+                    throw new NoSuchElementException("No more CSV records available");
+                }
+            }
+
+            return next;
+        }
+
+        @Override
+        public void remove() {
+            throw new UnsupportedOperationException();
+        }
+    }
+
+    static List<String> createHeaderNames(final Map<String, Integer> headerMap) {
+        return headerMap == null ? null
+            : headerMap.entrySet().stream().sorted(Map.Entry.comparingByValue()).map(Map.Entry::getKey)
+                .collect(Collectors.toList());
+    }
+
     /**
      * Creates a parser for the given {@link File}.
      *
@@ -229,6 +285,8 @@ public static CSVParser parse(final Reader reader, final CSVFormat format) throw
         return new CSVParser(reader, format);
     }
 
+    // the following objects are shared to reduce garbage
+
     /**
      * Creates a parser for the given {@link String}.
      *
@@ -277,13 +335,14 @@ public static CSVParser parse(final URL url, final Charset charset, final CSVFor
         return new CSVParser(new InputStreamReader(url.openStream(), charset), format);
     }
 
-    // the following objects are shared to reduce garbage
-
     private final CSVFormat format;
 
     /** A mapping of column names to column indices */
     private final Map<String, Integer> headerMap;
 
+    /** Preserve the column order to avoid re-computing it. */
+    private final List<String> headerNames;
+
     private final Lexer lexer;
 
     private final CSVRecordIterator csvRecordIterator;
@@ -349,14 +408,15 @@ public CSVParser(final Reader reader, final CSVFormat format) throws IOException
      */
     @SuppressWarnings("resource")
     public CSVParser(final Reader reader, final CSVFormat format, final long characterOffset, final long recordNumber)
-            throws IOException {
+        throws IOException {
         Assertions.notNull(reader, "reader");
         Assertions.notNull(format, "format");
 
         this.format = format;
         this.lexer = new Lexer(format, new ExtendedBufferedReader(reader));
         this.csvRecordIterator = new CSVRecordIterator();
-        this.headerMap = this.createHeaderMap();
+        this.headerMap = createHeaderMap(); // 1st
+        this.headerNames = createHeaderNames(this.headerMap); // 2nd
         this.characterOffset = characterOffset;
         this.recordNumber = recordNumber - 1;
     }
@@ -384,6 +444,53 @@ public void close() throws IOException {
         }
     }
 
+    /**
+     * Creates the name to index mapping if the format defines a header.
+     *
+     * @return null if the format has no header.
+     * @throws IOException if there is a problem reading the header or skipping the first record
+     */
+    private Map<String, Integer> createHeaderMap() throws IOException {
+        Map<String, Integer> hdrMap = null;
+        final String[] formatHeader = this.format.getHeader();
+        if (formatHeader != null) {
+            hdrMap = this.format.getIgnoreHeaderCase() ?
+                    new TreeMap<>(String.CASE_INSENSITIVE_ORDER) :
+                    new TreeMap<>();
+
+            String[] headerRecord = null;
+            if (formatHeader.length == 0) {
+                // read the header from the first line of the file
+                final CSVRecord nextRecord = this.nextRecord();
+                if (nextRecord != null) {
+                    headerRecord = nextRecord.values();
+                }
+            } else {
+                if (this.format.getSkipHeaderRecord()) {
+                    this.nextRecord();
+                }
+                headerRecord = formatHeader;
+            }
+
+            // build the name to index mappings
+            if (headerRecord != null) {
+                for (int i = 0; i < headerRecord.length; i++) {
+                    final String header = headerRecord[i];
+                    final boolean containsHeader = header == null ? false : hdrMap.containsKey(header);
+                    final boolean emptyHeader = header == null || header.trim().isEmpty();
+                    if (containsHeader && (!emptyHeader || !this.format.getAllowMissingColumnNames())) {
+                        throw new IllegalArgumentException("The header contains a duplicate name: \"" + header
+                                + "\" in " + Arrays.toString(headerRecord));
+                    }
+                    if (header != null) {
+                        hdrMap.put(header, Integer.valueOf(i));
+                    }
+                }
+            }
+        }
+        return hdrMap;
+    }
+
     /**
      * Returns the current line number in the input stream.
      *
@@ -409,11 +516,11 @@ public String getFirstEndOfLine() {
     }
 
     /**
-     * Returns a copy of the header map that iterates in column order.
+     * Returns a copy of the header map.
      * <p>
      * The map keys are column names. The map values are 0-based indices.
      * </p>
-     * @return a copy of the header map that iterates in column order.
+     * @return a copy of the header map.
      */
     public Map<String, Integer> getHeaderMap() {
         return this.headerMap == null ? null : new LinkedHashMap<>(this.headerMap);
@@ -454,53 +561,6 @@ public List<CSVRecord> getRecords() throws IOException {
         return records;
     }
 
-    /**
-     * Creates the name to index mapping if the format defines a header.
-     *
-     * @return null if the format has no header.
-     * @throws IOException if there is a problem reading the header or skipping the first record
-     */
-    private Map<String, Integer> createHeaderMap() throws IOException {
-        Map<String, Integer> hdrMap = null;
-        final String[] formatHeader = this.format.getHeader();
-        if (formatHeader != null) {
-            hdrMap = this.format.getIgnoreHeaderCase() ?
-                    new TreeMap<>(String.CASE_INSENSITIVE_ORDER) :
-                    new TreeMap<>();
-
-            String[] headerRecord = null;
-            if (formatHeader.length == 0) {
-                // read the header from the first line of the file
-                final CSVRecord nextRecord = this.nextRecord();
-                if (nextRecord != null) {
-                    headerRecord = nextRecord.values();
-                }
-            } else {
-                if (this.format.getSkipHeaderRecord()) {
-                    this.nextRecord();
-                }
-                headerRecord = formatHeader;
-            }
-
-            // build the name to index mappings
-            if (headerRecord != null) {
-                for (int i = 0; i < headerRecord.length; i++) {
-                    final String header = headerRecord[i];
-                    final boolean containsHeader = header == null ? false : hdrMap.containsKey(header);
-                    final boolean emptyHeader = header == null || header.trim().isEmpty();
-                    if (containsHeader && (!emptyHeader || !this.format.getAllowMissingColumnNames())) {
-                        throw new IllegalArgumentException("The header contains a duplicate name: \"" + header
-                                + "\" in " + Arrays.toString(headerRecord));
-                    }
-                    if (header != null) {
-                        hdrMap.put(header, Integer.valueOf(i));
-                    }
-                }
-            }
-        }
-        return hdrMap;
-    }
-
     /**
      * Gets whether this parser is closed.
      *
@@ -527,55 +587,6 @@ public Iterator<CSVRecord> iterator() {
         return csvRecordIterator;
     }
 
-    class CSVRecordIterator implements Iterator<CSVRecord> {
-        private CSVRecord current;
-
-        private CSVRecord getNextRecord() {
-            try {
-                return CSVParser.this.nextRecord();
-            } catch (final IOException e) {
-                throw new IllegalStateException(
-                        e.getClass().getSimpleName() + " reading next record: " + e.toString(), e);
-            }
-        }
-
-        @Override
-        public boolean hasNext() {
-            if (CSVParser.this.isClosed()) {
-                return false;
-            }
-            if (this.current == null) {
-                this.current = this.getNextRecord();
-            }
-
-            return this.current != null;
-        }
-
-        @Override
-        public CSVRecord next() {
-            if (CSVParser.this.isClosed()) {
-                throw new NoSuchElementException("CSVParser has been closed");
-            }
-            CSVRecord next = this.current;
-            this.current = null;
-
-            if (next == null) {
-                // hasNext() wasn't called before
-                next = this.getNextRecord();
-                if (next == null) {
-                    throw new NoSuchElementException("No more CSV records available");
-                }
-            }
-
-            return next;
-        }
-
-        @Override
-        public void remove() {
-            throw new UnsupportedOperationException();
-        }
-    }
-
     /**
      * Parses the next record from the current point in the stream.
      *
@@ -622,8 +633,8 @@ CSVRecord nextRecord() throws IOException {
         if (!this.recordList.isEmpty()) {
             this.recordNumber++;
             final String comment = sb == null ? null : sb.toString();
-            result = new CSVRecord(this.recordList.toArray(new String[this.recordList.size()]), this.headerMap, comment,
-                    this.recordNumber, startCharPosition);
+            result = new CSVRecord(this.recordList.toArray(new String[this.recordList.size()]), this.headerMap,
+                this.headerNames, comment, this.recordNumber, startCharPosition);
         }
         return result;
     }