Fix dotnet#2906

pgovind · Apr 28, 2020 · fe972e5 · fe972e5
1 parent 4cf5e04
commit fe972e5
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 23 deletions.
diff --git a/src/Microsoft.Data.Analysis/DataFrame.IO.cs b/src/Microsoft.Data.Analysis/DataFrame.IO.cs
@@ -198,55 +198,58 @@ public static DataFrame LoadCsv(Stream csvStream,
                                 Encoding encoding = null)
         {
             if (!csvStream.CanSeek)
+            {
                 throw new ArgumentException(Strings.NonSeekableStream, nameof(csvStream));
+            }
 
             var linesForGuessType = new List<string[]>();
             long rowline = 0;
             int numberOfColumns = dataTypes?.Length ?? 0;
 
             if (header == true && numberOfRowsToRead != -1)
+            {
                 numberOfRowsToRead++;
+            }
 
             List<DataFrameColumn> columns;
             long streamStart = csvStream.Position;
             // First pass: schema and number of rows.
             using (var streamReader = new StreamReader(csvStream, encoding ?? Encoding.UTF8, detectEncodingFromByteOrderMarks: true, DefaultStreamReaderBufferSize, leaveOpen: true))
             {
                 string line = null;
-                if (dataTypes == null)
+                line = streamReader.ReadLine();
+                while (line != null)
                 {
-                    line = streamReader.ReadLine();
-                    while (line != null)
+                    if ((numberOfRowsToRead == -1) || rowline < numberOfRowsToRead)
                     {
-                        if ((numberOfRowsToRead == -1) || rowline < numberOfRowsToRead)
+                        if (linesForGuessType.Count < guessRows)
                         {
-                            if (linesForGuessType.Count < guessRows)
+                            var spl = line.Split(separator);
+                            if (header && rowline == 0)
                             {
-                                var spl = line.Split(separator);
-                                if (header && rowline == 0)
-                                {
-                                    if (columnNames == null)
-                                        columnNames = spl;
-                                }
-                                else
+                                if (columnNames == null)
                                 {
-                                    linesForGuessType.Add(spl);
-                                    numberOfColumns = Math.Max(numberOfColumns, spl.Length);
+                                    columnNames = spl;
                                 }
                             }
+                            else
+                            {
+                                linesForGuessType.Add(spl);
+                                numberOfColumns = Math.Max(numberOfColumns, spl.Length);
+                            }
                         }
-                        ++rowline;
-                        if (rowline == guessRows)
-                        {
-                            break;
-                        }
-                        line = streamReader.ReadLine();
                     }
-
-                    if (linesForGuessType.Count == 0)
+                    ++rowline;
+                    if (rowline == guessRows)
                     {
-                        throw new FormatException(Strings.EmptyFile);
+                        break;
                     }
+                    line = streamReader.ReadLine();
+                }
+
+                if (linesForGuessType.Count == 0)
+                {
+                    throw new FormatException(Strings.EmptyFile);
                 }
 
                 columns = new List<DataFrameColumn>(numberOfColumns);

diff --git a/tests/Microsoft.Data.Analysis.Tests/DataFrame.IOTests.cs b/tests/Microsoft.Data.Analysis.Tests/DataFrame.IOTests.cs
@@ -176,6 +176,14 @@ Stream GetStream(string streamData)
             Assert.True(typeof(float) == df.Columns[4].DataType);
             Assert.True(typeof(string) == df.Columns[5].DataType);
             Assert.True(typeof(double) == df.Columns[6].DataType);
+
+            Assert.Equal("vendor_id", df.Columns[0].Name);
+            Assert.Equal("rate_code", df.Columns[1].Name);
+            Assert.Equal("passenger_count", df.Columns[2].Name);
+            Assert.Equal("trip_time_in_secs", df.Columns[3].Name);
+            Assert.Equal("trip_distance", df.Columns[4].Name);
+            Assert.Equal("payment_type", df.Columns[5].Name);
+            Assert.Equal("fare_amount", df.Columns[6].Name);
             VerifyColumnTypes(df);
 
             foreach (var column in df.Columns)