added BOM support (#26)

Now skipping the BOM if it exists

csv_files/valid_bom.csv

@@ -0,0 +1,4 @@
+header1, header2, header3
+line1, 1, 1.2
+line2, 2, 2.3
+line3, 3, 3.4

load.go

@@ -1,6 +1,7 @@
 package csvtag
 
 import (
+	"bytes"
 	"encoding/csv"
 	"fmt"
 	"io"
@@ -113,6 +114,8 @@ func LoadFromString(str string, destination interface{}, options ...CsvOptions)
 // @param separator: the separator used in the csv file.
 // @param header: the optional header if the file does not contain one.
 func readFile(file io.Reader, separator rune, header []string) ([]string, [][]string, error) {
+	file = skipBOM(file)
+
 	// Create and configure the csv reader.
 	reader := csv.NewReader(file)
 	reader.TrimLeadingSpace = true
@@ -141,6 +144,24 @@ func readFile(file io.Reader, separator rune, header []string) ([]string, [][]st
 	return header, content, nil
 }
 
+// Skip the Byte Order Mark (BOM) if it exists.
+// @param file: the io.Reader to read from.
+func skipBOM(file io.Reader) io.Reader {
+	// Read the first 3 bytes.
+	bom := make([]byte, 3)
+	_, err := file.Read(bom)
+	if err != nil {
+		return file
+	}
+
+	// If the first 3 bytes are not the BOM, reset the reader.
+	if bom[0] != 0xEF || bom[1] != 0xBB || bom[2] != 0xBF {
+		return io.MultiReader(bytes.NewReader(bom), file)
+	}
+
+	return file
+}
+
 // Map the provided content to the destination using the header and the tags.
 // @param header: the csv header to match with the struct's tags.
 // @param content: the content to put in destination.

load_test.go

@@ -55,6 +55,14 @@ func TestValideFile(t *testing.T) {
 	}
 }
 
+func TestValideFileWithBOM(t *testing.T) {
+	tabT := []test{}
+	err := LoadFromPath("csv_files/valid_bom.csv", &tabT)
+	if err != nil || checkValues(tabT) {
+		t.Fail()
+	}
+}
+
 func TestBool(t *testing.T) {
 	tabT := []testBool{}
 	err := LoadFromPath("csv_files/bool.csv", &tabT)