feat(pkg/csv2lp): add SkipHeaderLinesReader

pkg/csv2lp/skip_header_lines.go

@@ -0,0 +1,54 @@
+package csv2lp
+
+import (
+	"io"
+)
+
+// skipFirstLines is an io.Reader that skips first lines
+type skipFirstLines struct {
+	reader    io.Reader
+	skipLines int
+	// line is a mutable variable that increases until skipLines is reached
+	line int
+}
+
+// Read implements io.Reader
+func (state *skipFirstLines) Read(p []byte) (n int, err error) {
+skipHeaderLines:
+	for state.line < state.skipLines {
+		n, err := state.reader.Read(p)
+		if n == 0 {
+			return n, err
+		}
+		for i := 0; i < n; i++ {
+			if p[i] == '\n' {
+				state.line++
+				if state.line == state.skipLines {
+					// modify the buffer and return
+					if i == n-1 {
+						if err != nil {
+							return 0, err
+						}
+						// continue with the next chunk
+						break skipHeaderLines
+					} else {
+						// copy all bytes after the newline
+						for j := i + 1; j < n; j++ {
+							p[j-i-1] = p[j]
+						}
+						return n - i - 1, err
+					}
+				}
+			}
+		}
+	}
+	return state.reader.Read(p)
+}
+
+// SkipHeaderLinesReader wraps a reader to skip the first skipLines lines
+func SkipHeaderLinesReader(skipLines int, reader io.Reader) io.Reader {
+	return &skipFirstLines{
+		skipLines: skipLines,
+		reader:    reader,
+	}
+}

pkg/csv2lp/skip_header_lines_test.go

@@ -0,0 +1,82 @@
+package csv2lp
+
+import (
+	"io"
+	"strconv"
+	"strings"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+// simulates the reader that returns all data together with EOF
+type readOnceWithEOF struct {
+	reader io.Reader
+}
+
+func (r *readOnceWithEOF) Read(p []byte) (n int, err error) {
+	n, _ = r.reader.Read(p)
+	return n, io.EOF
+}
+
+// Test_SkipHeaderLines checks that first lines are skipped
+func Test_SkipHeaderLines(t *testing.T) {
+	input := "1\n2\n3\n4\n5\n6\n7\n8\n9\n0\n"
+
+	var tests = []struct {
+		skipCount int
+		result    string
+	}{
+		{
+			10,
+			"",
+		},
+		{
+			0,
+			"1\n2\n3\n4\n5\n6\n7\n8\n9\n0\n",
+		},
+		{
+			1,
+			"2\n3\n4\n5\n6\n7\n8\n9\n0\n",
+		},
+		{
+			5,
+			"6\n7\n8\n9\n0\n",
+		},
+		{
+			20,
+			"",
+		},
+	}
+	bufferSizes := []int{1, 2, 7, 0, len(input), len(input) + 1}
+
+	for i, test := range tests {
+		for _, bufferSize := range bufferSizes {
+			t.Run(strconv.Itoa(i)+"_"+strconv.Itoa(bufferSize), func(t *testing.T) {
+				var reader io.Reader
+				if bufferSize == 0 {
+					// emulate a reader that returns EOF together with data
+					bufferSize = len(input)
+					reader = SkipHeaderLinesReader(test.skipCount, &readOnceWithEOF{strings.NewReader(input)})
+				} else {
+					reader = SkipHeaderLinesReader(test.skipCount, strings.NewReader(input))
+				}
+				buffer := make([]byte, bufferSize)
+				result := make([]byte, 0, 100)
+				for {
+					n, err := reader.Read(buffer)
+					if n > 0 {
+						result = append(result, buffer[:n]...)
+					}
+					if err != nil {
+						if err != io.EOF {
+							require.Nil(t, err.Error())
+						}
+						break
+					}
+				}
+				require.Equal(t, test.result, string(result))
+			})
+		}
+	}
+}