consolidated go and scala repos into this repo as well

go/pkg/io/file_tools.go

@@ -0,0 +1,86 @@
+package io
+
+import (
+	"bufio"
+	"fmt"
+	"os"
+	"strings"
+)
+
+func FileToStringArray(filename string) ([]string, error) {
+	file, err := os.Open(filename)
+	if err != nil {
+		return nil, err
+	}
+	defer file.Close()
+
+	scanner := bufio.NewScanner(file)
+	scanner.Split(bufio.ScanLines)
+
+	var fileLines []string
+
+	for scanner.Scan() {
+		fileLines = append(fileLines, scanner.Text())
+	}
+	if err := scanner.Err(); err != nil {
+		return nil, err
+	}
+
+	file.Close()
+
+	return fileLines, nil
+}
+
+func WriteStringsToFile(content []string, filepath string) error {
+	file, err := os.Create(filepath)
+	if err != nil {
+		return err
+	}
+
+	w := bufio.NewWriter(file)
+
+	for _, line := range content {
+		_, err = fmt.Fprintln(w, line)
+		if err != nil {
+			return err
+		}
+	}
+
+	if err = w.Flush(); err != nil {
+		return err
+	}
+	if err = file.Close(); err != nil {
+		return err
+	}
+
+	return nil
+}
+
+// ParseFasta parses a .fasta formatted file returning a list of labels and a list of
+// content in separate string arrays.
+func ParseFasta(filename string) ([]string, []string, error) {
+	lines, err := FileToStringArray(filename)
+	if err != nil {
+		return nil, nil, err
+	}
+
+	var labels []string
+	var content []string
+	var currString []string
+	for _, line := range lines {
+		if line[0] == '>' {
+			labels = append(labels, line)
+			if len(currString) > 0 {
+				content = append(content, strings.Join(currString, ""))
+				currString = []string{}
+			}
+			continue
+		}
+		currString = append(currString, line)
+	}
+	if len(currString) > 0 {
+		content = append(content, strings.Join(currString, ""))
+	}
+
+	return labels, content, nil
+}