How to reorder a CSV file to group by contents of a particular column

user_657189 0
Viewed 74

I am very new go Golang and my question is not cleared also, but this is what I am trying to achieve. I have a csv file as follow, as I am mainly trying to re-arrange/sort last column(status=passed,failed/skipped)

test,test-cat,skipped
test,test-cat,failed
test,test-cat,passed
test,test-cat,skipped
test,test-cat,passed
test,test-cat,failed

Expecting last column to be grouped them together if it has same status.

test,test-cat,skipped
test,test-cat,skipped
test,test-cat,failed
test,test-cat,failed
test,test-cat,passed
test,test-cat,passed

With this codes I did, it does not look good:-) but it works as I wanted.

package main
import (
        "bufio"
        "fmt"
        "os"
        "strings"
)
func main() {
        var FailStat, SkipStat,PassStat []string
      
        file, err := os.Open("test.csv")

        if err != nil {
                fmt.Println(err)
        } else {
                scanner := bufio.NewScanner(file)
                for scanner.Scan() {
                        line := scanner.Text()
                        if strings.Contains(line, "failed") {
                                FailStat = append(FailStat, line)

                        }
                        if strings.Contains(line, "skipped") {
                                SkipStat = append(SkipStat, line)

                        }
                        if strings.Contains(line, "passed") {
                                PassStat = append(PassStat, line)

                        }                       
                }
        }
        file.Close()

        var finalstat []string
        finalstat = append(SkipStat, FailStat...)
        finalstat = append(finalstat, PassStat...)

        for _, line := range finalstat {
           fmt.Println(line)
   }
}

Test-Run:

$ ./readfile 
test,test-cat,skipped
test,test-cat,skipped
test,test-cat,failed
test,test-cat,failed
test,test-cat,passed
test,test-cat,passed

There must be a many better ways, please advice. Sorry for newbie question!

go csv
2 Answers

Inian's solution will work if the order of the status groupings doesn't matter (because of map's design, you should never expect to get the same ordering of the groups from run to run).

If you need the groups consistently ordered, that is actually sorted:

package main

import (
    "encoding/csv"
    "io"
    "log"
    "os"
    "sort"
    "strings"
)

type Row struct {
    Name, Category, Status string
}

func main() {
    in := `test,test-cat,skipped
test,test-cat,failed
test,test-cat,passed
test,test-cat,skipped
test,test-cat,passed
test,test-cat,failed
`
    r := csv.NewReader(strings.NewReader(in))

    rows := make([]Row, 0)
    for {
        record, err := r.Read()
        if err == io.EOF {
            break
        }
        if err != nil {
            log.Fatal(err)
        }
        row := Row{record[0], record[1], record[2]}
        rows = append(rows, row)
    }

    sort.Slice(rows, func(i, j int) bool { return rows[i].Status < rows[j].Status })

    w := csv.NewWriter(os.Stdout)

    for _, row := range rows {
        w.Write([]string{row.Name, row.Category, row.Status})
    }
    w.Flush()

    if err := w.Error(); err != nil {
        log.Fatal(err)
    }
}

and we get:

test,test-cat,failed
test,test-cat,failed
test,test-cat,passed
test,test-cat,passed
test,test-cat,skipped
test,test-cat,skipped

Change the < to > in the anonymous func for sort.Slice to reverse the order of the sort.

Go Playground

If you don't want to mess with the Row struct and convert between []Row and [][]string:

// ...
rows := make([][]string, 0)
for {
    row, err := r.Read()
    // ...
    rows = append(rows, row)
}

sort.Slice(rows, func(i, j int) bool { return rows[i][2] < rows[j][2] })

w := csv.NewWriter(os.Stdout)

for _, row := range rows {
    w.Write(row)
}
// ...

Go Playground

In a comment you mentioned wanting a specific order of the groups, and now I can see in your original code what you were aiming for

In which case Ianian's solution is going the right direction:

    // ...

    recordGroups := make(map[string][][]string)
    for {
        records, err := r.Read()
        if err == io.EOF {
            break
        }
        if err != nil {
            log.Fatal(err)
        }
        groupName := records[2]
        recordGroups[groupName] = append(recordGroups[groupName], records)
    }
    w := csv.NewWriter(os.Stdout)

    // Control the order with this slice of group names
    groupNames := []string{"failed", "passed", "skipped", "Bogus group!"}

    for _, groupName := range groupNames {
        recordGroup, ok := recordGroups[groupName]
        if !ok {
            log.Printf("did not find expected group %q\n", groupName)
            continue
        }
        for _, record := range recordGroup {
            if err := w.Write(record); err != nil {
                log.Fatalln("error writing record to csv:", err)
            }
        }
    }

    // ...

2009/11/10 23:00:00 did not find expected group "Bogus group!"
test,test-cat,failed
test,test-cat,failed
test,test-cat,skipped
test,test-cat,skipped
test,test-cat,passed
test,test-cat,passed

Go Playground

user_2468010

It is better of to the csv package provided in the standard library for this purpose. The logic involves creating a map of string to a slice of strings, where the key will be the column you want to group on and the value being the list of rows that are unique to it.

Once you populate the map, the subsequent action would be to print the result back in CSV format. The below example involves reading the input from a variable and printing back to stdout. You can refer to the other methods in the package to perform the same on a text file.

package main

import (
    "encoding/csv"
    "io"
    "log"
    "os"
    "strings"
)

func main() {
    in := `test,test-cat,skipped
test,test-cat,failed
test,test-cat,passed
test,test-cat,skipped
test,test-cat,passed
test,test-cat,failed
`
    r := csv.NewReader(strings.NewReader(in))
    dictMap := make(map[string][][]string)
    for {
        records, err := r.Read()
        if err == io.EOF {
            break
        }
        if err != nil {
            log.Fatal(err)
        }
        dictMap[records[2]] = append(dictMap[records[2]], records)
    }

    w := csv.NewWriter(os.Stdout)

    for _, records := range dictMap {
        for idx := range records {
            if err := w.Write(records[idx]); err != nil {
                log.Fatalln("error writing record to csv:", err)
            }
        }
    }

    w.Flush()

    if err := w.Error(); err != nil {
        log.Fatal(err)
    }
}

go-playground

user_52910150
Related