总结csv的内容

Question

Context I'm working on creating a little program that can summarize the contents of an absolute mess of a bill, which is in csv form.

The bill has three columns I'm interested in:

1. Event type. Here, I'm only interested in the rows where this column reads CHARGE
2. The cost. Self explanatory.
3. Resource name, containing Server and cluster names. The format is servername.clustername.

The idea is to select the rows that are labeled as charge, split them up first by cluster and then by server name, and sum up the total costs for each.

I can't help but feel like this should be easy, but I've been scratching my head on this for a while now, and just can't seem to figure it out. At this point I ought to state that I am fairly new to programming and entirely new to GO.

Here's what I have so far:

package main

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



func main() {
    rows := readBill("bill-2018-April.csv")
    rows = calculateSummary(rows)
    writeSummary("bill-2018-April-output", rows)

}

func readBill(name string) [][]string {

    f, err := os.Open(name)

    if err != nil {
        log.Fatalf("Cannot open '%s': %s
", name, err.Error())
    }

    defer f.Close()

    r := csv.NewReader(f)

    rows, err := r.ReadAll()

    if err != nil {
        log.Fatalln("Cannot read CSV data:", err.Error())
    }

    return rows
}

type charges struct {
    impactType string
    cost       float64
    resName    string
}
func createCharges(rows [][]string){
    charges:= []charges{}
    for i,r:=range rows {
        var c charges
        c.impactType :=r [i][10]
        c.cost := r [i][15]
        c.resName := r [i][20]
        charges = append()
    }
    return charges
} 

So, as far as I can tell, I should now have isolated the columns I am interested in (i.e. columns 10, 15 and 20). Is what I have so far even correct?

How would I go about singling out the rows reading "CHARGE" and slicing everything up by cluster and server?

Summing things up shouldn't be too tricky, but for whatever reason, this is really stumping me.

Answer 1

Just use two maps to store the sums per server and per cluster. And since you're not interested in the whole CSV but only some rows, reading everything is kind of wasteful. Just skip the rows you don't care about:

package main

import (
    "encoding/csv"
    "fmt"
    "io"
    "log"
    "strconv"
    "strings"
)

func main() {
    b := `
,,,,,,,,,,CHARGE,,,,,100.00,,,,,s1.c1
,,,,,,,,,,IGNORE,,,,,,,,,,
,,,,,,,,,,CHARGE,,,,,200.00,,,,,s2.c1
,,,,,,,,,,CHARGE,,,,,300.00,,,,,s3.c2
`

    r := csv.NewReader(strings.NewReader(b))

    byServer := make(map[string]float64)
    byCluster := make(map[string]float64)

    for i := 0; ; i++ {
        row, err := r.Read()
        if err == io.EOF {
            break
        }
        if err != nil {
            log.Fatal(err)
        }

        if row[10] != "CHARGE" {
            continue
        }

        cost, err := strconv.ParseFloat(row[15], 64)
        if err != nil {
            log.Fatalf("row %d: malformed cost: %v", i, err)
        }

        xs := strings.SplitN(row[20], ".", 2)
        if len(xs) != 2 {
            log.Fatalf("row %d: malformed resource name", i)
        }

        server, cluster := xs[0], xs[1]

        byServer[server] += cost
        byCluster[cluster] += cost
    }

    fmt.Printf("byServer: %+v
", byServer)
    fmt.Printf("byCluster: %+v
", byCluster)
}

// Output:
// byServer: map[s2:200 s3:300 s1:100]
// byCluster: map[c1:300 c2:300]

Try it on the playground: https://play.golang.org/p/1e9mJf4LyYE