move: log parsers

This commit is contained in:
Inanc Gumus
2019-08-28 20:23:38 +03:00
parent 0a121cd911
commit 9afbe8f350
123 changed files with 1018 additions and 1515 deletions

View File

@@ -0,0 +1,7 @@
SHELL := /bin/bash
r:
go run . < ../logs/log.txt
t:
time go run . < ../logs/log.txt

View File

@@ -0,0 +1,38 @@
package main
// func chartWriter(w io.Writer) outputFn {
// return func(res []result) error {
// return chartWrite(w, res)
// }
// }
// func chartWrite(w io.Writer, res []result) error {
// sort.Slice(res, func(i, j int) bool {
// return res[i].domain > res[j].domain
// })
// donut := chart.DonutChart{
// Title: "Total Visits Per Domain",
// TitleStyle: chart.Style{
// FontSize: 35,
// Show: true,
// FontColor: chart.ColorAlternateGreen,
// },
// Width: 1920,
// Height: 800,
// }
// for _, r := range res {
// v := chart.Value{
// Label: r.domain + r.page + ": " + strconv.Itoa(r.visits),
// Value: float64(r.visits),
// Style: chart.Style{
// FontSize: 14,
// },
// }
// donut.Values = append(donut.Values, v)
// }
// return donut.Render(chart.SVG, w)
// }

View File

@@ -0,0 +1,33 @@
package main
import (
"errors"
"fmt"
"strconv"
)
// field helps for field parsing
type field struct{ err error }
// uatoi parses an unsigned integer string and saves the error.
// it assumes that the val is unsigned.
// for ease of usability: it returns an int instead of uint.
func (f *field) uatoi(name, val string) int {
n, err := strconv.Atoi(val)
if err != nil || n < 0 {
f.err = fmt.Errorf("incorrect field -> %q = %q", name, val)
}
return n
}
func atoi(input []byte) (int, error) {
val := 0
for i := 0; i < len(input); i++ {
char := input[i]
if char < '0' || char > '9' {
return 0, errors.New("invalid number")
}
val = val*10 + int(char) - '0'
}
return val, nil
}

View File

@@ -0,0 +1,34 @@
package main
import "strings"
func noopFilter(r result) bool {
return true
}
func notUsing(filter filterFn) filterFn {
return func(r result) bool {
return !filter(r)
}
}
func domainExtFilter(domains ...string) filterFn {
return func(r result) bool {
for _, domain := range domains {
if strings.HasSuffix(r.domain, "."+domain) {
return true
}
}
return false
}
}
func domainFilter(domain string) filterFn {
return func(r result) bool {
return strings.Contains(r.domain, domain)
}
}
func orgDomainsFilter(r result) bool {
return strings.HasSuffix(r.domain, ".org")
}

View File

@@ -0,0 +1,20 @@
package main
// domainGrouper groups by domain.
// but it keeps the other fields.
// for example: it returns pages as well, but you shouldn't use them.
// exercise: write a function that erases superfluous data.
func domainGrouper(r result) string {
return r.domain
}
func pageGrouper(r result) string {
return r.domain + r.page
}
// groupBy allocates map unnecessarily
func noopGrouper(r result) string {
// with something like:
// return randomStrings()
return ""
}

View File

@@ -0,0 +1,44 @@
// For more tutorials: https://blog.learngoprogramming.com
//
// Copyright © 2018 Inanc Gumus
// Learn Go Programming Course
// License: https://creativecommons.org/licenses/by-nc-sa/4.0/
//
package main
import (
"fmt"
"os"
)
func main() {
p := pipeline{
read: textReader(os.Stdin),
write: textWriter(os.Stdout),
filter: notUsing(domainExtFilter("io", "com")),
group: domainGrouper,
}
// var p pipeline
// p.
// filterBy(notUsing(domainExtFilter("io", "com"))).
// groupBy(domainGrouper)
if err := p.start(); err != nil {
fmt.Println("> Err:", err)
}
}
// []outputter{textFile("results.txt"), chartFile("graph.png")}
// func outputs(w io.Writer) outputFn {
// tw := textWriter(w)
// cw := chartWriter(w)
// return func(rs []result) error {
// err := tw(rs)
// err = cw(rs)
// return err
// }
// }

View File

@@ -0,0 +1,78 @@
// For more tutorials: https://blog.learngoprogramming.com
//
// Copyright © 2018 Inanc Gumus
// Learn Go Programming Course
// License: https://creativecommons.org/licenses/by-nc-sa/4.0/
//
package main
import "os"
type (
processFn func(r result)
inputFn func(processFn) error
outputFn func([]result) error
filterFn func(result) (include bool)
groupFn func(result) (key string)
)
type pipeline struct {
read inputFn
write outputFn
filter filterFn
group groupFn
}
func (p *pipeline) filterBy(f filterFn) *pipeline { p.filter = f; return p }
func (p *pipeline) groupBy(f groupFn) *pipeline { p.group = f; return p }
func (p *pipeline) from(f inputFn) *pipeline { p.read = f; return p }
func (p *pipeline) to(f outputFn) *pipeline { p.write = f; return p }
func (p *pipeline) defaults() {
if p.filter == nil {
p.filter = noopFilter
}
if p.group == nil {
p.group = domainGrouper
}
if p.read == nil {
p.read = textReader(os.Stdin)
}
if p.write == nil {
p.write = textWriter(os.Stdout)
}
}
func (p *pipeline) start() error {
p.defaults()
// retrieve and process the lines
sum := make(map[string]result)
process := func(r result) {
if !p.filter(r) {
return
}
k := p.group(r)
sum[k] = r.add(sum[k])
}
// return err from input reader
if err := p.read(process); err != nil {
return err
}
// prepare the results for outputting
var out []result
for _, res := range sum {
out = append(out, res)
}
// return err from output reader
return p.write(out)
}

View File

@@ -0,0 +1,83 @@
package main
import (
"fmt"
"strings"
)
const fieldsLength = 4
// result stores the parsed result for a domain
type result struct {
domain string
page string
visits int
uniques int
}
// add adds the metrics of another result
func (r result) add(other result) result {
r.visits += other.visits
r.uniques += other.uniques
return r
}
// parseFields parses and returns the parsing result
func parseFields(line string) (r result, err error) {
fields := strings.Fields(line)
if len(fields) != fieldsLength {
return r, fmt.Errorf("wrong number of fields %q", fields)
}
r.domain = fields[0]
r.page = fields[1]
f := new(field)
r.visits = f.uatoi("visits", fields[2])
r.uniques = f.uatoi("uniques", fields[3])
return r, f.err
}
func fastParseFields(data []byte) (res result, err error) {
const separator = ' '
var findex int
for i, j := 0, 0; i < len(data); i++ {
c := data[i]
last := len(data) == i+1
if c != separator && !last {
continue
}
if last {
i = len(data)
}
switch fval := data[j:i]; findex {
case 0:
res.domain = string(fval)
case 1:
res.page = string(fval)
case 2:
res.visits, err = atoi(fval)
case 3:
res.uniques, err = atoi(fval)
}
if err != nil {
return res, err
}
j = i + 1
findex++
}
if findex != fieldsLength {
err = fmt.Errorf("wrong number of fields %q", data)
}
return res, err
}

View File

@@ -0,0 +1,39 @@
// For more tutorials: https://blog.learngoprogramming.com
//
// Copyright © 2018 Inanc Gumus
// Learn Go Programming Course
// License: https://creativecommons.org/licenses/by-nc-sa/4.0/
//
package main
import (
"bufio"
"fmt"
"io"
)
func textReader(r io.Reader) inputFn {
return func(process processFn) error {
var (
l = 1
in = bufio.NewScanner(r)
)
for in.Scan() {
r, err := fastParseFields(in.Bytes())
// r, err := parseFields(in.Text())
if err != nil {
return fmt.Errorf("line %d: %v", l, err)
}
process(r)
l++
}
if c, ok := r.(io.Closer); ok {
c.Close()
}
return in.Err()
}
}

View File

@@ -0,0 +1,50 @@
package main
import (
"fmt"
"io"
"sort"
"strings"
)
// TODO: sort by result key interfaces section
const (
// DOMAINS PAGES VISITS UNIQUES
// ^ ^ ^ ^
// | | | |
header = "%-25s %-10s %10s %10s\n"
line = "%-25s %-10s %10d %10d\n"
footer = "\n%-36s %10d %10d\n" // -> "" VISITS UNIQUES
dash = "-"
dashLength = 58
)
func textWriter(w io.Writer) outputFn {
return func(res []result) error {
sort.Slice(res, func(i, j int) bool {
return res[i].domain > res[j].domain
})
var total result
fmt.Fprintf(w, header, "DOMAINS", "PAGES", "VISITS", "UNIQUES")
fmt.Fprintf(w, strings.Repeat(dash, dashLength)+"\n")
for _, r := range res {
total = total.add(r)
fmt.Fprintf(w, line, r.domain, r.page, r.visits, r.uniques)
}
fmt.Fprintf(w, footer, "", total.visits, total.uniques)
return nil
}
}
func noWhere() outputFn {
return func(res []result) error {
return nil
}
}