remove: logparser v6

This commit is contained in:
Inanc Gumus
2019-08-28 23:46:28 +03:00
parent 10140e976d
commit 4629b59ef2
20 changed files with 0 additions and 777 deletions

View File

@ -1,5 +0,0 @@
r:
go run . < ../../logs/log.txt
t:
time go run . < ../../logs/log.txt

View File

@ -1,42 +0,0 @@
// 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"
"strings"
"github.com/inancgumus/learngo/logparser/v6/pipe"
"github.com/inancgumus/learngo/logparser/v6/pipe/group"
"github.com/inancgumus/learngo/logparser/v6/pipe/parse"
"github.com/inancgumus/learngo/logparser/v6/pipe/report"
)
// fromFile generates a default pipeline.
// Detects the correct parser by the file extension.
// Uses a TextReport and groups by domain.
func fromFile(path string) (*pipe.Pipeline, error) {
f, err := os.Open(path)
if err != nil {
return nil, err
}
var src pipe.Iterator
switch {
case strings.HasSuffix(path, ".txt"):
src = parse.FromText(f)
case strings.HasSuffix(path, ".jsonl"):
src = parse.FromJSON(f)
}
return pipe.New(
src,
report.AsText(os.Stdout),
group.By(group.Domain),
), nil
}

View File

@ -1,49 +0,0 @@
// 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 (
"log"
"os"
"github.com/inancgumus/learngo/logparser/v6/pipe"
"github.com/inancgumus/learngo/logparser/v6/pipe/filter"
"github.com/inancgumus/learngo/logparser/v6/pipe/group"
"github.com/inancgumus/learngo/logparser/v6/pipe/parse"
"github.com/inancgumus/learngo/logparser/v6/pipe/report"
)
func main() {
pipe := pipe.New(
parse.FromText(os.Stdin),
// parse.FromJSON(os.Stdin),
report.AsText(os.Stdout),
filter.By(filter.Not(filter.DomainExt("com", "io"))),
group.By(group.Domain),
new(logger),
)
if err := pipe.Run(); err != nil {
log.Fatalln(err)
}
}
type logger struct {
src pipe.Iterator
}
func (l *logger) Consume(records pipe.Iterator) error {
l.src = records
return nil
}
func (l *logger) Each(yield func(pipe.Record)) error {
return l.src.Each(func(r pipe.Record) {
yield(r)
})
}

View File

@ -1,38 +0,0 @@
// 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 filter
import (
"strings"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
// DomainExt filters a set of domain extensions.
func DomainExt(domains ...string) Func {
return func(r pipe.Record) bool {
for _, domain := range domains {
if strings.HasSuffix(r.Str("domain"), "."+domain) {
return true
}
}
return false
}
}
// Domain filters a domain if it contains the given text.
func Domain(text string) Func {
return func(r pipe.Record) bool {
return strings.Contains(r.Str("domain"), text)
}
}
// DomainOrg filters only the ".org" domains.
func DomainOrg(r pipe.Record) bool {
return strings.HasSuffix(r.Str("domain"), ".org")
}

View File

@ -1,50 +0,0 @@
// 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 filter
import "github.com/inancgumus/learngo/logparser/v6/pipe"
// Func represents a filtering pipeline func.
type Func func(pipe.Record) (pass bool)
// Filter the records.
type Filter struct {
src pipe.Iterator
filters []Func
}
// By returns a new filter pipeline.
func By(fn ...Func) *Filter {
return &Filter{filters: fn}
}
// Consume saves the iterator for later processing.
func (f *Filter) Consume(records pipe.Iterator) error {
f.src = records
return nil
}
// Each yields only the filtered records.
func (f *Filter) Each(yield func(pipe.Record)) error {
return f.src.Each(func(r pipe.Record) {
if !f.check(r) {
return
}
yield(r)
})
}
// check all the filters against the record.
func (f *Filter) check(r pipe.Record) bool {
for _, fi := range f.filters {
if !fi(r) {
return false
}
}
return true
}

View File

@ -1,15 +0,0 @@
// 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 filter
import "github.com/inancgumus/learngo/logparser/v6/pipe"
// Noop filter that does nothing.
func Noop(r pipe.Record) bool {
return true
}

View File

@ -1,17 +0,0 @@
// 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 filter
import "github.com/inancgumus/learngo/logparser/v6/pipe"
// Not reverses a filter. True becomes false, and vice versa.
func Not(filter Func) Func {
return func(r pipe.Record) bool {
return !filter(r)
}
}

View File

@ -1,18 +0,0 @@
// 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 group
import "github.com/inancgumus/learngo/logparser/v6/pipe"
// Domain groups the records by domain.
// It keeps the other fields intact.
// For example: It returns the page field as well.
// Exercise: Write a solution that removes the unnecessary data.
func Domain(r pipe.Record) string {
return r.Str("domain")
}

View File

@ -1,60 +0,0 @@
// 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 group
import (
"sort"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
// Func represents a grouping func that returns a grouping key.
type Func func(pipe.Record) (key string)
// Group records by a key.
type Group struct {
sum map[string]pipe.Record // metrics per group key
keys []string // unique group keys
key Func
}
// By returns a new Group.
// It takes a group func that returns a group key.
// The returned group will group the record using the key.
func By(key Func) *Group {
return &Group{
sum: make(map[string]pipe.Record),
key: key,
}
}
// Consume records for grouping.
func (g *Group) Consume(records pipe.Iterator) error {
return records.Each(func(r pipe.Record) {
k := g.key(r)
if _, ok := g.sum[k]; !ok {
g.keys = append(g.keys, k)
}
if r, ok := r.(pipe.Summer); ok {
g.sum[k] = r.Sum(g.sum[k])
}
})
}
// Each sorts and yields the grouped records.
func (g *Group) Each(yield func(pipe.Record)) error {
sort.Strings(g.keys)
for _, k := range g.keys {
yield(g.sum[k])
}
return nil
}

View File

@ -1,15 +0,0 @@
// 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 group
import "github.com/inancgumus/learngo/logparser/v6/pipe"
// Page groups records by page.
func Page(r pipe.Record) string {
return r.Str("domain") + r.Str("page")
}

View File

@ -1,36 +0,0 @@
// 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 pipe
import "fmt"
// logCount counts the yielded records.
type logCount struct {
Iterator
n int
}
// Each yields to the inner iterator while counting the records.
// Reports the record number on an error.
func (lc *logCount) Each(yield func(Record)) error {
err := lc.Iterator.Each(func(r Record) {
lc.n++
yield(r)
})
if err != nil {
// lc.n+1: iterator.each won't call yield on err
return fmt.Errorf("record %d: %v", lc.n+1, err)
}
return nil
}
// count returns the last read record number.
func (lc *logCount) count() int {
return lc.n
}

View File

@ -1,19 +0,0 @@
// 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 parse
import (
"io"
)
// readClose the reader if it's a io.Closer.
func readClose(r io.Reader) {
if rc, ok := r.(io.Closer); ok {
rc.Close()
}
}

View File

@ -1,47 +0,0 @@
// For more tutorials: https://bj.learngoprogramming.com
//
// Copyright © 2018 Inanc Gumus
// Learn Go Programming Course
// License: https://creativecommons.org/licenses/by-nc-sa/4.0/
//
package parse
import (
"encoding/json"
"io"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
// JSON parses json records.
type JSON struct {
reader io.Reader
}
// FromJSON creates a json parser.
func FromJSON(r io.Reader) *JSON {
return &JSON{reader: r}
}
// Each yields records from a json reader.
func (j *JSON) Each(yield func(pipe.Record)) error {
defer readClose(j.reader)
dec := json.NewDecoder(j.reader)
for {
var r record
err := dec.Decode(&r)
if err == io.EOF {
break
}
if err != nil {
return err
}
yield(r)
}
return nil
}

View File

@ -1,116 +0,0 @@
package parse
import (
"encoding/json"
"errors"
"fmt"
"strconv"
"strings"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
const fieldsLength = 4
// record stores fields of a log line.
type record struct {
Domain string
Page string
Visits int
Uniques int
}
// Str gets a string field by name.
func (r record) Str(field string) string {
switch field {
case "domain":
return r.Domain
case "page":
return r.Page
}
panic(fieldErr(field))
}
// Int gets an integer field by name.
func (r record) Int(field string) int {
switch field {
case "visits":
return r.Visits
case "uniques":
return r.Uniques
}
panic(fieldErr(field))
}
// Sum the numeric fields with another record.
func (r record) Sum(other pipe.Record) pipe.Record {
if other == nil {
return r
}
r.Visits += other.(record).Visits
r.Uniques += other.(record).Uniques
return r
}
// UnmarshalText to a *record.
func (r *record) UnmarshalText(p []byte) (err error) {
fields := strings.Fields(string(p))
if len(fields) != fieldsLength {
return fmt.Errorf("wrong number of fields %q", fields)
}
r.Domain, r.Page = fields[0], fields[1]
if r.Visits, err = parseStr("visits", fields[2]); err != nil {
return err
}
if r.Uniques, err = parseStr("uniques", fields[3]); err != nil {
return err
}
return validate(*r)
}
// UnmarshalJSON to a *record.
func (r *record) UnmarshalJSON(data []byte) error {
// `methodless` doesn't have any methods including UnmarshalJSON.
// This trick prevents the stack-overflow (infinite loop).
type methodless record
var m methodless
if err := json.Unmarshal(data, &m); err != nil {
return err
}
// Cast back to the record and save.
*r = record(m)
return validate(*r)
}
// parseStr helps UnmarshalText for string to positive int parsing.
func parseStr(name, v string) (int, error) {
n, err := strconv.Atoi(v)
if err != nil {
return 0, fmt.Errorf("Record.UnmarshalText %q: %v", name, err)
}
return n, nil
}
// validate whether a parsed record is valid or not.
func validate(r record) (err error) {
switch {
case r.Domain == "":
err = errors.New("record.domain cannot be empty")
case r.Page == "":
err = errors.New("record.page cannot be empty")
case r.Visits < 0:
err = errors.New("record.visits cannot be negative")
case r.Uniques < 0:
err = errors.New("record.uniques cannot be negative")
}
return
}
func fieldErr(field string) error {
return fmt.Errorf("record field: %q does not exist", field)
}

View File

@ -1,44 +0,0 @@
// For more tutorials: https://bp.learngoprogramming.com
//
// Copyright © 2018 Inanc Gumus
// Learn Go Programming Course
// License: https://creativecommons.org/licenses/by-nc-sa/4.0/
//
package parse
import (
"bufio"
"io"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
// Text parses text based log lines.
type Text struct {
reader io.Reader
}
// FromText creates a text parser.
func FromText(r io.Reader) *Text {
return &Text{reader: r}
}
// Each yields records from a text log.
func (p *Text) Each(yield func(pipe.Record)) error {
defer readClose(p.reader)
in := bufio.NewScanner(p.reader)
for in.Scan() {
r := new(record)
if err := r.UnmarshalText(in.Bytes()); err != nil {
return err
}
yield(r)
}
return in.Err()
}

View File

@ -1,26 +0,0 @@
// 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 pipe
// Iterator yields a record.
type Iterator interface {
Each(func(Record)) error
}
// Consumer consumes records from an iterator.
type Consumer interface {
Consume(Iterator) error
}
// Transform represents both a record consumer and producer.
// It has an input and output.
// It takes a single record and provides an iterator for all the records.
type Transform interface {
Consumer
Iterator // producer
}

View File

@ -1,48 +0,0 @@
// 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 pipe
import (
"fmt"
"os"
)
// Pipeline takes records from a source, transforms, and sends them to a destionation.
type Pipeline struct {
src Iterator
trans []Transform
dst Consumer
}
// New creates a new pipeline.
func New(src Iterator, dst Consumer, t ...Transform) *Pipeline {
return &Pipeline{
src: &logCount{Iterator: src},
dst: dst,
trans: t,
}
}
// Run the pipeline.
func (p *Pipeline) Run() error {
defer func() {
n := p.src.(*logCount).count()
fmt.Fprintf(os.Stderr, "%d records processed.\n", n)
}()
last := p.src
for _, t := range p.trans {
if err := t.Consume(last); err != nil {
return err
}
last = t
}
return p.dst.Consume(last)
}

View File

@ -1,19 +0,0 @@
// 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 pipe
// Record provides a generic interface for any sort of records.
type Record interface {
Str(field string) string
Int(field string) int
}
// Summer provides a method for summing the numeric fields.
type Summer interface {
Sum(Record) Record
}

View File

@ -1,49 +0,0 @@
package report
/*
// You need to run:
// go get -u github.com/wcharczuk/go-chart
// Chart renders a chart.
type Chart struct {
Title string
Width, Height int
w io.Writer
}
// AsChart returns a Chart report generator.
func AsChart(w io.Writer) *Chart {
return &Chart{w: w}
}
// Consume generates a chart report.
func (c *Chart) Consume(records pipe.Iterator) error {
w := os.Stdout
donut := chart.DonutChart{
Title: c.Title,
TitleStyle: chart.Style{
FontSize: 35,
Show: true,
FontColor: chart.ColorAlternateGreen,
},
Width: c.Width,
Height: c.Height,
}
records.Each(func(r pipe.Record) {
v := chart.Value{
Label: r.Str("domain") + r.Str("page") + ": " + strconv.Itoa(r.Int("visits")),
Value: float64(r.Int("visits")),
Style: chart.Style{
FontSize: 14,
},
}
donut.Values = append(donut.Values, v)
})
return donut.Render(chart.SVG, w)
}
*/

View File

@ -1,64 +0,0 @@
// 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 report
import (
"fmt"
"io"
"text/tabwriter"
"github.com/inancgumus/learngo/logparser/v6/pipe"
)
const (
minWidth = 0
tabWidth = 4
padding = 4
flags = 0
)
// Text report generator.
type Text struct {
w io.Writer
}
// AsText returns a Text report generator.
func AsText(w io.Writer) *Text {
return &Text{w: w}
}
// Consume generates a text report.
func (t *Text) Consume(records pipe.Iterator) error {
w := tabwriter.NewWriter(t.w, minWidth, tabWidth, padding, ' ', flags)
write := fmt.Fprintf
write(w, "DOMAINS\tPAGES\tVISITS\tUNIQUES\n")
write(w, "-------\t-----\t------\t-------\n")
var total pipe.Record
records.Each(func(r pipe.Record) {
if r, ok := r.(pipe.Summer); ok {
total = r.Sum(total)
}
write(w, "%s\t%s\t%d\t%d\n",
r.Str("domain"), r.Str("page"),
r.Int("visits"), r.Int("uniques"),
)
})
write(w, "\t\t\t\n")
write(w, "%s\t%s\t%d\t%d\n", "TOTAL", "",
total.Int("visits"),
total.Int("uniques"),
)
return w.Flush()
}