feat: add input(es) source support

.github/workflows/build.yml

@@ -16,6 +16,9 @@ jobs:
       - name: checkout repository
         uses: actions/checkout@v4
 
+      - name: fill version
+        run: sed -i -E "s/v[0-9]+.[0-9]+.[0-9]+/${{ github.ref_name }}/g" internal/opt/version.go
+
       - name: install golang
         uses: actions/setup-go@v4
         with:
@@ -52,5 +55,4 @@ jobs:
             dist/esgo2dump_${{ github.ref_name }}_windows_amd64.exe
             dist/esgo2dump_${{ github.ref_name }}_windows_arm64.exe
             dist/esgo2dump_${{ github.ref_name }}_darwin_amd64
-            dist/esgo2dump_${{ github.ref_name }}_darwin_amd64
             dist/esgo2dump_${{ github.ref_name }}_darwin_arm64

go.mod

@@ -4,6 +4,7 @@ go 1.18
 
 require (
 	github.com/elastic/go-elasticsearch/v7 v7.17.10
+	github.com/samber/lo v1.39.0
 	github.com/sirupsen/logrus v1.9.3
 	github.com/spf13/cobra v1.8.0
 )
@@ -12,5 +13,6 @@ require (
 	github.com/inconshreveable/mousetrap v1.1.0 // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
 	github.com/stretchr/testify v1.8.4 // indirect
+	golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17 // indirect
 	golang.org/x/sys v0.14.0 // indirect
 )

go.sum

@@ -9,6 +9,8 @@ github.com/inconshreveable/mousetrap v1.1.0/go.mod h1:vpF70FUmC8bwa3OWnCshd2FqLf
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
+github.com/samber/lo v1.39.0 h1:4gTz1wUhNYLhFSKl6O+8peW0v2F4BCY034GRpU9WnuA=
+github.com/samber/lo v1.39.0/go.mod h1:+m/ZKRl6ClXCE2Lgf3MsQlWfh4bn1bz6CXEOxnEXnEA=
 github.com/sirupsen/logrus v1.9.3 h1:dueUQJ1C2q9oE3F7wvmSGAaVtTmUizReu6fjN8uqzbQ=
 github.com/sirupsen/logrus v1.9.3/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
 github.com/spf13/cobra v1.8.0 h1:7aJaZx1B85qltLMc546zn58BxxfZdR/W22ej9CFoEf0=
@@ -19,6 +21,8 @@ github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+
 github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.8.4 h1:CcVxjf3Q8PM0mHUKJCdn+eZZtm5yQwehR5yeSVQQcUk=
 github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXlSw2iwfAo=
+golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17 h1:3MTrJm4PyNL9NBqvYDSj3DHl46qQakyfqfWo4jgfaEM=
+golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17/go.mod h1:lgLbSvA5ygNOMpwM/9anMpWVlVJ7Z+cHWq/eFuinpGE=
 golang.org/x/sys v0.0.0-20220715151400-c0bba94af5f8/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.14.0 h1:Vz7Qs629MkJkGyHxUlRHizWJRG2j8fbQKjELVSNhy7Q=
 golang.org/x/sys v0.14.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=

internal/cmd/cmd.go

@@ -21,6 +21,8 @@ esgo2dump --input=http://127.0.0.1:9200/some_index --output=http://192.168.1.1:9
 
 esgo2dump --input=https://username:password@127.0.0.1:9200/some_index --output=./data.json
 
+esgo2dump --input=http://127.0.0.1:9200/some_index --source='id;name;age;address' --output=./data.json
+
 esgo2dump --input=http://127.0.0.1:9200/some_index --output=./data.json --query='{"match": {"name": "some_name"}}'
 
 esgo2dump --input=http://127.0.0.1:9200/some_index --output=./data.json --query_file=my_queries.json`,
@@ -30,18 +32,23 @@ esgo2dump --input=http://127.0.0.1:9200/some_index --output=./data.json --query_
 	f_output string
 	f_limit  int
 	f_type   string
+	f_source string
 	f_query  string
 
 	f_query_file string
+
+	f_version bool
 )
 
 func init() {
 	rootCommand.Flags().BoolVar(&opt.Debug, "debug", false, "")
+	rootCommand.Flags().BoolVarP(&f_version, "version", "v", false, "print esgo2dump version")
 	rootCommand.Flags().IntVar(&opt.Timeout, "timeout", 30, "max timeout seconds per operation with limit")
 
 	rootCommand.Flags().StringVarP(&f_input, "input", "i", "", "*required: input file or es url (example :data.json / http://127.0.0.1:9200/my_index)")
 	rootCommand.Flags().StringVarP(&f_output, "output", "o", "output.json", "")
 	rootCommand.Flags().StringVarP(&f_type, "type", "t", "data", "data/mapping/setting")
+	rootCommand.Flags().StringVarP(&f_source, "source", "s", "", "query source, use ';' to separate")
 	rootCommand.Flags().StringVarP(&f_query, "query", "q", "", `query dsl, example: {"bool":{"must":[{"term":{"name":{"value":"some_name"}}}],"must_not":[{"range":{"age":{"gte":18,"lt":60}}}]}}`)
 	rootCommand.Flags().StringVar(&f_query_file, "query_file", "", `query json file (will execute line by line)`)
 	rootCommand.Flags().IntVarP(&f_limit, "limit", "l", 100, "")

internal/cmd/run.go

@@ -7,11 +7,13 @@ import (
 	"fmt"
 	"net/url"
 	"os"
+	"strings"
 
 	"github.com/loveuer/esgo2dump/internal/interfaces"
 	"github.com/loveuer/esgo2dump/internal/opt"
 	"github.com/loveuer/esgo2dump/internal/xes"
 	"github.com/loveuer/esgo2dump/internal/xfile"
+	"github.com/samber/lo"
 	"github.com/sirupsen/logrus"
 	"github.com/spf13/cobra"
 )
@@ -50,6 +52,11 @@ func run(cmd *cobra.Command, args []string) error {
 		logrus.SetLevel(logrus.DebugLevel)
 	}
 
+	if f_version {
+		logrus.Infof("esgo2dump (Version: %s)", opt.Version)
+		return nil
+	}
+
 	if err = check(cmd); err != nil {
 		return err
 	}
@@ -71,6 +78,10 @@ func run(cmd *cobra.Command, args []string) error {
 		return fmt.Errorf("with file input, query or query_file can't be supported")
 	}
 
+	if (f_source != "") && ioi.IsFile() {
+		return fmt.Errorf("with file input, source can't be supported")
+	}
+
 	switch f_type {
 	case "data":
 		if err = executeData(cmd.Context(), ioi, ioo); err != nil {
@@ -117,8 +128,15 @@ func executeData(ctx context.Context, input, output interfaces.DumpIO) error {
 		ch      = make(chan []*interfaces.ESSource, 1)
 		errCh   = make(chan error)
 		queries = make([]map[string]any, 0)
+		sources = make([]string, 0)
 	)
 
+	if f_source != "" {
+		sources = lo.Map(strings.Split(f_source, ";"), func(item string, idx int) string {
+			return strings.TrimSpace(item)
+		})
+	}
+
 	if f_query != "" {
 		query := make(map[string]any)
 		if err = json.Unmarshal([]byte(f_query), &query); err != nil {
@@ -142,6 +160,7 @@ func executeData(ctx context.Context, input, output interfaces.DumpIO) error {
 		}()
 
 		scanner := bufio.NewScanner(qf)
+		scanner.Buffer(make([]byte, 1*1024*1024), 5*1024*1024)
 		lineCount := 1
 		for scanner.Scan() {
 			line := scanner.Text()
@@ -174,24 +193,33 @@ func executeData(ctx context.Context, input, output interfaces.DumpIO) error {
 			close(ch)
 		}()
 
+	Loop:
 		for _, query := range queries {
+			for {
 				select {
 				case <-c.Done():
 					return
 				default:
-				if lines, err = input.ReadData(c, f_limit, query); err != nil {
+					if lines, err = input.ReadData(c, f_limit, query, sources); err != nil {
 						errCh <- err
 						return
 					}
 
+					logrus.Debugf("executeData: input read_data got lines=%d", len(lines))
+
 					if len(lines) == 0 {
 						input.ResetOffset()
-					continue
+						if query != nil {
+							bs, _ := json.Marshal(query)
+							logrus.Infof("Dump: query_file query=%s read done!!!", string(bs))
+						}
+						continue Loop
 					}
 
 					ch <- lines
 				}
 			}
+		}
 	}(ctx)
 
 	var (
@@ -219,6 +247,8 @@ func executeData(ctx context.Context, input, output interfaces.DumpIO) error {
 				return err
 			}
 
+			logrus.Debugf("executeData: output write_data succeed lines=%d", succeed)
+
 			if succeed != len(docs) {
 				return fmt.Errorf("cmd.run: got lines=%d, only succeed=%d", len(docs), succeed)
 			}

internal/interfaces/dumpio.go

@@ -3,7 +3,7 @@ package interfaces
 import "context"
 
 type DumpIO interface {
-	ReadData(context.Context, int, map[string]any) ([]*ESSource, error)
+	ReadData(context.Context, int, map[string]any, []string) ([]*ESSource, error)
 	WriteData(ctx context.Context, docs []*ESSource) (int, error)
 
 	ResetOffset()

internal/opt/var.go

@@ -7,4 +7,7 @@ const (
 var (
 	Debug   bool
 	Timeout int
+
+	BuffSize    = 5 * 1024 * 1024   // 5M
+	MaxBuffSize = 100 * 1024 * 1024 // 100M, default elastic_search doc max size
 )

internal/opt/version.go

@@ -0,0 +1,3 @@
+package opt
+
+const Version = "v0.1.2"

internal/xes/xes.go

@@ -128,7 +128,26 @@ func (c *client) Close() error {
 }
 
 func (c *client) ResetOffset() {
+	defer func() {
 		c.scrollId = ""
+	}()
+
+	bs, _ := json.Marshal(map[string]string{
+		"scroll_id": c.scrollId,
+	})
+
+	rr, err := c.c.ClearScroll(
+		c.c.ClearScroll.WithContext(util.Timeout(3)),
+		c.c.ClearScroll.WithBody(bytes.NewReader(bs)),
+	)
+	if err != nil {
+		logrus.Warnf("ResetOffset: clear scroll id=%s err=%v", c.scrollId, err)
+		return
+	}
+
+	if rr.StatusCode != 200 {
+		logrus.Warnf("ResetOffset: clear scroll id=%s msg=%s", c.scrollId, rr.String())
+	}
 }
 func (c *client) WriteData(ctx context.Context, docs []*interfaces.ESSource) (int, error) {
 	var (
@@ -152,8 +171,6 @@ func (c *client) WriteData(ctx context.Context, docs []*interfaces.ESSource) (in
 			return 0, err
 		}
 
-		logrus.Debugf("xes.Write: doc content=%s", string(bs))
-
 		if err = indexer.Add(context.Background(), esutil.BulkIndexerItem{
 			Action:     "index",
 			Index:      c.index,
@@ -184,7 +201,7 @@ func (c *client) WriteData(ctx context.Context, docs []*interfaces.ESSource) (in
 	return count, nil
 }
 
-func (c *client) ReadData(ctx context.Context, i int, query map[string]any) ([]*interfaces.ESSource, error) {
+func (c *client) ReadData(ctx context.Context, i int, query map[string]any, source []string) ([]*interfaces.ESSource, error) {
 	var (
 		err    error
 		resp   *esapi.Response
@@ -197,7 +214,11 @@ func (c *client) ReadData(ctx context.Context, i int, query map[string]any) ([]*
 			c.c.Search.WithIndex(c.index),
 			c.c.Search.WithSize(i),
 			c.c.Search.WithFrom(0),
-			c.c.Search.WithScroll(time.Duration(opt.ScrollDurationSeconds) * time.Second),
+			c.c.Search.WithScroll(time.Duration(opt.Timeout*2) * time.Second),
+		}
+
+		if len(source) > 0 {
+			qs = append(qs, c.c.Search.WithSourceIncludes(source...))
 		}
 
 		if query != nil && len(query) > 0 {
@@ -225,7 +246,7 @@ func (c *client) ReadData(ctx context.Context, i int, query map[string]any) ([]*
 
 	if resp, err = c.c.Scroll(
 		c.c.Scroll.WithScrollID(c.scrollId),
-		c.c.Scroll.WithScroll(time.Duration(opt.ScrollDurationSeconds)*time.Second),
+		c.c.Scroll.WithScroll(time.Duration(opt.Timeout*2)*time.Second),
 	); err != nil {
 		return result.Hits.Hits, nil
 	}

internal/xes/xes_test.go

@@ -1,6 +1,9 @@
 package xes
 
 import (
+	"bufio"
+	"fmt"
+	"os"
 	"testing"
 
 	elastic "github.com/elastic/go-elasticsearch/v7"
@@ -37,3 +40,68 @@ func TestGetESMapping(t *testing.T) {
 
 	t.Log("get source:", r.String())
 }
+
+func TestScanWithInterrupt(t *testing.T) {
+	filename := "test_scan.txt"
+	f, err := os.OpenFile(filename, os.O_RDWR|os.O_CREATE, 0644)
+	if err != nil {
+		t.Error(1, err)
+		return
+	}
+	defer func() {
+		os.Remove(filename)
+	}()
+	f.WriteString(`line 01
+line 02
+line 03
+line 04
+line 05
+line 06
+line 07
+line 08
+line 09
+line 10
+line 11
+line 12
+line 13
+line 14
+line 15`)
+	f.Close()
+
+	of, err := os.Open(filename)
+	if err != nil {
+		t.Error(2, err)
+		return
+	}
+
+	scanner := bufio.NewScanner(of)
+
+	count := 0
+	for scanner.Scan() {
+		text := scanner.Text()
+		fmt.Printf("[line: %2d] = %s\n", count, text)
+		count++
+
+		if count > 5 {
+			break
+		}
+	}
+
+	count = 0
+	for scanner.Scan() {
+		text := scanner.Text()
+		fmt.Printf("[line: %2d] = %s\n", count, text)
+		count++
+
+		if count > 5 {
+			break
+		}
+	}
+
+	count = 0
+	for scanner.Scan() {
+		text := scanner.Text()
+		fmt.Printf("[line: %2d] = %s\n", count, text)
+		count++
+	}
+}

internal/xfile/xfile.go

@@ -7,8 +7,9 @@ import (
 	"io"
 	"os"
 
+	"github.com/loveuer/esgo2dump/internal/opt"
+
 	"github.com/loveuer/esgo2dump/internal/interfaces"
-	"github.com/sirupsen/logrus"
 )
 
 type client struct {
@@ -111,7 +112,7 @@ func (c *client) WriteData(ctx context.Context, docs []*interfaces.ESSource) (in
 	return count, nil
 }
 
-func (c *client) ReadData(ctx context.Context, i int, _ map[string]any) ([]*interfaces.ESSource, error) {
+func (c *client) ReadData(ctx context.Context, i int, _ map[string]any, _ []string) ([]*interfaces.ESSource, error) {
 	var (
 		err   error
 		count = 0
@@ -121,8 +122,6 @@ func (c *client) ReadData(ctx context.Context, i int, _ map[string]any) ([]*inte
 	for c.scanner.Scan() {
 		line := c.scanner.Text()
 
-		logrus.Debugf("xfile.Read: line=%s", line)
-
 		item := new(interfaces.ESSource)
 		if err = json.Unmarshal([]byte(line), item); err != nil {
 			return list, err
@@ -152,6 +151,8 @@ func NewClient(file *os.File, ioType interfaces.IO) (interfaces.DumpIO, error) {
 
 	if ioType == interfaces.IOInput {
 		c.scanner = bufio.NewScanner(c.f)
+		buf := make([]byte, opt.BuffSize)
+		c.scanner.Buffer(buf, opt.MaxBuffSize)
 	}
 
 	return c, nil