How to debug geth with goland? - ethereum

I would like to be able to debug through Geth's Go code - using Goland.
It looks to me as if the entry point for Geth is main.go, however, when I try to run this I get the following errors:
GOROOT=/usr/local/Cellar/go/1.11/libexec #gosetup
GOPATH=/Users/codet/go #gosetup
/usr/local/Cellar/go/1.11/libexec/bin/go build -o /private/var/folders/6t/tdn_s_2x2rx1scn774cwvlph0000gn/T/___geth_go -gcflags "all=-N -l" /Users/codet/go/src/github.com/ethereum/go-ethereum/cmd/geth/main.go #gosetup
# command-line-arguments
cmd/geth/main.go:136:3: undefined: configFileFlag
cmd/geth/main.go:177:3: undefined: initCommand
cmd/geth/main.go:178:3: undefined: importCommand
cmd/geth/main.go:179:3: undefined: exportCommand
cmd/geth/main.go:180:3: undefined: importPreimagesCommand
cmd/geth/main.go:181:3: undefined: exportPreimagesCommand
cmd/geth/main.go:182:3: undefined: copydbCommand
cmd/geth/main.go:183:3: undefined: removedbCommand
cmd/geth/main.go:184:3: undefined: dumpCommand
cmd/geth/main.go:186:3: undefined: monitorCommand
cmd/geth/main.go:186:3: too many errors
Compilation finished with exit code 2
I add debug point to function main, but it didn't work.
func main() {
if err := app.Run(os.Args); err != nil {
fmt.Fprintln(os.Stderr, err)
os.Exit(1)
}
}

Dude, try this command go build -x path-to-go-ethereum/cmd/geth/*go, you will see what happens with flag -x. Should compile the package instead.
In goland, right click the geth folder, just choose run->go bulid, you will get what you expect.

Related

Error on importing go-sql-driver/sql from github repository

As the title says, I have an error when importing go-mysql-driver package. I have installed the go-my-sql driver in my machine but the error still persists. I use XAMPP for local hosting and here’s the block of program.
package model
import (
"database/sql"
"fmt"
_ "github.com/go-sql-driver/mysql"
)
type Table interface {
Name() string
Field() ([]string, []interface{})
}
func Connect(username string, password string, host string, database string) (*sql.DB, error) {
conn := fmt.Sprintf("%s:%s#tcp(%s:3306)/%s", username, password, host, database)
db, err := sql.Open("mysql", conn)
return db, err
}
func CreateDB(db *sql.DB, name string) error {
query := fmt.Sprintf("CREATE DATABASE %v", name)
_, err := db.Exec(query)
return err
}
func CreateTable(db *sql.DB, query string) error {
_, err := db.Exec(query)
return err
}
func DropDB(db *sql.DB, name string) error {
query := fmt.Sprintf("DROP DATABASE %v", name)
_, err := db.Exec(query)
return err
}
could not import github.com/go-sql-driver/mysql (no required modules provides package "github.com/go-sql-driver/mysql")
screenshot of what's happening
It seems that you read the tutorial for an older go version.
Go 1.17 requires dependencies must be explicitly in go.mod.
Maybe you could try go module first (https://go.dev/blog/using-go-modules)
Your IDE is not showing you the whole picture. By running go run main.go (or whatever main file you have) on the command line, you can see the same error as you're seeing on your IDE with some extra:
$ go run main.go
main.go:7:5: no required module provides package github.com/go-sql-driver/mysql; to add it:
go get github.com/go-sql-driver/mysql
By issuing the suggested command go get github.com/go-sql-driver/mysql, you'll get the dependency added to your go.mod file and the contents of the package will be downloaded to your machine.
Your next execution will work:
$ go run main.go
Hello world
I've made some small modifications to your code to work, I'll add them here for completeness:
I've used the same source, but changed the package name to main.
I've added a main function to the bottom of the file:
func main() {
fmt.Println("Hello world")
_, err := Connect("username", "password", "localhost", "db")
if err != nil {
panic(err)
}
}
I've saved to a file named main.go
I've initialized the go.mod file by running go mod init test and go mod tidy, then I took the steps described on the beginning of the answer.

How to build libvirt-go for windows on mac

I installed the libs and can run/debug the libvirt related codes in idea IDE, but after running below build command on MacBook
CGO_ENABLED=0 GOOS=windows GOARCH=amd64 go build -o zagent.exe cmd/test/main.go
I got
cmd/test/main.go:11:22: undefined: libvirt.NewConnect
Thank you for your attention, below is the codes:
package main
import (
_logUtils "github.com/easysoft/zagent/internal/pkg/lib/log"
"github.com/libvirt/libvirt-go"
)
func main() {
connStr := "***"
LibvirtConn, err := libvirt.NewConnect(connStr)
if err != nil {
_logUtils.Errorf(err.Error())
return
}
active, err := LibvirtConn.IsAlive()
if err != nil {
_logUtils.Errorf(err.Error())
return
}
if !active {
_logUtils.Errorf("not active")
}
}
The libvirt-go package is a CGo based API to the underlying libvirt.so library. You cannot set CGO_ENABLED=0 and expect it to still work. AFAI, Go toolchain doesn't allow cross-compiling with CGo either.

How to disable default error logger in Go-Gorm

I am using GORM with MySQL, I have encountered and handled the error Error 1062: Duplicate entry. The problem is that it's still printed to the console.
Code in gym/models/auth.go:49:
func AddAuth(username, password string) error {
passwordHash, err := auth.HashPassword(password, argon2Conf)
if err != nil {
return err
}
userAuth := Auth{
Username: username,
Password: passwordHash,
}
return db.Create(&userAuth).Error
}
I am handling the error in the handler function:
func SignUpHandler(c *gin.Context) {
var form user
if err := c.ShouldBind(&form); err != nil {
c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
return
}
if err := models.AddAuth(form.Username, form.Password); err == nil {
c.JSON(http.StatusOK, gin.H{"status": "you are signed in"})
} else {
// I think I have handled the sql error here
c.JSON(http.StatusBadRequest, gin.H{"error": "sign in failed"})
}
}
When I send a POST request, the error is correctly handled and I get the correct response with {"error": "sign in failed"}. But the console still prints this error message:
(/...../gym/models/auth.go:49)
[2019-04-28 23:37:06] Error 1062: Duplicate entry '123123' for key 'username'
[GIN] 2019/04/28 - 23:37:06 | 400 | 136.690908ms | ::1 | POST /signup
I am confused since I handled the error but it still gets printed. How to prevent this error from getting printed to the error log? Or am I handle the error correct?
UPDATE: for gorm v2:
Use the Logger in gorm.Config:
db, err := gorm.Open(sqlite.Open("test.db"), &gorm.Config{
Logger: logger.Default.LogMode(logger.Silent),
})
For gorm v1:
Use db.LogMode to silence the default error logger.
LogMode set log mode, true for detailed logs, false for no log, default, will only print error logs.
db.LogMode(false) should do the job!
I don't have enough reputation to comment but just to add to the answer by ifnotak, you can log sql conditionally by controlling it through an environment variable. This can be handy during debugging.
gormConfig := &gorm.Config{}
if !logSql {
// I use an env variable LOG_SQL to set logSql to either true or false.
gormConfig.Logger = logger.Default.LogMode(logger.Silent)
}
db, err := gorm.Open(dialector, gormConfig)

How to execute Mysql Script in golang using exec.Command

Hi i am trying to execute a script to fill data into a database using Golang
func executeTestScript(){
cmd := exec.Command("/usr/local/mysql/bin/mysql", "-h127.0.0.1", "-P3333", "-uusr", "-pPassxxx", "-Ddtb_test", "< /Users/XXX/Documents/test/scripts/olds/SCRIPT_XXX.sql")
var out, stderr bytes.Buffer
cmd.Stdout = &out
cmd.Stderr = &stderr
err := cmd.Run()
if err != nil {
fmt.Println(fmt.Sprintf("Error executing query. Command Output: %+v\n: %+v, %v", out.String(), stderr.String(), err))
log.Fatalf("Error executing query. Command Output: %+v\n: %+v, %v", out.String(), stderr.String(), err)
}
}
The problem is that i am getting the error:
ERROR 1049 (42000): Unknown database '< /Users/XXX/Documents/test/scripts/olds/SCRIPT_XXX.sql'
i think the problem is the last param (the sql script path) that the exec thinks is the dbname
The following command in the terminal is working:
/usr/local/mysql/bin/mysql --host=127.0.0.1 --port=3333 --user=usr --password=Passxxx --database=dtb_test < /Users/XXX/Documents/roseula/scripts/olds/SCRIPT_XXX.sql
but i try to replicate in Go to automatize the execution of the script.
The script have drop tables, create tables, inserts, and PK with FK relationships its a very complete one so i cant execute line by line, because of that i decided to execute de mysql program to insert the data in the database.
Any suggestions?
+1 to answer from #MatteoRagni for showing how to do stdin redirection in Golang.
But here's a simple alternative that I use:
cmd := exec.Command("/usr/local/mysql/bin/mysql", "-h127.0.0.1", "-P3333",
"-uusr", "-pPassxxx", "-Ddtb_test",
"-e", "source /Users/XXX/Documents/test/scripts/olds/SCRIPT_XXX.sql")
You don't have to make the mysql client read the script using stdin redirection. Instead, you can make the mysql client execute a specific command, which is source <scriptname>.
P.S.: I also would not put the host, port, user, and password in your code. That means you have to recompile your program any time you change those connection parameters. Also it's not secure to use passwords in plaintext on the command-line. Instead, I'd put all the connection parameters into a defaults file and use mysql --defaults-file=FILENAME.
This is a little example that runs something like:
cat < test.txt
that is what I think you are missing in your code:
package main
import (
"fmt"
"os/exec"
"os"
)
func main() {
cmd := exec.Command("cat")
file, _ := os.Open("test.txt")
cmd.Stdin = file
out, _ := cmd.Output()
fmt.Printf("%s\n", out)
}
That prints in the console the content of test.txt, as read by cat. You will need to adapt it to your problem.
Something like:
func executeTestScript(){
cmd := exec.Command("/usr/local/mysql/bin/mysql", "-h127.0.0.1", "-P3333", "-uusr", "-pPassxxx", "-Ddtb_test")
dump, dump_err = os.Open("/Users/XXX/Documents/test/scripts/olds/SCRIPT_XXX.sql")
if dump_err != nil {
/* Handle the error if file not opened */
}
cmd.Stdin = dump
var out, stderr bytes.Buffer
cmd.Stdout = &out
cmd.Stderr = &stderr
err := cmd.Run()
if err != nil {
fmt.Println(fmt.Sprintf("Error executing query. Command Output: %+v\n: %+v, %v", out.String(), stderr.String(), err))
log.Fatalf("Error executing query. Command Output: %+v\n: %+v, %v", out.String(), stderr.String(), err)
}
}
if I'm not wrong...

Why doesn't Perl 6's try handle a non-zero exit in shell()?

This try catches the exception:
try die X::AdHoc;
say "Got to the end";
The output shows that the program continues:
Got to the end
If I attempt it with shell and a command that doesn't exit with 0, the try doesn't catch it:
try shell('/usr/bin/false');
say "Got to the end";
The output doesn't look like an exception:
The spawned command '/usr/bin/false' exited unsuccessfully (exit code: 1)
in block <unit> at ... line ...
What's going on that this makes it through the try?
The answer is really provided by Jonathan Worthington:
https://irclog.perlgeek.de/perl6-dev/2017-04-04#i_14372945
In short, shell() returns a Proc object. The moment that object is sunk, it will throw the exception that it has internally if running the program failed.
$ 6 'dd shell("/usr/bin/false")'
Proc.new(in => IO::Pipe, out => IO::Pipe, err => IO::Pipe, exitcode => 1, signal => 0, command => ["/usr/bin/false"])
So, what you need to do is catch the Proc object in a variable, to prevent it from being sunk:
$ 6 'my $result = shell("/usr/bin/false"); say "Got to the end"'
Got to the end
And then you can use $result.exitcode to see whether it was successful or not.