Retire python2.7 scripts and try to migrate to golang.

• "aaabbb".startswith("aa")

strings.HasPrefix("aaabbb", "aa")

• "aaabbb".endswith("bb")

strings.HasSuffix("aaabbb", "bb")

• "a"*16

strings.Repeat("a", 16)

• "a" in "aa"

strings.Contains("aa", "a")

• "abcb".replace("b", "c")

strings.ReplaceAll("abcb", "b", "c")

• "1|2|3".split("|", 1)

Notice that 1 means split the string by at most 1 |, however split function in golang use the number of parts as a parameter.

strings.SplitN("1|2|3", "|", 2)

Python's float function tranforms many things to float. According to this, "infinity" can also be tranformed to float.

• Stackoverflow: Converting unknown interface to float64 in Golang

No need.

• Ref: Stackoverflow: Equivalent of python's encode('utf8') in golang

• threading.Lock: Golang's sync.Mutex only provides Lock and Unlock operation.
  • No TryLock?
    • This issue has been discussed in 2013, and golang seems not to implement this function.
    • github.com/LK4D4/trylock
    • Or implement by channel:

// A Lock
ch := make(chan bool, 1)

// Lock
ch <- true

// Trylock?
select {
case ch <- true:
    // Trylock successfully
default:
    // Fail to lock 
}

// Islock?
len(ch) == 1

// Unlock
<-ch

• go 1.18: Mutex.TryLock()

• string.Template("${a}${b}").safe_substitute({"a":1,"b":2})

This function doesn't has a proper replace go package. We need to implement PEP0292 ourself. strings.Replacer will apply to most of the cases.

func SafeSubstitute(temp string, vars map[string]string) string {
	replacerList := []string{}
	for varName, varValue := range vars {
		replacerList = append(replacerList, "${"+varName+"}")
		replacerList = append(replacerList, varValue)
		replacerList = append(replacerList, "$"+varName)
		replacerList = append(replacerList, varValue)
	}
	replacerList = append(replacerList, "$$")
	replacerList = append(replacerList, "$")
	rep := strings.NewReplacer(replacerList...)
	return rep.Replace(temp)
}

• int(time.time()): Unix second

time.Now().Unix()

• urllib.quote(s): Quote path section of the URL

url.PathEscape(s)

• urllib.quote_plus(s): Quote query section of the URL

url.QueryEscape(s)

• urlparse.urlparse

  • u = urlparse('http://www.aaa.bb:80/c/d.html?a=1')

    • u.schema: http
    • u.netloc: www.aaa.bb:80
    • u.hostname: www.aaa.bb
    • u.query: a=1

  • u, _ := url.Parse("http://www.aaa.bb:80/c/d.html?a=1")

    • u.Schema: http
    • u.Host: www.aaa.bb:80
      • host, port, _ := net.SplitHostPort(u.Host)
        • host: www.aaa.bb
        • post: 80
        • Notice that net.SplitHostPort("www.kimo.com") return err and empty host and empty port
    • u.RawQuery: a=1

• os.path.join(a, b, c)

path.Join(a, b, c)

• base64.urlsafe_b64encode(s)

Example: base64.urlsafe_b64encode("a") = "YQ=="

base64.RawURLEncoding.EncodeToString([]byte("a"))

EncodeToString("a") return "YQ" which has no "=" padding.

• base64.urlsafe_b64decode(s)

Example: base64.urlsafe_b64decode("YQ==") = "a" Throw exception if no "=" padding

base64.RawURLEncoding.DecodeToString([]byte("YQ"))

DecodeToString("YQ") return "a". DecodeToString("YQ==") return error: "=" is an illegal character.

Support ipv4, ipv6

• github.com/yl2chen/cidranger
• Check/Get operation need about 100ns~300ns (DB records $\approx$ 1200)