use string for value from pflag (#190)

and leave the conversion to mapstructure

provider/pflag/pflag.go

@@ -86,96 +86,45 @@ func (f PFlag) Load() (map[string]any, error) { //nolint:cyclop
 				return
 			}
 
-			val, _ := f.flagVal(set, flag) // Ignore error as it uses whatever returned.
 			// Skip zero default value to avoid overriding values set by other loader.
-			if !flag.Changed && (exists(keys) || reflect.ValueOf(val).IsZero()) {
+			if !flag.Changed && (exists(keys) || zeroDefaultValue(flag)) {
 				return
 			}
 
-			maps.Insert(values, keys, val)
+			maps.Insert(values, keys, flag.Value.String())
 		},
 	)
 
 	return values, nil
 }
 
-//nolint:cyclop,funlen,gocyclo,wrapcheck
-func (f PFlag) flagVal(set *pflag.FlagSet, flag *pflag.Flag) (any, error) {
+func zeroDefaultValue(flag *pflag.Flag) bool { //nolint:cyclop
 	switch flag.Value.Type() {
-	case "int":
-		return set.GetInt(flag.Name)
-	case "uint":
-		return set.GetUint(flag.Name)
-	case "int8":
-		return set.GetInt8(flag.Name)
-	case "uint8":
-		return set.GetUint8(flag.Name)
-	case "int16":
-		return set.GetInt16(flag.Name)
-	case "uint16":
-		return set.GetUint16(flag.Name)
-	case "int32":
-		return set.GetInt32(flag.Name)
-	case "uint32":
-		return set.GetUint32(flag.Name)
-	case "int64":
-		return set.GetInt64(flag.Name)
-	case "uint64":
-		return set.GetUint64(flag.Name)
-	case "float":
-		return set.GetFloat64(flag.Name)
-	case "float32":
-		return set.GetFloat32(flag.Name)
-	case "float64":
-		return set.GetFloat64(flag.Name)
 	case "bool":
-		return set.GetBool(flag.Name)
+		return flag.DefValue == "false"
 	case "duration":
-		return set.GetDuration(flag.Name)
-	case "ip":
-		return set.GetIP(flag.Name)
-	case "ipMask":
-		return set.GetIPv4Mask(flag.Name)
-	case "ipNet":
-		return set.GetIPNet(flag.Name)
-	case "count":
-		return set.GetCount(flag.Name)
-	case "bytesHex":
-		return set.GetBytesHex(flag.Name)
-	case "bytesBase64":
-		return set.GetBytesBase64(flag.Name)
+		// Beginning in Go 1.7, duration zero values are "0s"
+		return flag.DefValue == "0" || flag.DefValue == "0s"
+	case "count", "int", "int8", "int16", "int32", "int64",
+		"uint", "uint8", "uint16", "uint32", "uint64",
+		"float32", "float64":
+		return flag.DefValue == "0"
 	case "string":
-		return set.GetString(flag.Name)
-	case "stringSlice":
-		return set.GetStringSlice(flag.Name)
-	case "intSlice":
-		return set.GetIntSlice(flag.Name)
-	case "uintSlice":
-		return set.GetUintSlice(flag.Name)
-	case "int32Slice":
-		return set.GetInt32Slice(flag.Name)
-	case "int64Slice":
-		return set.GetInt64Slice(flag.Name)
-	case "float32Slice":
-		return set.GetFloat32Slice(flag.Name)
-	case "float64Slice":
-		return set.GetFloat64Slice(flag.Name)
-	case "boolSlice":
-		return set.GetBoolSlice(flag.Name)
-	case "durationSlice":
-		return set.GetDurationSlice(flag.Name)
-	case "ipSlice":
-		return set.GetIPSlice(flag.Name)
-	case "stringArray":
-		return set.GetStringArray(flag.Name)
-	case "stringToString":
-		return set.GetStringToString(flag.Name)
-	case "stringToInt":
-		return set.GetStringToInt(flag.Name)
-	case "stringToInt64":
-		return set.GetStringToInt64(flag.Name)
+		return flag.DefValue == ""
+	case "ip", "ipMask", "ipNet":
+		return flag.DefValue == "<nil>"
+	case
+		"boolSlice", "durationSlice", "ipSlice",
+		"bytesHex", "bytesBase64", "stringArray", "stringSlice", "stringToString", "stringToInt", "stringToInt64",
+		"intSlice", "int32Slice", "int64Slice", "uintSlice", "float32Slice", "float64Slice":
+		return flag.DefValue == "[]"
 	default:
-		return flag.Value.String(), nil
+		switch flag.DefValue {
+		case "false", "<nil>", "", "0":
+			return true
+		default:
+			return false
+		}
 	}
 }