remove uuid (#253)

* fix: negative hex
improve performance of checkVidFormat

* fix: remove uuid

README.md

@@ -235,7 +235,7 @@ schema:
 
 * `index`: **Optional**. The column number in the CSV file. Started with 0. The default value is 0.
 * `concatItems`: **Optional**. The concat item can be `string`, `int` or mixed. `string` represents a constant, and `int` represents an index column. Then connect all items.If set, the above `index` will have no effect.
-* `function`: **Optional**. Functions to generate the VIDs. Currently, we only support function `hash` and `uuid`.
+* `function`: **Optional**. Functions to generate the VIDs. Currently, we only support function `hash`.
 * `type`: **Optional**. The type for VIDs. The default value is `string`.
 * `prefix`: **Optional**. Add prefix to the original vid. When `function` is specified also, `prefix` is applied to the original vid before `function`.
 
@@ -271,7 +271,7 @@ schema:
       function: hash
     dstVID:
       index: 1
-      function: uuid
+      function: hash
     rank:
       index: 2
     props:
@@ -348,7 +348,7 @@ Take vertex course as example:
 ```csv
 :LABEL,:VID,course.name,building.name:string,:IGNORE,course.credits:int
 +,"hash(""Math"")",Math,No5,1,3
-+,"uuid(""English"")",English,"No11 B\",2,6
++,"hash(""English"")",English,"No11 B\",2,6
 ```
 
 ##### LABEL (optional)
@@ -367,10 +367,10 @@ Indicates the column is the insertion (+) or deletion (-) operation.
 :VID
 123,
 "hash(""Math"")",
-"uuid(""English"")"
+"hash(""English"")"
 ```
 
-In the `:VID` column, in addition to the common integer values (such as 123), you can also use the two built-in functions `hash` and `uuid` to automatically generate the VID for the vertices (for example, hash("Math")).
+In the `:VID` column, in addition to the common integer values (such as 123), you can also use the two built-in function `hash` to automatically generate the VID for the vertices (for example, hash("Math")).
 
 > **NOTE**: The double quotes (") are escaped in the CSV file. For example, `hash("Math")` must be written as `"hash(""Math"")"`.
 

README_zh-CN.md

@@ -208,7 +208,7 @@ schema:
 
 - `index`：**可选**。在 CSV 文件中的列标，从 0 开始计数。默认值 0。
 - `concatItems`: **可选**. 连接项可以是`string`、`int`或者混合。`string`代表常量，`int`表示索引列。然后连接所有的项。如果设置了，上面的`index`将不生效。
-- `function`：**可选**。用来生成 VID 时的函数，有 `hash` 和 `uuid` 两种函数可选。
+- `function`：**可选**。用来生成 VID 时的函数，支持 `hash` 函数。
 - `prefix`: **可选**。给 原始vid 添加的前缀，当同时指定了 `function` 时, 生成 VID 的方法是先添加 `prefix` 前缀, 再用 `function`生成 VID。
 
 ##### `schema.vertex.tags`
@@ -243,7 +243,7 @@ schema:
       function: hash
     dstVID:
       index: 1
-      function: uuid
+      function: hash
     rank:
       index: 2
     props:
@@ -320,7 +320,7 @@ example 中 course 点的示例：
 ```csv
 :LABEL,:VID,course.name,building.name:string,:IGNORE,course.credits:int
 +,"hash(""Math"")",Math,No5,1,3
-+,"uuid(""English"")",English,"No11 B\",2,6
++,"hash(""English"")",English,"No11 B\",2,6
 ```
 
 ##### LABEL (可选）
@@ -339,10 +339,10 @@ example 中 course 点的示例：
 :VID
 123,
 "hash(""Math"")",
-"uuid(""English"")"
+"hash(""English"")"
 ```
 
-在 `:VID` 这列除了常见的整数值（例如 123），还可以使用 `hash` 和 `uuid` 两个内置函数来自动计算生成点的 VID（例如 hash("Math")）。
+在 `:VID` 这列除了常见的整数值（例如 123），还可以使用 `hash` 内置函数来自动计算生成点的 VID（例如 hash("Math")）。
 
 > 需要注意的是在 CSV 文件中对双引号(")的转义处理。如 `hash("Math")` 要写成 `"hash(""Math"")"`。
 

examples/v1/choose-hex.csv

@@ -3,3 +3,4 @@
 0x201,00102,3
 0X202,02102,3
 0x2af,0,3
+-0X202,-02102,6

examples/v1/example.yaml

@@ -297,7 +297,7 @@ files:
           type: int
         dstVID:
           index: 1
-          # function: uuid
+          # function: hash
           type: int
         rank:
           index: 2

examples/v2/example.yaml

@@ -314,7 +314,6 @@ files:
       vertex:
         vid:
           index: 1
-          # function: uuid
         tags:
           - name: student
             props:
@@ -344,7 +343,6 @@ files:
           # function: hash
         dstVID:
           index: 1
-          # function: uuid
         rank:
           index: 2
         props:
@@ -383,7 +381,6 @@ files:
           # function: hash
         dstVID:
           index: 1
-          # function: uuid
         rank:
           index: 2
         props:

pkg/config/config.go

@@ -7,25 +7,21 @@ import (
 	"net/url"
 	"os"
 	"path/filepath"
-	"regexp"
 	"strings"
 	"time"
 
 	"github.com/vesoft-inc/nebula-importer/pkg/base"
 	ierrors "github.com/vesoft-inc/nebula-importer/pkg/errors"
 	"github.com/vesoft-inc/nebula-importer/pkg/logger"
 	"github.com/vesoft-inc/nebula-importer/pkg/picker"
+	"github.com/vesoft-inc/nebula-importer/pkg/utils"
 	"gopkg.in/yaml.v2"
 )
 
 const (
 	dbNULL = "NULL"
 )
 
-var (
-	reTimestampInteger = regexp.MustCompile(`^(0[xX][0-9a-fA-F]+|0[0-7]+|\d+)$`)
-)
-
 type NebulaClientConnection struct {
 	User     *string `json:"user" yaml:"user"`
 	Password *string `json:"password" yaml:"password"`
@@ -546,10 +542,9 @@ func (v *VID) FormatValue(record base.Record) (string, error) {
 func (v *VID) checkFunction(prefix string) error {
 	if v.Function != nil {
 		switch strings.ToLower(*v.Function) {
-		// FIXME: uuid is not supported in nebula-graph-v2, and hash returns int which is not the valid vid type.
-		case "", "hash", "uuid":
+		case "", "hash":
 		default:
-			return fmt.Errorf("Invalid %s.function: %s, only following values are supported: \"\", hash, uuid", prefix, *v.Function)
+			return fmt.Errorf("Invalid %s.function: %s, only following values are supported: \"\", hash", prefix, *v.Function)
 		}
 	}
 	return nil
@@ -628,11 +623,16 @@ func (r *Rank) validateAndReset(prefix string, defaultVal int) error {
 	return nil
 }
 
-var re = regexp.MustCompile(`^(0[xX][0-9a-fA-F]+|0[0-7]+|[+-]?\d+|hash\(".+"\)|uuid\(".+"\))$`)
-
 func checkVidFormat(vid string, isInt bool) error {
-	if isInt && !re.MatchString(vid) {
-		return fmt.Errorf("Invalid vid format: %s", vid)
+	if isInt {
+		if utils.IsInteger(vid) {
+			return nil
+		}
+		vidLen := len(vid)
+		if vidLen > 8 /* hash("") */ && strings.HasSuffix(vid, "\")") && strings.HasPrefix(vid, "hash(\"") {
+			return nil
+		}
+		return fmt.Errorf("Invalid vid format: " + vid)
 	}
 	return nil
 }

pkg/config/config_test.go

@@ -1181,3 +1181,12 @@ func TestParseFunction(t *testing.T) {
 		})
 	}
 }
+
+func Benchmark_checkVidFormat(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		_ = checkVidFormat("-0xfedcba9876543210", true)
+		_ = checkVidFormat("-076543210", true)
+		_ = checkVidFormat("-9876543210", true)
+		_ = checkVidFormat("hash(\"abcdefg\")", true)
+	}
+}

pkg/picker/converter-type.go

@@ -3,6 +3,8 @@ package picker
 import (
 	"fmt"
 	"strings"
+
+	"github.com/vesoft-inc/nebula-importer/pkg/utils"
 )
 
 var (
@@ -113,7 +115,7 @@ func (tc TypeStringConverter) Convert(v *Value) (*Value, error) {
 }
 
 func (tc TypeTimestampConverter) Convert(v *Value) (*Value, error) {
-	if isUnsignedInteger(v.Val) {
+	if utils.IsUnsignedInteger(v.Val) {
 		return tc.fc.Convert(v)
 	}
 	return tc.fsc.Convert(v)

pkg/utils/string.go

@@ -0,0 +1,46 @@
+package utils
+
+func IsInteger(s string) bool {
+	if s == "" {
+		return false
+	}
+	if s[0] == '+' || s[0] == '-' {
+		s = s[1:]
+	}
+	return IsUnsignedInteger(s)
+}
+
+func IsUnsignedInteger(s string) bool {
+	switch len(s) {
+	case 0:
+		return false
+	case 1:
+		return IsDigit(s[0])
+	case 2:
+		return IsDigit(s[0]) && IsDigit(s[1])
+	}
+	return isUnsignedIntegerSlow(s)
+}
+
+func isUnsignedIntegerSlow(s string) bool {
+	f := IsDigit
+	if len(s) > 2 && s[0] == '0' && (s[1] == 'x' || s[1] == 'X') {
+		s = s[2:]
+		f = IsHexDigit
+	}
+
+	for _, b := range []byte(s) {
+		if !f(b) {
+			return false
+		}
+	}
+	return true
+}
+
+func IsDigit(b byte) bool {
+	return '0' <= b && b <= '9'
+}
+
+func IsHexDigit(b byte) bool {
+	return IsDigit(b) || ('a' <= b && b <= 'f') || ('A' <= b && b <= 'F')
+}