Expr is an engine that can evaluate expressions.
The purpose of the package is to allow users to use expressions inside configuration for more complex logic. It is a perfect candidate for the foundation of a business rule engine. The idea is to let configure things in a dynamic way without recompile of a program:
# Get the special price if
user.Group in ["good_customers", "collaborator"]
# Promote article to the homepage when
article.CommentCount > 100 and article.Category not in ["misc"]
# Send an alert when
product.Stock < 15Inspired by
- Symfony's The ExpressionLanguage component,
- Rob Pike's talk Lexical Scanning in Go.
go get -u github.com/antonmedv/expr
// Evaluate expression on data.
result, err := expr.Eval("expression", data)
// Or precompile expression to ast first.
node, err := expr.Parse("expression")
// And run later.
result, err := expr.Run(node, data)See The Expression Syntax to learn the syntax of the Expr expressions.
You can pass variables into the expression, which can be of any valid Go type (including structs):
// Maps
data := map[string]interface{}{
"Foo": ...
"Bar": ...
}
// Structs
data := Payload{
Foo: ...
Bar: ...
}
// Pass object
result, err := expr.Eval("Foo == Bar", data)Expr uses reflection for accessing and iterating passed data. For example you can pass nested structures without any modification or preparation:
type Cookie struct {
Key string
Value string
}
type User struct {
UserAgent string
Cookies []Cookie
}
type Request struct {
User user
}
req := Request{User{
Cookies: []cookie{{"origin", "www"}},
UserAgent: "Firefox",
}}
ok, err := expr.Eval(`User.UserAgent matches "Firefox" and User.Cookies[0].Value == "www"`, req)You can also pass functions into the expression:
data := map[string]interface{}{
"Request": req,
"Values": func(xs []Cookie) []string {
vs := make([]string, 0)
for _, x := range xs {
vs = append(vs, x.Value)
}
return vs
},
}
ok, err := expr.Eval(`"www" in Values(Request.User.Cookies)`, data)If you planning to execute some expression lots times, it's good to compile it first and only one time:
// Precompile
node, err := expr.Parse(expression)
// Run
ok, err := expr.Run(node, data)It is possible to check used variables and functions during parsing of the expression.
expression := `Request.User.UserAgent matches "Firefox" && "www" in Values(Request.User.Cookies)`
node, err := expr.Parse(expression, expr.Names("Request"), expr.Funcs("Values"))Only Request and Values will be available inside expression, otherwise parse error.
Compiled ast can be compiled back to string expression using String():
node, err := expr.Parse(expression)
code := fmt.Sprintf("%v", node)Inside Expr engine there is no distinguish between int, uint and float types (as in JavaScript).
All numbers inside Expr engine represented as float64.
You should remember about it if you use any of binary operators (+, -, /, *, etc).
Otherwise type remain unchanged.
data := map[string]int{
"Foo": 1,
"Bar": 2,
}
out, err := expr.Eval(`Foo`, data) // int
out, err := expr.Eval(`Foo + Bar`, data) // float64MIT
