Created
July 9, 2018 08:01
-
-
Save renskiy/d2725afa620be9cd4bdf0f15e9339655 to your computer and use it in GitHub Desktop.
Syntax analyser of strings like "(foo (bar baz :arg 42))"
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| package include_parser | |
| import ( | |
| "regexp" | |
| "strings" | |
| "errors" | |
| "fmt" | |
| ) | |
| var whitespaces = regexp.MustCompile("\\s+") | |
| const maxNestingLevel = 3 | |
| type Node struct { | |
| name string | |
| nodes map[string]Node | |
| args map[string]string | |
| } | |
| func New(value string) (*Node, error) { | |
| if tokens, err := tokenize(value); err != nil { | |
| return nil, err | |
| } else { | |
| if nodes, _, err := parse(tokens); err != nil { | |
| return nil, err | |
| } else { | |
| return &Node{"", nodes, nil}, nil | |
| } | |
| } | |
| } | |
| func tokenize(params string) ([]string, error) { | |
| params = whitespaces.ReplaceAllString(params, " ") | |
| params = strings.Trim(params, "'\" ") | |
| var ( | |
| tokens []string | |
| start, nestingLevel int | |
| ) | |
| capture := func(index int) error { | |
| if start < index { | |
| var token = params[start:index] | |
| if strings.HasPrefix(token, ":") && len(token) == 1 { | |
| return errors.New(fmt.Sprintf("parameter without name (position: %v)", index - 1)) | |
| } | |
| tokens = append(tokens, token) | |
| } | |
| start = index + 1 | |
| return nil | |
| } | |
| for index, char := range params { | |
| switch char { | |
| case '(': | |
| start = index + 1 | |
| tokens = append(tokens, string(char)) | |
| nestingLevel++ | |
| case ')': | |
| if err := capture(index); err != nil { | |
| return nil, err | |
| } | |
| tokens = append(tokens, string(char)) | |
| nestingLevel-- | |
| case ' ': | |
| if err := capture(index); err != nil { | |
| return nil, err | |
| } | |
| } | |
| if nestingLevel < 0 { | |
| return nil, errors.New(fmt.Sprintf("unexpected closing parentheses (position: %v)", index)) | |
| } | |
| if nestingLevel > maxNestingLevel { | |
| return nil, errors.New(fmt.Sprintf("parentheses nesting level exceeded (position: %v)", index)) | |
| } | |
| } | |
| if nestingLevel > 0 { | |
| return nil, errors.New("unclosed parentheses") | |
| } | |
| if start < len(params) { | |
| tokens = append(tokens, params[start:]) | |
| } | |
| return tokens, nil | |
| } | |
| func parse(tokens []string, startRef ...*int) (map[string]Node, map[string]string, error) { | |
| var ( | |
| nodes = make(map[string]Node) | |
| args = make(map[string]string) | |
| start *int | |
| ) | |
| if len(startRef) > 0 { | |
| start = startRef[0] | |
| } else { | |
| startInitial := 0 | |
| start = &startInitial | |
| } | |
| for *start < len(tokens) { | |
| token := tokens[*start] | |
| if token == "(" { | |
| *start++ | |
| token = tokens[*start] | |
| if strings.HasPrefix(token, ":") { | |
| return nil, nil, errors.New(fmt.Sprintf("parameter %v for object without name", token)) | |
| } | |
| if token != ")" { | |
| *start++ | |
| if childNodes, nodeArgs, err := parse(tokens, start); err != nil { | |
| return nil, nil, errors.New(fmt.Sprintf("parameter %v for object without name", token)) | |
| } else { | |
| nodes[token] = Node{token, childNodes, nodeArgs} | |
| } | |
| } | |
| } else if token == ")" { | |
| break | |
| } else if strings.HasPrefix(token, ":") { | |
| *start++ | |
| if *start >= len(tokens) { | |
| return nil, nil, errors.New(fmt.Sprintf("not provided value for %v", token)) | |
| } | |
| if argValue := tokens[*start]; argValue == ")" { | |
| return nil, nil, errors.New(fmt.Sprintf("not provided value for %v", token)) | |
| } else { | |
| args[token[1:]] = argValue | |
| } | |
| } else { | |
| nodes[token] = Node{token, nil, nil} | |
| } | |
| *start++ | |
| } | |
| return nodes, args, nil | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment