prometheus/util/strutil/quote.go

// Copyright 2015 The Prometheus Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

package strutil

import (
	"errors"
	"unicode/utf8"
)

// ErrSyntax indicates that a value does not have the right syntax for the target type.
var ErrSyntax = errors.New("invalid syntax")

// Unquote interprets s as a single-quoted, double-quoted, or backquoted
// Prometheus query language string literal, returning the string value that s
// quotes.
//
// NOTE: This function as well as the necessary helper functions below
// (unquoteChar, contains, unhex) and associated tests have been adapted from
// the corresponding functions in the "strconv" package of the Go standard
// library to work for Prometheus-style strings. Go's special-casing for single
// quotes was removed and single quoted strings are now treated the same as
// double quoted ones.
func Unquote(s string) (t string, err error) {
	n := len(s)
	if n < 2 {
		return "", ErrSyntax
	}
	quote := s[0]
	if quote != s[n-1] {
		return "", ErrSyntax
	}
	s = s[1 : n-1]

	if quote == '`' {
		if contains(s, '`') {
			return "", ErrSyntax
		}
		return s, nil
	}
	if quote != '"' && quote != '\'' {
		return "", ErrSyntax
	}
	if contains(s, '\n') {
		return "", ErrSyntax
	}

	// Is it trivial?  Avoid allocation.
	if !contains(s, '\\') && !contains(s, quote) {
		return s, nil
	}

	var runeTmp [utf8.UTFMax]byte
	buf := make([]byte, 0, 3*len(s)/2) // Try to avoid more allocations.
	for len(s) > 0 {
		c, multibyte, ss, err := unquoteChar(s, quote)
		if err != nil {
			return "", err
		}
		s = ss
		if c < utf8.RuneSelf || !multibyte {
			buf = append(buf, byte(c))
		} else {
			n := utf8.EncodeRune(runeTmp[:], c)
			buf = append(buf, runeTmp[:n]...)
		}
	}
	return string(buf), nil
}

// unquoteChar decodes the first character or byte in the escaped string
// or character literal represented by the string s.
// It returns four values:
//
//	1) value, the decoded Unicode code point or byte value;
//	2) multibyte, a boolean indicating whether the decoded character requires a multibyte UTF-8 representation;
//	3) tail, the remainder of the string after the character; and
//	4) an error that will be nil if the character is syntactically valid.
//
// The second argument, quote, specifies the type of literal being parsed
// and therefore which escaped quote character is permitted.
// If set to a single quote, it permits the sequence \' and disallows unescaped '.
// If set to a double quote, it permits \" and disallows unescaped ".
// If set to zero, it does not permit either escape and allows both quote characters to appear unescaped.
func unquoteChar(s string, quote byte) (value rune, multibyte bool, tail string, err error) {
	// easy cases
	switch c := s[0]; {
	case c == quote && (quote == '\'' || quote == '"'):
		err = ErrSyntax
		return
	case c >= utf8.RuneSelf:
		r, size := utf8.DecodeRuneInString(s)
		return r, true, s[size:], nil
	case c != '\\':
		return rune(s[0]), false, s[1:], nil
	}

	// Hard case: c is backslash.
	if len(s) <= 1 {
		err = ErrSyntax
		return
	}
	c := s[1]
	s = s[2:]

	switch c {
	case 'a':
		value = '\a'
	case 'b':
		value = '\b'
	case 'f':
		value = '\f'
	case 'n':
		value = '\n'
	case 'r':
		value = '\r'
	case 't':
		value = '\t'
	case 'v':
		value = '\v'
	case 'x', 'u', 'U':
		n := 0
		switch c {
		case 'x':
			n = 2
		case 'u':
			n = 4
		case 'U':
			n = 8
		}
		var v rune
		if len(s) < n {
			err = ErrSyntax
			return
		}
		for j := 0; j < n; j++ {
			x, ok := unhex(s[j])
			if !ok {
				err = ErrSyntax
				return
			}
			v = v<<4 | x
		}
		s = s[n:]
		if c == 'x' {
			// Single-byte string, possibly not UTF-8.
			value = v
			break
		}
		if v > utf8.MaxRune {
			err = ErrSyntax
			return
		}
		value = v
		multibyte = true
	case '0', '1', '2', '3', '4', '5', '6', '7':
		v := rune(c) - '0'
		if len(s) < 2 {
			err = ErrSyntax
			return
		}
		for j := 0; j < 2; j++ { // One digit already; two more.
			x := rune(s[j]) - '0'
			if x < 0 || x > 7 {
				err = ErrSyntax
				return
			}
			v = (v << 3) | x
		}
		s = s[2:]
		if v > 255 {
			err = ErrSyntax
			return
		}
		value = v
	case '\\':
		value = '\\'
	case '\'', '"':
		if c != quote {
			err = ErrSyntax
			return
		}
		value = rune(c)
	default:
		err = ErrSyntax
		return
	}
	tail = s
	return
}

// contains reports whether the string contains the byte c.
func contains(s string, c byte) bool {
	for i := 0; i < len(s); i++ {
		if s[i] == c {
			return true
		}
	}
	return false
}

func unhex(b byte) (v rune, ok bool) {
	c := rune(b)
	switch {
	case '0' <= c && c <= '9':
		return c - '0', true
	case 'a' <= c && c <= 'f':
		return c - 'a' + 10, true
	case 'A' <= c && c <= 'F':
		return c - 'A' + 10, true
	}
	return
}