gitea/vendor/golang.org/x/tools/internal/module/module.go

541 lines
17 KiB
Go
Raw Normal View History

// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package module defines the module.Version type
// along with support code.
package module
// IMPORTANT NOTE
//
// This file essentially defines the set of valid import paths for the go command.
// There are many subtle considerations, including Unicode ambiguity,
// security, network, and file system representations.
//
// This file also defines the set of valid module path and version combinations,
// another topic with many subtle considerations.
//
// Changes to the semantics in this file require approval from rsc.
import (
"fmt"
"sort"
"strings"
"unicode"
"unicode/utf8"
"golang.org/x/tools/internal/semver"
)
// A Version is defined by a module path and version pair.
type Version struct {
Path string
// Version is usually a semantic version in canonical form.
// There are two exceptions to this general rule.
// First, the top-level target of a build has no specific version
// and uses Version = "".
// Second, during MVS calculations the version "none" is used
// to represent the decision to take no version of a given module.
Version string `json:",omitempty"`
}
// Check checks that a given module path, version pair is valid.
// In addition to the path being a valid module path
// and the version being a valid semantic version,
// the two must correspond.
// For example, the path "yaml/v2" only corresponds to
// semantic versions beginning with "v2.".
func Check(path, version string) error {
if err := CheckPath(path); err != nil {
return err
}
if !semver.IsValid(version) {
return fmt.Errorf("malformed semantic version %v", version)
}
_, pathMajor, _ := SplitPathVersion(path)
if !MatchPathMajor(version, pathMajor) {
if pathMajor == "" {
pathMajor = "v0 or v1"
}
if pathMajor[0] == '.' { // .v1
pathMajor = pathMajor[1:]
}
return fmt.Errorf("mismatched module path %v and version %v (want %v)", path, version, pathMajor)
}
return nil
}
// firstPathOK reports whether r can appear in the first element of a module path.
// The first element of the path must be an LDH domain name, at least for now.
// To avoid case ambiguity, the domain name must be entirely lower case.
func firstPathOK(r rune) bool {
return r == '-' || r == '.' ||
'0' <= r && r <= '9' ||
'a' <= r && r <= 'z'
}
// pathOK reports whether r can appear in an import path element.
// Paths can be ASCII letters, ASCII digits, and limited ASCII punctuation: + - . _ and ~.
// This matches what "go get" has historically recognized in import paths.
// TODO(rsc): We would like to allow Unicode letters, but that requires additional
// care in the safe encoding (see note below).
func pathOK(r rune) bool {
if r < utf8.RuneSelf {
return r == '+' || r == '-' || r == '.' || r == '_' || r == '~' ||
'0' <= r && r <= '9' ||
'A' <= r && r <= 'Z' ||
'a' <= r && r <= 'z'
}
return false
}
// fileNameOK reports whether r can appear in a file name.
// For now we allow all Unicode letters but otherwise limit to pathOK plus a few more punctuation characters.
// If we expand the set of allowed characters here, we have to
// work harder at detecting potential case-folding and normalization collisions.
// See note about "safe encoding" below.
func fileNameOK(r rune) bool {
if r < utf8.RuneSelf {
// Entire set of ASCII punctuation, from which we remove characters:
// ! " # $ % & ' ( ) * + , - . / : ; < = > ? @ [ \ ] ^ _ ` { | } ~
// We disallow some shell special characters: " ' * < > ? ` |
// (Note that some of those are disallowed by the Windows file system as well.)
// We also disallow path separators / : and \ (fileNameOK is only called on path element characters).
// We allow spaces (U+0020) in file names.
const allowed = "!#$%&()+,-.=@[]^_{}~ "
if '0' <= r && r <= '9' || 'A' <= r && r <= 'Z' || 'a' <= r && r <= 'z' {
return true
}
for i := 0; i < len(allowed); i++ {
if rune(allowed[i]) == r {
return true
}
}
return false
}
// It may be OK to add more ASCII punctuation here, but only carefully.
// For example Windows disallows < > \, and macOS disallows :, so we must not allow those.
return unicode.IsLetter(r)
}
// CheckPath checks that a module path is valid.
func CheckPath(path string) error {
if err := checkPath(path, false); err != nil {
return fmt.Errorf("malformed module path %q: %v", path, err)
}
i := strings.Index(path, "/")
if i < 0 {
i = len(path)
}
if i == 0 {
return fmt.Errorf("malformed module path %q: leading slash", path)
}
if !strings.Contains(path[:i], ".") {
return fmt.Errorf("malformed module path %q: missing dot in first path element", path)
}
if path[0] == '-' {
return fmt.Errorf("malformed module path %q: leading dash in first path element", path)
}
for _, r := range path[:i] {
if !firstPathOK(r) {
return fmt.Errorf("malformed module path %q: invalid char %q in first path element", path, r)
}
}
if _, _, ok := SplitPathVersion(path); !ok {
return fmt.Errorf("malformed module path %q: invalid version", path)
}
return nil
}
// CheckImportPath checks that an import path is valid.
func CheckImportPath(path string) error {
if err := checkPath(path, false); err != nil {
return fmt.Errorf("malformed import path %q: %v", path, err)
}
return nil
}
// checkPath checks that a general path is valid.
// It returns an error describing why but not mentioning path.
// Because these checks apply to both module paths and import paths,
// the caller is expected to add the "malformed ___ path %q: " prefix.
// fileName indicates whether the final element of the path is a file name
// (as opposed to a directory name).
func checkPath(path string, fileName bool) error {
if !utf8.ValidString(path) {
return fmt.Errorf("invalid UTF-8")
}
if path == "" {
return fmt.Errorf("empty string")
}
if strings.Contains(path, "..") {
return fmt.Errorf("double dot")
}
if strings.Contains(path, "//") {
return fmt.Errorf("double slash")
}
if path[len(path)-1] == '/' {
return fmt.Errorf("trailing slash")
}
elemStart := 0
for i, r := range path {
if r == '/' {
if err := checkElem(path[elemStart:i], fileName); err != nil {
return err
}
elemStart = i + 1
}
}
if err := checkElem(path[elemStart:], fileName); err != nil {
return err
}
return nil
}
// checkElem checks whether an individual path element is valid.
// fileName indicates whether the element is a file name (not a directory name).
func checkElem(elem string, fileName bool) error {
if elem == "" {
return fmt.Errorf("empty path element")
}
if strings.Count(elem, ".") == len(elem) {
return fmt.Errorf("invalid path element %q", elem)
}
if elem[0] == '.' && !fileName {
return fmt.Errorf("leading dot in path element")
}
if elem[len(elem)-1] == '.' {
return fmt.Errorf("trailing dot in path element")
}
charOK := pathOK
if fileName {
charOK = fileNameOK
}
for _, r := range elem {
if !charOK(r) {
return fmt.Errorf("invalid char %q", r)
}
}
// Windows disallows a bunch of path elements, sadly.
// See https://docs.microsoft.com/en-us/windows/desktop/fileio/naming-a-file
short := elem
if i := strings.Index(short, "."); i >= 0 {
short = short[:i]
}
for _, bad := range badWindowsNames {
if strings.EqualFold(bad, short) {
return fmt.Errorf("disallowed path element %q", elem)
}
}
return nil
}
// CheckFilePath checks whether a slash-separated file path is valid.
func CheckFilePath(path string) error {
if err := checkPath(path, true); err != nil {
return fmt.Errorf("malformed file path %q: %v", path, err)
}
return nil
}
// badWindowsNames are the reserved file path elements on Windows.
// See https://docs.microsoft.com/en-us/windows/desktop/fileio/naming-a-file
var badWindowsNames = []string{
"CON",
"PRN",
"AUX",
"NUL",
"COM1",
"COM2",
"COM3",
"COM4",
"COM5",
"COM6",
"COM7",
"COM8",
"COM9",
"LPT1",
"LPT2",
"LPT3",
"LPT4",
"LPT5",
"LPT6",
"LPT7",
"LPT8",
"LPT9",
}
// SplitPathVersion returns prefix and major version such that prefix+pathMajor == path
// and version is either empty or "/vN" for N >= 2.
// As a special case, gopkg.in paths are recognized directly;
// they require ".vN" instead of "/vN", and for all N, not just N >= 2.
func SplitPathVersion(path string) (prefix, pathMajor string, ok bool) {
if strings.HasPrefix(path, "gopkg.in/") {
return splitGopkgIn(path)
}
i := len(path)
dot := false
for i > 0 && ('0' <= path[i-1] && path[i-1] <= '9' || path[i-1] == '.') {
if path[i-1] == '.' {
dot = true
}
i--
}
if i <= 1 || i == len(path) || path[i-1] != 'v' || path[i-2] != '/' {
return path, "", true
}
prefix, pathMajor = path[:i-2], path[i-2:]
if dot || len(pathMajor) <= 2 || pathMajor[2] == '0' || pathMajor == "/v1" {
return path, "", false
}
return prefix, pathMajor, true
}
// splitGopkgIn is like SplitPathVersion but only for gopkg.in paths.
func splitGopkgIn(path string) (prefix, pathMajor string, ok bool) {
if !strings.HasPrefix(path, "gopkg.in/") {
return path, "", false
}
i := len(path)
if strings.HasSuffix(path, "-unstable") {
i -= len("-unstable")
}
for i > 0 && ('0' <= path[i-1] && path[i-1] <= '9') {
i--
}
if i <= 1 || path[i-1] != 'v' || path[i-2] != '.' {
// All gopkg.in paths must end in vN for some N.
return path, "", false
}
prefix, pathMajor = path[:i-2], path[i-2:]
if len(pathMajor) <= 2 || pathMajor[2] == '0' && pathMajor != ".v0" {
return path, "", false
}
return prefix, pathMajor, true
}
// MatchPathMajor reports whether the semantic version v
// matches the path major version pathMajor.
func MatchPathMajor(v, pathMajor string) bool {
if strings.HasPrefix(pathMajor, ".v") && strings.HasSuffix(pathMajor, "-unstable") {
pathMajor = strings.TrimSuffix(pathMajor, "-unstable")
}
if strings.HasPrefix(v, "v0.0.0-") && pathMajor == ".v1" {
// Allow old bug in pseudo-versions that generated v0.0.0- pseudoversion for gopkg .v1.
// For example, gopkg.in/yaml.v2@v2.2.1's go.mod requires gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405.
return true
}
m := semver.Major(v)
if pathMajor == "" {
return m == "v0" || m == "v1" || semver.Build(v) == "+incompatible"
}
return (pathMajor[0] == '/' || pathMajor[0] == '.') && m == pathMajor[1:]
}
// CanonicalVersion returns the canonical form of the version string v.
// It is the same as semver.Canonical(v) except that it preserves the special build suffix "+incompatible".
func CanonicalVersion(v string) string {
cv := semver.Canonical(v)
if semver.Build(v) == "+incompatible" {
cv += "+incompatible"
}
return cv
}
// Sort sorts the list by Path, breaking ties by comparing Versions.
func Sort(list []Version) {
sort.Slice(list, func(i, j int) bool {
mi := list[i]
mj := list[j]
if mi.Path != mj.Path {
return mi.Path < mj.Path
}
// To help go.sum formatting, allow version/file.
// Compare semver prefix by semver rules,
// file by string order.
vi := mi.Version
vj := mj.Version
var fi, fj string
if k := strings.Index(vi, "/"); k >= 0 {
vi, fi = vi[:k], vi[k:]
}
if k := strings.Index(vj, "/"); k >= 0 {
vj, fj = vj[:k], vj[k:]
}
if vi != vj {
return semver.Compare(vi, vj) < 0
}
return fi < fj
})
}
// Safe encodings
//
// Module paths appear as substrings of file system paths
// (in the download cache) and of web server URLs in the proxy protocol.
// In general we cannot rely on file systems to be case-sensitive,
// nor can we rely on web servers, since they read from file systems.
// That is, we cannot rely on the file system to keep rsc.io/QUOTE
// and rsc.io/quote separate. Windows and macOS don't.
// Instead, we must never require two different casings of a file path.
// Because we want the download cache to match the proxy protocol,
// and because we want the proxy protocol to be possible to serve
// from a tree of static files (which might be stored on a case-insensitive
// file system), the proxy protocol must never require two different casings
// of a URL path either.
//
// One possibility would be to make the safe encoding be the lowercase
// hexadecimal encoding of the actual path bytes. This would avoid ever
// needing different casings of a file path, but it would be fairly illegible
// to most programmers when those paths appeared in the file system
// (including in file paths in compiler errors and stack traces)
// in web server logs, and so on. Instead, we want a safe encoding that
// leaves most paths unaltered.
//
// The safe encoding is this:
// replace every uppercase letter with an exclamation mark
// followed by the letter's lowercase equivalent.
//
// For example,
// github.com/Azure/azure-sdk-for-go -> github.com/!azure/azure-sdk-for-go.
// github.com/GoogleCloudPlatform/cloudsql-proxy -> github.com/!google!cloud!platform/cloudsql-proxy
// github.com/Sirupsen/logrus -> github.com/!sirupsen/logrus.
//
// Import paths that avoid upper-case letters are left unchanged.
// Note that because import paths are ASCII-only and avoid various
// problematic punctuation (like : < and >), the safe encoding is also ASCII-only
// and avoids the same problematic punctuation.
//
// Import paths have never allowed exclamation marks, so there is no
// need to define how to encode a literal !.
//
// Although paths are disallowed from using Unicode (see pathOK above),
// the eventual plan is to allow Unicode letters as well, to assume that
// file systems and URLs are Unicode-safe (storing UTF-8), and apply
// the !-for-uppercase convention. Note however that not all runes that
// are different but case-fold equivalent are an upper/lower pair.
// For example, U+004B ('K'), U+006B ('k'), and U+212A ('' for Kelvin)
// are considered to case-fold to each other. When we do add Unicode
// letters, we must not assume that upper/lower are the only case-equivalent pairs.
// Perhaps the Kelvin symbol would be disallowed entirely, for example.
// Or perhaps it would encode as "!!k", or perhaps as "(212A)".
//
// Also, it would be nice to allow Unicode marks as well as letters,
// but marks include combining marks, and then we must deal not
// only with case folding but also normalization: both U+00E9 ('é')
// and U+0065 U+0301 ('e' followed by combining acute accent)
// look the same on the page and are treated by some file systems
// as the same path. If we do allow Unicode marks in paths, there
// must be some kind of normalization to allow only one canonical
// encoding of any character used in an import path.
// EncodePath returns the safe encoding of the given module path.
// It fails if the module path is invalid.
func EncodePath(path string) (encoding string, err error) {
if err := CheckPath(path); err != nil {
return "", err
}
return encodeString(path)
}
// EncodeVersion returns the safe encoding of the given module version.
// Versions are allowed to be in non-semver form but must be valid file names
// and not contain exclamation marks.
func EncodeVersion(v string) (encoding string, err error) {
if err := checkElem(v, true); err != nil || strings.Contains(v, "!") {
return "", fmt.Errorf("disallowed version string %q", v)
}
return encodeString(v)
}
func encodeString(s string) (encoding string, err error) {
haveUpper := false
for _, r := range s {
if r == '!' || r >= utf8.RuneSelf {
// This should be disallowed by CheckPath, but diagnose anyway.
// The correctness of the encoding loop below depends on it.
return "", fmt.Errorf("internal error: inconsistency in EncodePath")
}
if 'A' <= r && r <= 'Z' {
haveUpper = true
}
}
if !haveUpper {
return s, nil
}
var buf []byte
for _, r := range s {
if 'A' <= r && r <= 'Z' {
buf = append(buf, '!', byte(r+'a'-'A'))
} else {
buf = append(buf, byte(r))
}
}
return string(buf), nil
}
// DecodePath returns the module path of the given safe encoding.
// It fails if the encoding is invalid or encodes an invalid path.
func DecodePath(encoding string) (path string, err error) {
path, ok := decodeString(encoding)
if !ok {
return "", fmt.Errorf("invalid module path encoding %q", encoding)
}
if err := CheckPath(path); err != nil {
return "", fmt.Errorf("invalid module path encoding %q: %v", encoding, err)
}
return path, nil
}
// DecodeVersion returns the version string for the given safe encoding.
// It fails if the encoding is invalid or encodes an invalid version.
// Versions are allowed to be in non-semver form but must be valid file names
// and not contain exclamation marks.
func DecodeVersion(encoding string) (v string, err error) {
v, ok := decodeString(encoding)
if !ok {
return "", fmt.Errorf("invalid version encoding %q", encoding)
}
if err := checkElem(v, true); err != nil {
return "", fmt.Errorf("disallowed version string %q", v)
}
return v, nil
}
func decodeString(encoding string) (string, bool) {
var buf []byte
bang := false
for _, r := range encoding {
if r >= utf8.RuneSelf {
return "", false
}
if bang {
bang = false
if r < 'a' || 'z' < r {
return "", false
}
buf = append(buf, byte(r+'A'-'a'))
continue
}
if r == '!' {
bang = true
continue
}
if 'A' <= r && r <= 'Z' {
return "", false
}
buf = append(buf, byte(r))
}
if bang {
return "", false
}
return string(buf), true
}