// Copyright 2019 Google Inc. All rights reserved. // // 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 main import ( "bufio" "bytes" "encoding/json" "io" "os" "regexp" "strings" "unicode" ) type jsonWhitelist struct { Paths []string IgnoreMatchingLines []string } type whitelist struct { path string ignoreMatchingLines []string } func parseWhitelists(whitelists []string, whitelistFiles []string) ([]whitelist, error) { var ret []whitelist add := func(path string, ignoreMatchingLines []string) { for _, x := range ret { if x.path == path { x.ignoreMatchingLines = append(x.ignoreMatchingLines, ignoreMatchingLines...) return } } ret = append(ret, whitelist{ path: path, ignoreMatchingLines: ignoreMatchingLines, }) } for _, file := range whitelistFiles { newWhitelists, err := parseWhitelistFile(file) if err != nil { return nil, err } for _, w := range newWhitelists { add(w.path, w.ignoreMatchingLines) } } for _, s := range whitelists { colon := strings.IndexRune(s, ':') var ignoreMatchingLines []string if colon >= 0 { ignoreMatchingLines = []string{s[colon+1:]} } add(s, ignoreMatchingLines) } return ret, nil } func parseWhitelistFile(file string) ([]whitelist, error) { r, err := os.Open(file) if err != nil { return nil, err } defer r.Close() d := json.NewDecoder(newJSONCommentStripper(r)) var jsonWhitelists []jsonWhitelist err = d.Decode(&jsonWhitelists) var whitelists []whitelist for _, w := range jsonWhitelists { for _, p := range w.Paths { whitelists = append(whitelists, whitelist{ path: p, ignoreMatchingLines: w.IgnoreMatchingLines, }) } } return whitelists, err } func filterModifiedPaths(l [][2]*ZipArtifactFile, whitelists []whitelist) ([][2]*ZipArtifactFile, error) { outer: for i := 0; i < len(l); i++ { for _, w := range whitelists { if match, err := Match(w.path, l[i][0].Name); err != nil { return l, err } else if match { if match, err := diffIgnoringMatchingLines(l[i][0], l[i][1], w.ignoreMatchingLines); err != nil { return l, err } else if match || len(w.ignoreMatchingLines) == 0 { l = append(l[:i], l[i+1:]...) i-- } continue outer } } } if len(l) == 0 { l = nil } return l, nil } func filterNewPaths(l []*ZipArtifactFile, whitelists []whitelist) ([]*ZipArtifactFile, error) { outer: for i := 0; i < len(l); i++ { for _, w := range whitelists { if match, err := Match(w.path, l[i].Name); err != nil { return l, err } else if match && len(w.ignoreMatchingLines) == 0 { l = append(l[:i], l[i+1:]...) i-- } continue outer } } if len(l) == 0 { l = nil } return l, nil } func diffIgnoringMatchingLines(a *ZipArtifactFile, b *ZipArtifactFile, ignoreMatchingLines []string) (match bool, err error) { lineMatchesIgnores := func(b []byte) (bool, error) { for _, m := range ignoreMatchingLines { if match, err := regexp.Match(m, b); err != nil { return false, err } else if match { return match, nil } } return false, nil } filter := func(z *ZipArtifactFile) ([]byte, error) { var ret []byte r, err := z.Open() if err != nil { return nil, err } s := bufio.NewScanner(r) for s.Scan() { if match, err := lineMatchesIgnores(s.Bytes()); err != nil { return nil, err } else if !match { ret = append(ret, "\n"...) ret = append(ret, s.Bytes()...) } } return ret, nil } bufA, err := filter(a) if err != nil { return false, err } bufB, err := filter(b) if err != nil { return false, err } return bytes.Compare(bufA, bufB) == 0, nil } func applyWhitelists(diff zipDiff, whitelists []whitelist) (zipDiff, error) { var err error diff.modified, err = filterModifiedPaths(diff.modified, whitelists) if err != nil { return diff, err } diff.onlyInA, err = filterNewPaths(diff.onlyInA, whitelists) if err != nil { return diff, err } diff.onlyInB, err = filterNewPaths(diff.onlyInB, whitelists) if err != nil { return diff, err } return diff, nil } func newJSONCommentStripper(r io.Reader) *jsonCommentStripper { return &jsonCommentStripper{ r: bufio.NewReader(r), } } type jsonCommentStripper struct { r *bufio.Reader b []byte err error } func (j *jsonCommentStripper) Read(buf []byte) (int, error) { for len(j.b) == 0 { if j.err != nil { return 0, j.err } j.b, j.err = j.r.ReadBytes('\n') if isComment(j.b) { j.b = nil } } n := copy(buf, j.b) j.b = j.b[n:] return n, nil } var commentPrefix = []byte("//") func isComment(b []byte) bool { for len(b) > 0 && unicode.IsSpace(rune(b[0])) { b = b[1:] } return bytes.HasPrefix(b, commentPrefix) }