1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
|
package pkglint
// Checks for patch files.
import "strings"
func CheckLinesPatch(lines *Lines, pkg *Package) {
(&PatchChecker{lines, NewLinesLexer(lines), false, false}).Check(pkg)
}
type PatchChecker struct {
lines *Lines
llex *LinesLexer
seenDocumentation bool
previousLineEmpty bool
}
const (
rePatchUniFileDel = `^---[\t ]([^\t ]+)(?:[\t ]+(.*))?$`
rePatchUniFileAdd = `^\+\+\+[\t ]([^\t ]+)(?:[\t ]+(.*))?$`
rePatchUniHunk = `^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@(.*)$`
)
func (ck *PatchChecker) Check(pkg *Package) {
if ck.lines.CheckCvsID(0, ``, "") {
ck.llex.Skip()
}
if ck.llex.EOF() {
ck.lines.Lines[0].Errorf("Patch files must not be empty.")
return
}
ck.previousLineEmpty = ck.llex.SkipEmptyOrNote()
var patchedFiles []Path
for !ck.llex.EOF() {
line := ck.llex.CurrentLine()
if ck.llex.SkipRegexp(rePatchUniFileDel) {
if m := ck.llex.NextRegexp(rePatchUniFileAdd); m != nil {
patchedFile := NewPath(m[1])
ck.checkBeginDiff(line, len(patchedFiles))
ck.checkUnifiedDiff(patchedFile)
patchedFiles = append(patchedFiles, patchedFile)
continue
}
ck.llex.Undo()
}
if m := ck.llex.NextRegexp(rePatchUniFileAdd); m != nil {
patchedFile := NewPath(m[1])
if ck.llex.SkipRegexp(rePatchUniFileDel) {
ck.checkBeginDiff(line, len(patchedFiles))
ck.llex.PreviousLine().Warnf("Unified diff headers should be first ---, then +++.")
ck.checkUnifiedDiff(patchedFile)
patchedFiles = append(patchedFiles, patchedFile)
continue
}
ck.llex.Undo()
}
if ck.llex.SkipRegexp(`^\*\*\*[\t ]([^\t ]+)(.*)$`) {
if ck.llex.SkipRegexp(`^---[\t ]([^\t ]+)(.*)$`) {
ck.checkBeginDiff(line, len(patchedFiles))
line.Warnf("Please use unified diffs (diff -u) for patches.")
return
}
ck.llex.Undo()
}
ck.llex.Skip()
ck.previousLineEmpty = ck.isEmptyLine(line.Text)
if !ck.previousLineEmpty {
ck.seenDocumentation = true
}
}
nPatched := len(patchedFiles)
if nPatched > 1 && !matches(ck.lines.Filename.String(), `\bCVE\b`) {
ck.lines.Whole().Warnf("Contains patches for %d files, should be only one.", nPatched)
}
if nPatched == 0 {
ck.lines.Whole().Errorf("Contains no patch.")
}
if len(patchedFiles) == 1 {
ck.checkCanonicalPatchName(patchedFiles[0])
}
CheckLinesTrailingEmptyLines(ck.lines)
sha1Before := computePatchSha1Hex(ck.lines)
if SaveAutofixChanges(ck.lines) && pkg != nil {
linesAfter := Load(ck.lines.Filename, 0)
sha1After := computePatchSha1Hex(linesAfter)
pkg.AutofixDistinfo(sha1Before, sha1After)
}
}
// See https://www.gnu.org/software/diffutils/manual/html_node/Detailed-Unified.html
func (ck *PatchChecker) checkUnifiedDiff(patchedFile Path) {
isConfigure := ck.isConfigure(patchedFile)
hasHunks := false
for {
m := ck.llex.NextRegexp(rePatchUniHunk)
if m == nil {
break
}
text := m[0]
hasHunks = true
linesToDel := toInt(m[2], 1)
linesToAdd := toInt(m[4], 1)
ck.checktextUniHunkCr()
ck.checktextCvsID(text)
for !ck.llex.EOF() && (linesToDel > 0 || linesToAdd > 0 || hasPrefix(ck.llex.CurrentLine().Text, "\\")) {
line := ck.llex.CurrentLine()
ck.llex.Skip()
text := line.Text
switch {
case text == "":
// There should be a space here, but that was a trailing space and
// has been trimmed down somewhere on its way. Doesn't matter,
// all the patch programs can handle this situation.
linesToDel--
linesToAdd--
case hasPrefix(text, " "), hasPrefix(text, "\t"):
linesToDel--
linesToAdd--
ck.checktextCvsID(text)
case hasPrefix(text, "-"):
linesToDel--
case hasPrefix(text, "+"):
linesToAdd--
ck.checktextCvsID(text)
ck.checkConfigure(text[1:], isConfigure)
ck.checkAddedLine(text[1:])
case hasPrefix(text, "\\"):
// \ No newline at end of file (or a translation of that message)
default:
line.Errorf("Invalid line in unified patch hunk: %s", text)
return
}
}
// When these two counts are equal, they may refer to context
// lines that consist only of whitespace and have therefore
// been lost during transmission. There is no way to detect
// this by looking only at the patch file.
if linesToAdd != linesToDel {
line := ck.llex.PreviousLine()
line.Warnf("Premature end of patch hunk (expected %d lines to be deleted and %d lines to be added).",
linesToDel, linesToAdd)
}
}
if !hasHunks {
ck.llex.CurrentLine().Errorf("No patch hunks for %q.", patchedFile.String())
}
if !ck.llex.EOF() {
line := ck.llex.CurrentLine()
if !ck.isEmptyLine(line.Text) && !matches(line.Text, rePatchUniFileDel) {
line.Warnf("Empty line or end of file expected.")
line.Explain(
"This line is not part of the patch anymore, although it may look so.",
"To make this situation clear, there should be an",
"empty line above this line.",
"If the line doesn't contain useful information, it should be removed.")
}
}
}
func (ck *PatchChecker) checkBeginDiff(line *Line, patchedFiles int) {
if !ck.seenDocumentation && patchedFiles == 0 {
line.Errorf("Each patch must be documented.")
line.Explain(
"Pkgsrc tries to have as few patches as possible.",
"Therefore, each patch must document why it is necessary.",
"Typical reasons are portability or security.",
"A typical documented patch looks like this:",
"",
"\t$"+"NetBSD$",
"",
"\tPortability fixes for GCC 4.8 on Linux.",
"\tSee https://github.com/org/repo/issues/7",
"",
"Patches that are related to a security issue should mention the",
"corresponding CVE identifier.",
"",
"Each patch should be sent to the upstream maintainers of the",
"package, so that they can include it in future versions.",
"After submitting a patch upstream, the corresponding bug report should",
"be mentioned in this file, to prevent duplicate work.")
}
if !ck.previousLineEmpty {
fix := line.Autofix()
fix.Notef("Empty line expected.")
fix.InsertAbove("")
fix.Apply()
}
}
func (ck *PatchChecker) checkConfigure(addedText string, isConfigure bool) {
if !isConfigure {
return
}
if !hasSuffix(addedText, ": Avoid regenerating within pkgsrc") {
return
}
line := ck.llex.PreviousLine()
line.Errorf("This code must not be included in patches.")
line.Explain(
"It is generated automatically by pkgsrc after the patch phase.",
"",
"For more details, look for \"configure-scripts-override\" in",
"mk/configure/gnu-configure.mk.")
}
func (ck *PatchChecker) checkAddedLine(addedText string) {
dirs := regcomp(`(?:^|[^.@)}])(/usr/pkg|/var|/etc)([^\w-]|$)`)
for _, m := range dirs.FindAllStringSubmatchIndex(addedText, -1) {
before := addedText[:m[2]]
dir := NewPath(addedText[m[2]:m[3]])
ck.checkAddedAbsPath(before, dir, addedText[m[4]:])
}
}
func (ck *PatchChecker) checkAddedAbsPath(before string, dir Path, after string) {
line := ck.llex.PreviousLine()
// Remove the #define from C and C++ macros.
before = replaceAll(before, `^[ \t]*#[ \t]*define[ \t]*\w+[ \t]*(.+)[ \t]*$`, "$1")
// Remove the "set(VAR" from CMakeLists.txt.
before = replaceAll(before, `^[ \t]*set\(\w+[ \t]*`, "")
// Ignore comments in shell programs.
if m, first := match1(before, `^[ \t]*#[ \t]*(\w*)`); m && first != "define" {
return
}
// Ignore paths inside C-style comments.
if contains(before, "/*") && contains(after, "*/") {
return
}
// Ignore composed C string literals such as PREFIX "/etc".
if matches(before, `\w+[ \t]*"$`) {
return
}
// Ignore shell literals such as $PREFIX/etc.
// But keep compiler options like -I/usr/pkg even though they look
// like a relative pathname.
if matches(before, `\w$`) && !matches(before, `(^|[ \t])-(I|L|R|rpath|Wl,-R)$`) {
return
}
switch dir {
case "/usr/pkg":
line.Errorf("Patches must not hard-code the pkgsrc PREFIX.")
line.Explain(
"Not every pkgsrc installation uses /usr/pkg as its PREFIX.",
"To keep the PREFIX configurable, the patch files should contain",
"the placeholder @PREFIX@ instead.",
"",
"In the pre-configure stage, this placeholder should then be",
"replaced with the actual configuration directory",
"using a SUBST block containing SUBST_VARS.dirs=PREFIX.",
"See mk/subst.mk for details.")
case "/var":
afterPath := NewPath(after)
if afterPath.HasPrefixPath("/tmp") || afterPath.HasPrefixPath("/shm") {
break
}
line.Errorf("Patches must not hard-code the pkgsrc VARBASE.")
line.Explain(
"Not every pkgsrc installation uses /var as its directory",
"for writable files.",
"To keep the VARBASE configurable, the patch files should",
"contain the placeholder @VARBASE@ instead.",
"",
"In the pre-configure stage, this placeholder should then be",
"replaced with the actual configuration directory",
"using a SUBST block containing SUBST_VARS.dirs=VARBASE.",
"See mk/subst.mk for details.")
default:
line.Errorf("Patches must not hard-code the pkgsrc PKG_SYSCONFDIR.")
line.Explain(
"Not every pkgsrc installation uses /etc as its directory",
"for configuration files.",
"To keep the PKG_SYSCONFDIR configurable, the patch files should",
"contain the placeholder @PKG_SYSCONFDIR@ instead.",
"",
"In the pre-configure stage, this placeholder should then be",
"replaced with the actual configuration directory",
"using a SUBST block containing SUBST_VARS.dirs=PKG_SYSCONFDIR.",
"See mk/subst.mk for details.")
}
}
func (ck *PatchChecker) checktextUniHunkCr() {
line := ck.llex.PreviousLine()
if !hasSuffix(line.Text, "\r") {
return
}
// This code has been introduced around 2006.
// As of 2018, this might be fixed by now.
fix := line.Autofix()
fix.Errorf("The hunk header must not end with a CR character.")
fix.Explain(
"The MacOS X patch utility cannot handle these.")
fix.Replace("\r\n", "\n")
fix.Apply()
}
func (ck *PatchChecker) checktextCvsID(text string) {
if strings.IndexByte(text, '$') == -1 {
return
}
if m, tagname := match1(text, `\$(Author|Date|Header|Id|Locker|Log|Name|RCSfile|Revision|Source|State|NetBSD)(?::[^\$]*)?\$`); m {
if matches(text, rePatchUniHunk) {
ck.llex.PreviousLine().Warnf("Found CVS tag \"$%s$\". Please remove it.", tagname)
} else {
ck.llex.PreviousLine().Warnf("Found CVS tag \"$%s$\". Please remove it by reducing the number of context lines using pkgdiff or \"diff -U[210]\".", tagname)
}
}
}
func (ck *PatchChecker) checkCanonicalPatchName(patched Path) {
patch := ck.lines.BaseName
if matches(patch, `^patch-[a-z][a-z]$`) {
// This naming scheme is only accepted for historic reasons.
// It has has absolutely no benefit.
return
}
if matches(patch, `^patch-[A-Z]+-[0-9]+`) {
return
}
// The patch name only needs to correspond very roughly to the patched file.
// There are varying schemes in use that transform a filename to a patch name.
normalize := func(s string) string {
return strings.ToLower(replaceAll(s, `[^A-Za-z0-9]+`, "*"))
}
patchedNorm := normalize(patched.Clean().String())
patchNorm := normalize(strings.TrimPrefix(patch, "patch-"))
if patchNorm == patchedNorm {
return
}
if hasSuffix(patchedNorm, patchNorm) && patchNorm == normalize(patched.Base()) {
return
}
// See pkgtools/pkgdiff/files/mkpatches, function patch_name.
canon1 := replaceAll(patched.Clean().String(), `_`, "__")
canon2 := replaceAll(canon1, `[/\s]`, "_")
canonicalName := "patch-" + canon2
ck.lines.Whole().Warnf(
"The patch file should be named %q to match the patched file %q.",
canonicalName, patched.String())
}
// isEmptyLine tests whether a line provides essentially no interesting content.
// The focus here is on human-generated content that is intended for other human readers.
// Therefore text that is typical for patch generators is considered empty as well.
func (*PatchChecker) isEmptyLine(text string) bool {
return text == "" ||
hasPrefix(text, "index ") ||
hasPrefix(text, "Index: ") ||
hasPrefix(text, "diff ") ||
hasPrefix(text, "=============")
}
func (*PatchChecker) isConfigure(filename Path) bool {
switch filename.Base() {
case "configure", "configure.in", "configure.ac":
return true
}
return false
}
|