format-patch: make rfc2047 encoding more strict
RFC 2047 requires more characters to be encoded than it is currently done. Especially, RFC 2047 distinguishes between allowed remaining characters in encoded words in addresses (From, To, etc.) and other headers, such as Subject. Make add_rfc2047() and is_rfc2047_special() location dependent and include all non-allowed characters to hopefully be RFC 2047 conformant. This especially fixes a problem, where RFC 822 specials (e. g. ".") were left unencoded in addresses, which was solved with a non-standard-conforming workaround in the past (which is going to be removed in a follow-up patch). Signed-off-by: Jan H. Schönherr <schnhrr@cs.tu-berlin.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jan H. Schönherr committed
Oct 18, 2012 at 16:43 UTC
0fcec2ce5467740eb613da6a977deca3e7e866ba
3 files changed
+72
-12
git-compat-util.h
+2
@@ -466,6 +466,7 @@ extern const char tolower_trans_tbl[256];
466
#undef isdigit
467
#undef isalpha
468
#undef isalnum
469
+#undef isprint
470
#undef islower
471
#undef isupper
472
#undef tolower
@@ -483,6 +484,7 @@ extern unsigned char sane_ctype[256];
484
#define isdigit(x) sane_istest(x,GIT_DIGIT)
485
#define isalpha(x) sane_istest(x,GIT_ALPHA)
486
#define isalnum(x) sane_istest(x,GIT_ALPHA | GIT_DIGIT)
487
+#define isprint(x) ((x) >= 0x20 && (x) <= 0x7e)
488
#define islower(x) sane_iscase(x, 1)
489
#define isupper(x) sane_iscase(x, 0)
490
#define is_glob_special(x) sane_istest(x,GIT_GLOB_SPECIAL)
pretty.c
+59
-8
@@ -272,16 +272,65 @@ static void add_rfc822_quoted(struct strbuf *out, const char *s, int len)
272
strbuf_addch(out, '"');
273
}
274
275
-static int is_rfc2047_special(char ch)
275
+enum rfc2047_type {
276
+ RFC2047_SUBJECT,
277
+ RFC2047_ADDRESS,
278
+};
279
+
280
+static int is_rfc2047_special(char ch, enum rfc2047_type type)
281
{
277
- if (ch == ' ' || ch == '\n')
282
+ /*
283
+ * rfc2047, section 4.2:
284
+ *
285
+ * 8-bit values which correspond to printable ASCII characters other
286
+ * than "=", "?", and "_" (underscore), MAY be represented as those
287
+ * characters. (But see section 5 for restrictions.) In
288
+ * particular, SPACE and TAB MUST NOT be represented as themselves
289
+ * within encoded words.
290
+ */
291
+
292
+ /*
293
+ * rule out non-ASCII characters and non-printable characters (the
294
+ * non-ASCII check should be redundant as isprint() is not localized
295
+ * and only knows about ASCII, but be defensive about that)
296
+ */
297
+ if (non_ascii(ch) || !isprint(ch))
298
+ return 1;
299
+
300
+ /*
301
+ * rule out special printable characters (' ' should be the only
302
+ * whitespace character considered printable, but be defensive and use
303
+ * isspace())
304
+ */
305
+ if (isspace(ch) || ch == '=' || ch == '?' || ch == '_')
306
return 1;
307
280
- return (non_ascii(ch) || (ch == '=') || (ch == '?') || (ch == '_'));
308
+ /*
309
+ * rfc2047, section 5.3:
310
+ *
311
+ * As a replacement for a 'word' entity within a 'phrase', for example,
312
+ * one that precedes an address in a From, To, or Cc header. The ABNF
313
+ * definition for 'phrase' from RFC 822 thus becomes:
314
+ *
315
+ * phrase = 1*( encoded-word / word )
316
+ *
317
+ * In this case the set of characters that may be used in a "Q"-encoded
318
+ * 'encoded-word' is restricted to: <upper and lower case ASCII
319
+ * letters, decimal digits, "!", "*", "+", "-", "/", "=", and "_"
320
+ * (underscore, ASCII 95.)>. An 'encoded-word' that appears within a
321
+ * 'phrase' MUST be separated from any adjacent 'word', 'text' or
322
+ * 'special' by 'linear-white-space'.
323
+ */
324
+
325
+ if (type != RFC2047_ADDRESS)
326
+ return 0;
327
+
328
+ /* '=' and '_' are special cases and have been checked above */
329
+ return !(isalnum(ch) || ch == '!' || ch == '*' || ch == '+' || ch == '-' || ch == '/');
330
}
331
332
static void add_rfc2047(struct strbuf *sb, const char *line, int len,
284
- const char *encoding)
333
+ const char *encoding, enum rfc2047_type type)
334
{
335
static const int max_length = 78; /* per rfc2822 */
336
static const int max_encoded_length = 76; /* per rfc2047 */
@@ -304,7 +353,7 @@ needquote:
353
line_len += strlen(encoding) + 5; /* 5 for =??q? */
354
for (i = 0; i < len; i++) {
355
unsigned ch = line[i] & 0xFF;
307
- int is_special = is_rfc2047_special(ch);
356
+ int is_special = is_rfc2047_special(ch, type);
357
358
/*
359
* According to RFC 2047, we could encode the special character
@@ -358,11 +407,13 @@ void pp_user_info(const struct pretty_print_context *pp,
407
display_name_length = name_tail - line;
408
strbuf_addstr(sb, "From: ");
409
if (!has_rfc822_specials(line, display_name_length)) {
361
- add_rfc2047(sb, line, display_name_length, encoding);
410
+ add_rfc2047(sb, line, display_name_length,
411
+ encoding, RFC2047_ADDRESS);
412
} else {
413
struct strbuf quoted = STRBUF_INIT;
414
add_rfc822_quoted("ed, line, display_name_length);
365
- add_rfc2047(sb, quoted.buf, quoted.len, encoding);
415
+ add_rfc2047(sb, quoted.buf, quoted.len,
416
+ encoding, RFC2047_ADDRESS);
417
strbuf_release("ed);
418
}
419
if (namelen - display_name_length + last_line_length(sb) > 78) {
@@ -1294,7 +1345,7 @@ void pp_title_line(const struct pretty_print_context *pp,
1345
strbuf_grow(sb, title.len + 1024);
1346
if (pp->subject) {
1347
strbuf_addstr(sb, pp->subject);
1297
- add_rfc2047(sb, title.buf, title.len, encoding);
1348
+ add_rfc2047(sb, title.buf, title.len, encoding, RFC2047_SUBJECT);
1349
} else {
1350
strbuf_addbuf(sb, &title);
1351
}
t/t4014-format-patch.sh
+11
-4
@@ -818,21 +818,28 @@ check_author() {
818
cat >expect <<'EOF'
819
From: "Foo B. Bar" <author@example.com>
820
EOF
821
-test_expect_success 'format-patch quotes dot in headers' '
821
+test_expect_success 'format-patch quotes dot in from-headers' '
822
check_author "Foo B. Bar"
823
'
824
825
cat >expect <<'EOF'
826
From: "Foo \"The Baz\" Bar" <author@example.com>
827
EOF
828
-test_expect_success 'format-patch quotes double-quote in headers' '
828
+test_expect_success 'format-patch quotes double-quote in from-headers' '
829
check_author "Foo \"The Baz\" Bar"
830
'
831
832
cat >expect <<'EOF'
833
-From: =?UTF-8?q?"F=C3=B6o=20B.=20Bar"?= <author@example.com>
833
+From: =?UTF-8?q?F=C3=B6o=20Bar?= <author@example.com>
834
EOF
835
-test_expect_success 'rfc2047-encoded headers also double-quote 822 specials' '
835
+test_expect_success 'format-patch uses rfc2047-encoded from-headers when necessary' '
836
+ check_author "Föo Bar"
837
+'
838
+
839
+cat >expect <<'EOF'
840
+From: =?UTF-8?q?F=C3=B6o=20B=2E=20Bar?= <author@example.com>
841
+EOF
842
+test_expect_failure 'rfc2047-encoded from-headers leave no rfc822 specials' '
843
check_author "Föo B. Bar"
844
'
845