]>
Commit | Line | Data |
---|---|---|
2744b234 LT |
1 | /* |
2 | * Another stupid program, this one parsing the headers of an | |
3 | * email to figure out authorship and subject | |
4 | */ | |
f1f909e3 | 5 | #include "cache.h" |
34488e3c | 6 | #include "builtin.h" |
b45974a6 | 7 | #include "utf8.h" |
3b6121f6 | 8 | #include "strbuf.h" |
2744b234 | 9 | |
8e919277 JH |
10 | #define MAX_BOUNDARIES 5 |
11 | ||
c69f2395 | 12 | struct mailinfo { |
173aef7c JH |
13 | FILE *input; |
14 | FILE *output; | |
8f63588a | 15 | FILE *patchfile; |
173aef7c | 16 | |
c69f2395 JH |
17 | struct strbuf name; |
18 | struct strbuf email; | |
849106d5 JH |
19 | int keep_subject; |
20 | int keep_non_patch_brackets_in_subject; | |
6200b751 | 21 | int add_message_id; |
ad57ef9d JH |
22 | int use_scissors; |
23 | int use_inbody_headers; | |
28be2d08 | 24 | const char *metainfo_charset; |
13c6df26 | 25 | |
8e919277 JH |
26 | struct strbuf *content[MAX_BOUNDARIES]; |
27 | struct strbuf **content_top; | |
f1e037b9 | 28 | struct strbuf charset; |
6200b751 | 29 | char *message_id; |
ab50e38b JH |
30 | enum { |
31 | TE_DONTCARE, TE_QP, TE_BASE64 | |
32 | } transfer_encoding; | |
43550efa | 33 | int patch_lines; |
13c6df26 JH |
34 | int filter_stage; /* still reading log or are we copying patch? */ |
35 | int header_stage; /* still checking in-body headers? */ | |
d895bf0f JH |
36 | struct strbuf **p_hdr_data; |
37 | struct strbuf **s_hdr_data; | |
05e625e5 JH |
38 | |
39 | struct strbuf log_message; | |
c69f2395 | 40 | }; |
2744b234 | 41 | |
30f50c34 JH |
42 | static void cleanup_space(struct strbuf *sb) |
43 | { | |
44 | size_t pos, cnt; | |
45 | for (pos = 0; pos < sb->len; pos++) { | |
46 | if (isspace(sb->buf[pos])) { | |
47 | sb->buf[pos] = ' '; | |
48 | for (cnt = 0; isspace(sb->buf[pos + cnt + 1]); cnt++); | |
49 | strbuf_remove(sb, pos + 1, cnt); | |
50 | } | |
51 | } | |
52 | } | |
08e6710f | 53 | |
3b6121f6 | 54 | static void get_sane_name(struct strbuf *out, struct strbuf *name, struct strbuf *email) |
2744b234 | 55 | { |
3b6121f6 LS |
56 | struct strbuf *src = name; |
57 | if (name->len < 3 || 60 < name->len || strchr(name->buf, '@') || | |
58 | strchr(name->buf, '<') || strchr(name->buf, '>')) | |
59 | src = email; | |
60 | else if (name == out) | |
61 | return; | |
62 | strbuf_reset(out); | |
63 | strbuf_addbuf(out, src); | |
2744b234 LT |
64 | } |
65 | ||
c69f2395 | 66 | static void parse_bogus_from(struct mailinfo *mi, const struct strbuf *line) |
e0e3ba20 JH |
67 | { |
68 | /* John Doe <johndoe> */ | |
e0e3ba20 | 69 | |
3b6121f6 | 70 | char *bra, *ket; |
e0e3ba20 JH |
71 | /* This is fallback, so do not bother if we already have an |
72 | * e-mail address. | |
34488e3c | 73 | */ |
c69f2395 | 74 | if (mi->email.len) |
3b6121f6 | 75 | return; |
e0e3ba20 | 76 | |
3b6121f6 | 77 | bra = strchr(line->buf, '<'); |
e0e3ba20 | 78 | if (!bra) |
3b6121f6 | 79 | return; |
e0e3ba20 JH |
80 | ket = strchr(bra, '>'); |
81 | if (!ket) | |
3b6121f6 | 82 | return; |
e0e3ba20 | 83 | |
c69f2395 JH |
84 | strbuf_reset(&mi->email); |
85 | strbuf_add(&mi->email, bra + 1, ket - bra - 1); | |
3b6121f6 | 86 | |
c69f2395 JH |
87 | strbuf_reset(&mi->name); |
88 | strbuf_add(&mi->name, line->buf, bra - line->buf); | |
89 | strbuf_trim(&mi->name); | |
90 | get_sane_name(&mi->name, &mi->name, &mi->email); | |
e0e3ba20 JH |
91 | } |
92 | ||
c69f2395 | 93 | static void handle_from(struct mailinfo *mi, const struct strbuf *from) |
2744b234 | 94 | { |
2dec02b1 | 95 | char *at; |
3b6121f6 LS |
96 | size_t el; |
97 | struct strbuf f; | |
2744b234 | 98 | |
3b6121f6 LS |
99 | strbuf_init(&f, from->len); |
100 | strbuf_addbuf(&f, from); | |
101 | ||
102 | at = strchr(f.buf, '@'); | |
103 | if (!at) { | |
c69f2395 | 104 | parse_bogus_from(mi, from); |
3b6121f6 LS |
105 | return; |
106 | } | |
2744b234 LT |
107 | |
108 | /* | |
109 | * If we already have one email, don't take any confusing lines | |
110 | */ | |
c69f2395 | 111 | if (mi->email.len && strchr(at + 1, '@')) { |
3b6121f6 LS |
112 | strbuf_release(&f); |
113 | return; | |
114 | } | |
2744b234 | 115 | |
d4a9ce78 | 116 | /* Pick up the string around '@', possibly delimited with <> |
3b6121f6 | 117 | * pair; that is the email part. |
d4a9ce78 | 118 | */ |
3b6121f6 | 119 | while (at > f.buf) { |
2744b234 | 120 | char c = at[-1]; |
d4a9ce78 JH |
121 | if (isspace(c)) |
122 | break; | |
123 | if (c == '<') { | |
124 | at[-1] = ' '; | |
2744b234 | 125 | break; |
d4a9ce78 | 126 | } |
2744b234 LT |
127 | at--; |
128 | } | |
3b6121f6 | 129 | el = strcspn(at, " \n\t\r\v\f>"); |
c69f2395 JH |
130 | strbuf_reset(&mi->email); |
131 | strbuf_add(&mi->email, at, el); | |
e9d7d10a | 132 | strbuf_remove(&f, at - f.buf, el + (at[el] ? 1 : 0)); |
2744b234 | 133 | |
08e6710f KS |
134 | /* The remainder is name. It could be |
135 | * | |
136 | * - "John Doe <john.doe@xz>" (a), or | |
137 | * - "john.doe@xz (John Doe)" (b), or | |
138 | * - "John (zzz) Doe <john.doe@xz> (Comment)" (c) | |
139 | * | |
140 | * but we have removed the email part, so | |
141 | * | |
142 | * - remove extra spaces which could stay after email (case 'c'), and | |
143 | * - trim from both ends, possibly removing the () pair at the end | |
144 | * (cases 'a' and 'b'). | |
d4a9ce78 | 145 | */ |
08e6710f | 146 | cleanup_space(&f); |
3b6121f6 | 147 | strbuf_trim(&f); |
0d4ede9f PB |
148 | if (f.buf[0] == '(' && f.len && f.buf[f.len - 1] == ')') { |
149 | strbuf_remove(&f, 0, 1); | |
3b6121f6 | 150 | strbuf_setlen(&f, f.len - 1); |
0d4ede9f | 151 | } |
2744b234 | 152 | |
c69f2395 | 153 | get_sane_name(&mi->name, &f, &mi->email); |
3b6121f6 | 154 | strbuf_release(&f); |
2744b234 LT |
155 | } |
156 | ||
3b6121f6 | 157 | static void handle_header(struct strbuf **out, const struct strbuf *line) |
62c1f6b4 | 158 | { |
3b6121f6 LS |
159 | if (!*out) { |
160 | *out = xmalloc(sizeof(struct strbuf)); | |
161 | strbuf_init(*out, line->len); | |
162 | } else | |
163 | strbuf_reset(*out); | |
87ab7992 | 164 | |
3b6121f6 | 165 | strbuf_addbuf(*out, line); |
d4a9ce78 JH |
166 | } |
167 | ||
168 | /* NOTE NOTE NOTE. We do not claim we do full MIME. We just attempt | |
169 | * to have enough heuristics to grok MIME encoded patches often found | |
170 | * on our mailing lists. For example, we do not even treat header lines | |
171 | * case insensitively. | |
172 | */ | |
173 | ||
3b6121f6 | 174 | static int slurp_attr(const char *line, const char *name, struct strbuf *attr) |
d4a9ce78 | 175 | { |
554fe20d | 176 | const char *ends, *ap = strcasestr(line, name); |
d4a9ce78 JH |
177 | size_t sz; |
178 | ||
176943b9 JH |
179 | strbuf_setlen(attr, 0); |
180 | if (!ap) | |
d4a9ce78 | 181 | return 0; |
d4a9ce78 JH |
182 | ap += strlen(name); |
183 | if (*ap == '"') { | |
184 | ap++; | |
185 | ends = "\""; | |
186 | } | |
187 | else | |
188 | ends = "; \t"; | |
189 | sz = strcspn(ap, ends); | |
3b6121f6 | 190 | strbuf_add(attr, ap, sz); |
d4a9ce78 JH |
191 | return 1; |
192 | } | |
193 | ||
f1e037b9 | 194 | static void handle_content_type(struct mailinfo *mi, struct strbuf *line) |
d4a9ce78 | 195 | { |
3b6121f6 LS |
196 | struct strbuf *boundary = xmalloc(sizeof(struct strbuf)); |
197 | strbuf_init(boundary, line->len); | |
87ab7992 | 198 | |
3b6121f6 LS |
199 | if (slurp_attr(line->buf, "boundary=", boundary)) { |
200 | strbuf_insert(boundary, 0, "--", 2); | |
8e919277 | 201 | if (++mi->content_top >= &mi->content[MAX_BOUNDARIES]) { |
87ab7992 DZ |
202 | fprintf(stderr, "Too many boundaries to handle\n"); |
203 | exit(1); | |
204 | } | |
8e919277 | 205 | *(mi->content_top) = boundary; |
3b6121f6 | 206 | boundary = NULL; |
b893f091 | 207 | } |
f1e037b9 | 208 | slurp_attr(line->buf, "charset=", &mi->charset); |
3b6121f6 LS |
209 | |
210 | if (boundary) { | |
211 | strbuf_release(boundary); | |
212 | free(boundary); | |
d4a9ce78 | 213 | } |
d4a9ce78 JH |
214 | } |
215 | ||
6200b751 | 216 | static void handle_message_id(struct mailinfo *mi, const struct strbuf *line) |
452dfbed | 217 | { |
6200b751 JH |
218 | if (mi->add_message_id) |
219 | mi->message_id = strdup(line->buf); | |
452dfbed PB |
220 | } |
221 | ||
ab50e38b JH |
222 | static void handle_content_transfer_encoding(struct mailinfo *mi, |
223 | const struct strbuf *line) | |
d4a9ce78 | 224 | { |
3b6121f6 | 225 | if (strcasestr(line->buf, "base64")) |
ab50e38b | 226 | mi->transfer_encoding = TE_BASE64; |
3b6121f6 | 227 | else if (strcasestr(line->buf, "quoted-printable")) |
ab50e38b | 228 | mi->transfer_encoding = TE_QP; |
d4a9ce78 | 229 | else |
ab50e38b | 230 | mi->transfer_encoding = TE_DONTCARE; |
2744b234 LT |
231 | } |
232 | ||
8e919277 | 233 | static int is_multipart_boundary(struct mailinfo *mi, const struct strbuf *line) |
d4a9ce78 | 234 | { |
8e919277 JH |
235 | struct strbuf *content_top = *(mi->content_top); |
236 | ||
237 | return ((content_top->len <= line->len) && | |
238 | !memcmp(line->buf, content_top->buf, content_top->len)); | |
d4a9ce78 JH |
239 | } |
240 | ||
849106d5 | 241 | static void cleanup_subject(struct mailinfo *mi, struct strbuf *subject) |
2744b234 | 242 | { |
17635fc9 JH |
243 | size_t at = 0; |
244 | ||
245 | while (at < subject->len) { | |
246 | char *pos; | |
247 | size_t remove; | |
248 | ||
249 | switch (subject->buf[at]) { | |
2744b234 | 250 | case 'r': case 'R': |
17635fc9 | 251 | if (subject->len <= at + 3) |
3b6121f6 | 252 | break; |
d5b4d80d JH |
253 | if ((subject->buf[at + 1] == 'e' || |
254 | subject->buf[at + 1] == 'E') && | |
255 | subject->buf[at + 2] == ':') { | |
17635fc9 | 256 | strbuf_remove(subject, at, 3); |
2744b234 LT |
257 | continue; |
258 | } | |
17635fc9 | 259 | at++; |
2744b234 LT |
260 | break; |
261 | case ' ': case '\t': case ':': | |
17635fc9 | 262 | strbuf_remove(subject, at, 1); |
2744b234 | 263 | continue; |
2744b234 | 264 | case '[': |
17635fc9 JH |
265 | pos = strchr(subject->buf + at, ']'); |
266 | if (!pos) | |
267 | break; | |
268 | remove = pos - subject->buf + at + 1; | |
849106d5 | 269 | if (!mi->keep_non_patch_brackets_in_subject || |
17635fc9 JH |
270 | (7 <= remove && |
271 | memmem(subject->buf + at, remove, "PATCH", 5))) | |
272 | strbuf_remove(subject, at, remove); | |
ee2d1cb4 | 273 | else { |
17635fc9 | 274 | at += remove; |
ee2d1cb4 TR |
275 | /* |
276 | * If the input had a space after the ], keep | |
277 | * it. We don't bother with finding the end of | |
278 | * the space, since we later normalize it | |
279 | * anyway. | |
280 | */ | |
281 | if (isspace(subject->buf[at])) | |
282 | at += 1; | |
283 | } | |
17635fc9 | 284 | continue; |
2744b234 | 285 | } |
17635fc9 | 286 | break; |
2744b234 | 287 | } |
17635fc9 | 288 | strbuf_trim(subject); |
34488e3c | 289 | } |
2744b234 | 290 | |
c1b40bd7 | 291 | #define MAX_HDR_PARSED 10 |
538dfe73 | 292 | static const char *header[MAX_HDR_PARSED] = { |
87ab7992 | 293 | "From","Subject","Date", |
d4a9ce78 JH |
294 | }; |
295 | ||
3b6121f6 | 296 | static inline int cmp_header(const struct strbuf *line, const char *hdr) |
d4a9ce78 | 297 | { |
3b6121f6 LS |
298 | int len = strlen(hdr); |
299 | return !strncasecmp(line->buf, hdr, len) && line->len > len && | |
300 | line->buf[len] == ':' && isspace(line->buf[len + 1]); | |
301 | } | |
d4a9ce78 | 302 | |
2da1f366 JK |
303 | static int is_format_patch_separator(const char *line, int len) |
304 | { | |
85de86a1 ES |
305 | static const char SAMPLE[] = |
306 | "From e6807f3efca28b30decfecb1732a56c7db1137ee Mon Sep 17 00:00:00 2001\n"; | |
2da1f366 JK |
307 | const char *cp; |
308 | ||
309 | if (len != strlen(SAMPLE)) | |
310 | return 0; | |
311 | if (!skip_prefix(line, "From ", &cp)) | |
312 | return 0; | |
313 | if (strspn(cp, "0123456789abcdef") != 40) | |
314 | return 0; | |
315 | cp += 40; | |
316 | return !memcmp(SAMPLE + (cp - line), cp, strlen(SAMPLE) - (cp - line)); | |
317 | } | |
318 | ||
3b6121f6 | 319 | static struct strbuf *decode_q_segment(const struct strbuf *q_seg, int rfc2047) |
d4a9ce78 | 320 | { |
3b6121f6 | 321 | const char *in = q_seg->buf; |
d4a9ce78 | 322 | int c; |
3b6121f6 LS |
323 | struct strbuf *out = xmalloc(sizeof(struct strbuf)); |
324 | strbuf_init(out, q_seg->len); | |
325 | ||
326 | while ((c = *in++) != 0) { | |
d4a9ce78 JH |
327 | if (c == '=') { |
328 | int d = *in++; | |
329 | if (d == '\n' || !d) | |
330 | break; /* drop trailing newline */ | |
3b6121f6 | 331 | strbuf_addch(out, (hexval(d) << 4) | hexval(*in++)); |
75731930 | 332 | continue; |
d4a9ce78 | 333 | } |
75731930 JH |
334 | if (rfc2047 && c == '_') /* rfc2047 4.2 (2) */ |
335 | c = 0x20; | |
3b6121f6 | 336 | strbuf_addch(out, c); |
d4a9ce78 | 337 | } |
3b6121f6 | 338 | return out; |
d4a9ce78 JH |
339 | } |
340 | ||
3b6121f6 | 341 | static struct strbuf *decode_b_segment(const struct strbuf *b_seg) |
d4a9ce78 JH |
342 | { |
343 | /* Decode in..ep, possibly in-place to ot */ | |
344 | int c, pos = 0, acc = 0; | |
3b6121f6 LS |
345 | const char *in = b_seg->buf; |
346 | struct strbuf *out = xmalloc(sizeof(struct strbuf)); | |
347 | strbuf_init(out, b_seg->len); | |
d4a9ce78 | 348 | |
3b6121f6 | 349 | while ((c = *in++) != 0) { |
d4a9ce78 JH |
350 | if (c == '+') |
351 | c = 62; | |
352 | else if (c == '/') | |
353 | c = 63; | |
354 | else if ('A' <= c && c <= 'Z') | |
355 | c -= 'A'; | |
356 | else if ('a' <= c && c <= 'z') | |
357 | c -= 'a' - 26; | |
358 | else if ('0' <= c && c <= '9') | |
359 | c -= '0' - 52; | |
d4a9ce78 JH |
360 | else |
361 | continue; /* garbage */ | |
362 | switch (pos++) { | |
363 | case 0: | |
364 | acc = (c << 2); | |
365 | break; | |
366 | case 1: | |
3b6121f6 | 367 | strbuf_addch(out, (acc | (c >> 4))); |
d4a9ce78 JH |
368 | acc = (c & 15) << 4; |
369 | break; | |
370 | case 2: | |
3b6121f6 | 371 | strbuf_addch(out, (acc | (c >> 2))); |
d4a9ce78 JH |
372 | acc = (c & 3) << 6; |
373 | break; | |
374 | case 3: | |
3b6121f6 | 375 | strbuf_addch(out, (acc | c)); |
d4a9ce78 JH |
376 | acc = pos = 0; |
377 | break; | |
378 | } | |
379 | } | |
3b6121f6 | 380 | return out; |
d4a9ce78 JH |
381 | } |
382 | ||
28be2d08 JH |
383 | static void convert_to_utf8(struct mailinfo *mi, |
384 | struct strbuf *line, const char *charset) | |
d4a9ce78 | 385 | { |
b59d398b LT |
386 | char *out; |
387 | ||
28c6bfe9 | 388 | if (!mi->metainfo_charset || !charset || !*charset) |
08a94a14 | 389 | return; |
b45974a6 | 390 | |
28be2d08 | 391 | if (same_encoding(mi->metainfo_charset, charset)) |
7296096c | 392 | return; |
28be2d08 | 393 | out = reencode_string(line->buf, mi->metainfo_charset, charset); |
bb1091a4 | 394 | if (!out) |
d7530708 | 395 | die("cannot convert from %s to %s", |
28be2d08 | 396 | charset, mi->metainfo_charset); |
3b6121f6 | 397 | strbuf_attach(line, out, strlen(out), strlen(out)); |
d4a9ce78 JH |
398 | } |
399 | ||
28be2d08 | 400 | static void decode_header(struct mailinfo *mi, struct strbuf *it) |
d4a9ce78 | 401 | { |
3b6121f6 LS |
402 | char *in, *ep, *cp; |
403 | struct strbuf outbuf = STRBUF_INIT, *dec; | |
404 | struct strbuf charset_q = STRBUF_INIT, piecebuf = STRBUF_INIT; | |
d4a9ce78 | 405 | |
3b6121f6 LS |
406 | in = it->buf; |
407 | while (in - it->buf <= it->len && (ep = strstr(in, "=?")) != NULL) { | |
408 | int encoding; | |
409 | strbuf_reset(&charset_q); | |
410 | strbuf_reset(&piecebuf); | |
b75bf2c3 | 411 | |
d4a9ce78 | 412 | if (in != ep) { |
353aaf2f KS |
413 | /* |
414 | * We are about to process an encoded-word | |
415 | * that begins at ep, but there is something | |
416 | * before the encoded word. | |
417 | */ | |
418 | char *scan; | |
419 | for (scan = in; scan < ep; scan++) | |
420 | if (!isspace(*scan)) | |
421 | break; | |
422 | ||
423 | if (scan != ep || in == it->buf) { | |
424 | /* | |
425 | * We should not lose that "something", | |
426 | * unless we have just processed an | |
427 | * encoded-word, and there is only LWS | |
428 | * before the one we are about to process. | |
429 | */ | |
430 | strbuf_add(&outbuf, in, ep - in); | |
431 | } | |
d4a9ce78 JH |
432 | } |
433 | /* E.g. | |
434 | * ep : "=?iso-2022-jp?B?GyR...?= foo" | |
435 | * ep : "=?ISO-8859-1?Q?Foo=FCbar?= baz" | |
436 | */ | |
437 | ep += 2; | |
3b6121f6 LS |
438 | |
439 | if (ep - it->buf >= it->len || !(cp = strchr(ep, '?'))) | |
3a8fcdaf | 440 | goto release_return; |
3b6121f6 LS |
441 | |
442 | if (cp + 3 - it->buf > it->len) | |
3a8fcdaf | 443 | goto release_return; |
3b6121f6 | 444 | strbuf_add(&charset_q, ep, cp - ep); |
3b6121f6 | 445 | |
d4a9ce78 JH |
446 | encoding = cp[1]; |
447 | if (!encoding || cp[2] != '?') | |
3a8fcdaf | 448 | goto release_return; |
d4a9ce78 JH |
449 | ep = strstr(cp + 3, "?="); |
450 | if (!ep) | |
3a8fcdaf | 451 | goto release_return; |
3b6121f6 | 452 | strbuf_add(&piecebuf, cp + 3, ep - cp - 3); |
d4a9ce78 JH |
453 | switch (tolower(encoding)) { |
454 | default: | |
3a8fcdaf | 455 | goto release_return; |
d4a9ce78 | 456 | case 'b': |
3b6121f6 | 457 | dec = decode_b_segment(&piecebuf); |
d4a9ce78 JH |
458 | break; |
459 | case 'q': | |
3b6121f6 | 460 | dec = decode_q_segment(&piecebuf, 1); |
d4a9ce78 JH |
461 | break; |
462 | } | |
28c6bfe9 | 463 | convert_to_utf8(mi, dec, charset_q.buf); |
8dabdfcc | 464 | |
3b6121f6 LS |
465 | strbuf_addbuf(&outbuf, dec); |
466 | strbuf_release(dec); | |
467 | free(dec); | |
d4a9ce78 JH |
468 | in = ep + 2; |
469 | } | |
3b6121f6 LS |
470 | strbuf_addstr(&outbuf, in); |
471 | strbuf_reset(it); | |
472 | strbuf_addbuf(it, &outbuf); | |
3a8fcdaf | 473 | release_return: |
3b6121f6 LS |
474 | strbuf_release(&outbuf); |
475 | strbuf_release(&charset_q); | |
476 | strbuf_release(&piecebuf); | |
d4a9ce78 JH |
477 | } |
478 | ||
6200b751 JH |
479 | static int check_header(struct mailinfo *mi, |
480 | const struct strbuf *line, | |
481 | struct strbuf *hdr_data[], int overwrite) | |
4f0f9d46 JH |
482 | { |
483 | int i, ret = 0, len; | |
484 | struct strbuf sb = STRBUF_INIT; | |
6200b751 | 485 | |
4f0f9d46 JH |
486 | /* search for the interesting parts */ |
487 | for (i = 0; header[i]; i++) { | |
488 | int len = strlen(header[i]); | |
489 | if ((!hdr_data[i] || overwrite) && cmp_header(line, header[i])) { | |
490 | /* Unwrap inline B and Q encoding, and optionally | |
491 | * normalize the meta information to utf8. | |
492 | */ | |
493 | strbuf_add(&sb, line->buf + len + 2, line->len - len - 2); | |
28be2d08 | 494 | decode_header(mi, &sb); |
4f0f9d46 JH |
495 | handle_header(&hdr_data[i], &sb); |
496 | ret = 1; | |
497 | goto check_header_out; | |
498 | } | |
499 | } | |
500 | ||
501 | /* Content stuff */ | |
502 | if (cmp_header(line, "Content-Type")) { | |
503 | len = strlen("Content-Type: "); | |
504 | strbuf_add(&sb, line->buf + len, line->len - len); | |
28be2d08 | 505 | decode_header(mi, &sb); |
4f0f9d46 | 506 | strbuf_insert(&sb, 0, "Content-Type: ", len); |
f1e037b9 | 507 | handle_content_type(mi, &sb); |
4f0f9d46 JH |
508 | ret = 1; |
509 | goto check_header_out; | |
510 | } | |
511 | if (cmp_header(line, "Content-Transfer-Encoding")) { | |
512 | len = strlen("Content-Transfer-Encoding: "); | |
513 | strbuf_add(&sb, line->buf + len, line->len - len); | |
28be2d08 | 514 | decode_header(mi, &sb); |
ab50e38b | 515 | handle_content_transfer_encoding(mi, &sb); |
4f0f9d46 JH |
516 | ret = 1; |
517 | goto check_header_out; | |
518 | } | |
519 | if (cmp_header(line, "Message-Id")) { | |
520 | len = strlen("Message-Id: "); | |
521 | strbuf_add(&sb, line->buf + len, line->len - len); | |
28be2d08 | 522 | decode_header(mi, &sb); |
6200b751 | 523 | handle_message_id(mi, &sb); |
4f0f9d46 JH |
524 | ret = 1; |
525 | goto check_header_out; | |
526 | } | |
527 | ||
528 | /* for inbody stuff */ | |
529 | if (starts_with(line->buf, ">From") && isspace(line->buf[5])) { | |
530 | ret = is_format_patch_separator(line->buf + 1, line->len - 1); | |
531 | goto check_header_out; | |
532 | } | |
533 | if (starts_with(line->buf, "[PATCH]") && isspace(line->buf[7])) { | |
534 | for (i = 0; header[i]; i++) { | |
535 | if (!strcmp("Subject", header[i])) { | |
536 | handle_header(&hdr_data[i], line); | |
537 | ret = 1; | |
538 | goto check_header_out; | |
539 | } | |
540 | } | |
541 | } | |
542 | ||
543 | check_header_out: | |
544 | strbuf_release(&sb); | |
545 | return ret; | |
546 | } | |
547 | ||
ab50e38b | 548 | static void decode_transfer_encoding(struct mailinfo *mi, struct strbuf *line) |
d4a9ce78 | 549 | { |
3b6121f6 | 550 | struct strbuf *ret; |
d4a9ce78 | 551 | |
ab50e38b | 552 | switch (mi->transfer_encoding) { |
d4a9ce78 | 553 | case TE_QP: |
3b6121f6 LS |
554 | ret = decode_q_segment(line, 0); |
555 | break; | |
d4a9ce78 | 556 | case TE_BASE64: |
3b6121f6 LS |
557 | ret = decode_b_segment(line); |
558 | break; | |
d4a9ce78 | 559 | case TE_DONTCARE: |
9aa23094 | 560 | default: |
3b6121f6 | 561 | return; |
d4a9ce78 | 562 | } |
3b6121f6 LS |
563 | strbuf_reset(line); |
564 | strbuf_addbuf(line, ret); | |
565 | strbuf_release(ret); | |
566 | free(ret); | |
d4a9ce78 JH |
567 | } |
568 | ||
3b6121f6 | 569 | static inline int patchbreak(const struct strbuf *line) |
f0658cf2 | 570 | { |
3b6121f6 LS |
571 | size_t i; |
572 | ||
f0658cf2 | 573 | /* Beginning of a "diff -" header? */ |
59556548 | 574 | if (starts_with(line->buf, "diff -")) |
f0658cf2 DZ |
575 | return 1; |
576 | ||
577 | /* CVS "Index: " line? */ | |
59556548 | 578 | if (starts_with(line->buf, "Index: ")) |
f0658cf2 DZ |
579 | return 1; |
580 | ||
581 | /* | |
582 | * "--- <filename>" starts patches without headers | |
583 | * "---<sp>*" is a manual separator | |
584 | */ | |
3b6121f6 LS |
585 | if (line->len < 4) |
586 | return 0; | |
587 | ||
59556548 | 588 | if (starts_with(line->buf, "---")) { |
f0658cf2 | 589 | /* space followed by a filename? */ |
3b6121f6 | 590 | if (line->buf[3] == ' ' && !isspace(line->buf[4])) |
f0658cf2 DZ |
591 | return 1; |
592 | /* Just whitespace? */ | |
3b6121f6 LS |
593 | for (i = 3; i < line->len; i++) { |
594 | unsigned char c = line->buf[i]; | |
f0658cf2 DZ |
595 | if (c == '\n') |
596 | return 1; | |
597 | if (!isspace(c)) | |
598 | break; | |
599 | } | |
600 | return 0; | |
601 | } | |
602 | return 0; | |
603 | } | |
604 | ||
200c75f0 JH |
605 | static int is_scissors_line(const struct strbuf *line) |
606 | { | |
607 | size_t i, len = line->len; | |
608 | int scissors = 0, gap = 0; | |
609 | int first_nonblank = -1; | |
610 | int last_nonblank = 0, visible, perforation = 0, in_perforation = 0; | |
611 | const char *buf = line->buf; | |
612 | ||
613 | for (i = 0; i < len; i++) { | |
614 | if (isspace(buf[i])) { | |
615 | if (in_perforation) { | |
616 | perforation++; | |
617 | gap++; | |
618 | } | |
619 | continue; | |
620 | } | |
621 | last_nonblank = i; | |
622 | if (first_nonblank < 0) | |
623 | first_nonblank = i; | |
624 | if (buf[i] == '-') { | |
625 | in_perforation = 1; | |
626 | perforation++; | |
627 | continue; | |
628 | } | |
629 | if (i + 1 < len && | |
9974e290 JN |
630 | (!memcmp(buf + i, ">8", 2) || !memcmp(buf + i, "8<", 2) || |
631 | !memcmp(buf + i, ">%", 2) || !memcmp(buf + i, "%<", 2))) { | |
200c75f0 JH |
632 | in_perforation = 1; |
633 | perforation += 2; | |
634 | scissors += 2; | |
635 | i++; | |
636 | continue; | |
637 | } | |
638 | in_perforation = 0; | |
639 | } | |
640 | ||
641 | /* | |
642 | * The mark must be at least 8 bytes long (e.g. "-- >8 --"). | |
643 | * Even though there can be arbitrary cruft on the same line | |
644 | * (e.g. "cut here"), in order to avoid misidentification, the | |
645 | * perforation must occupy more than a third of the visible | |
646 | * width of the line, and dashes and scissors must occupy more | |
647 | * than half of the perforation. | |
648 | */ | |
649 | ||
650 | visible = last_nonblank - first_nonblank + 1; | |
651 | return (scissors && 8 <= visible && | |
652 | visible < perforation * 3 && | |
653 | gap * 2 < perforation); | |
654 | } | |
655 | ||
13c6df26 | 656 | static int handle_commit_msg(struct mailinfo *mi, struct strbuf *line) |
d4a9ce78 | 657 | { |
4933910a | 658 | assert(!mi->filter_stage); |
2744b234 | 659 | |
13c6df26 | 660 | if (mi->header_stage) { |
470b4526 | 661 | if (!line->len || (line->len == 1 && line->buf[0] == '\n')) |
3b6121f6 | 662 | return 0; |
d25e5159 LS |
663 | } |
664 | ||
ad57ef9d | 665 | if (mi->use_inbody_headers && mi->header_stage) { |
d895bf0f | 666 | mi->header_stage = check_header(mi, line, mi->s_hdr_data, 0); |
13c6df26 | 667 | if (mi->header_stage) |
87ab7992 | 668 | return 0; |
d25e5159 LS |
669 | } else |
670 | /* Only trim the first (blank) line of the commit message | |
671 | * when ignoring in-body headers. | |
672 | */ | |
13c6df26 | 673 | mi->header_stage = 0; |
8b4525fb | 674 | |
86747c13 | 675 | /* normalize the log message to UTF-8. */ |
f1e037b9 | 676 | convert_to_utf8(mi, line, mi->charset.buf); |
86747c13 | 677 | |
ad57ef9d | 678 | if (mi->use_scissors && is_scissors_line(line)) { |
200c75f0 | 679 | int i; |
05e625e5 JH |
680 | |
681 | strbuf_setlen(&mi->log_message, 0); | |
13c6df26 | 682 | mi->header_stage = 1; |
200c75f0 JH |
683 | |
684 | /* | |
685 | * We may have already read "secondary headers"; purge | |
686 | * them to give ourselves a clean restart. | |
687 | */ | |
688 | for (i = 0; header[i]; i++) { | |
d895bf0f JH |
689 | if (mi->s_hdr_data[i]) |
690 | strbuf_release(mi->s_hdr_data[i]); | |
691 | mi->s_hdr_data[i] = NULL; | |
200c75f0 JH |
692 | } |
693 | return 0; | |
694 | } | |
695 | ||
f0658cf2 | 696 | if (patchbreak(line)) { |
6200b751 | 697 | if (mi->message_id) |
05e625e5 JH |
698 | strbuf_addf(&mi->log_message, |
699 | "Message-Id: %s\n", mi->message_id); | |
87ab7992 DZ |
700 | return 1; |
701 | } | |
8b4525fb | 702 | |
05e625e5 | 703 | strbuf_addbuf(&mi->log_message, line); |
d4a9ce78 | 704 | return 0; |
2744b234 LT |
705 | } |
706 | ||
43550efa | 707 | static void handle_patch(struct mailinfo *mi, const struct strbuf *line) |
2744b234 | 708 | { |
8f63588a | 709 | fwrite(line->buf, 1, line->len, mi->patchfile); |
43550efa | 710 | mi->patch_lines++; |
2744b234 LT |
711 | } |
712 | ||
13c6df26 | 713 | static void handle_filter(struct mailinfo *mi, struct strbuf *line) |
2744b234 | 714 | { |
13c6df26 | 715 | switch (mi->filter_stage) { |
87ab7992 | 716 | case 0: |
13c6df26 | 717 | if (!handle_commit_msg(mi, line)) |
d4a9ce78 | 718 | break; |
13c6df26 | 719 | mi->filter_stage++; |
87ab7992 | 720 | case 1: |
43550efa | 721 | handle_patch(mi, line); |
3b6121f6 | 722 | break; |
2744b234 LT |
723 | } |
724 | } | |
725 | ||
9cc243f7 JH |
726 | static int is_rfc2822_header(const struct strbuf *line) |
727 | { | |
728 | /* | |
729 | * The section that defines the loosest possible | |
730 | * field name is "3.6.8 Optional fields". | |
731 | * | |
732 | * optional-field = field-name ":" unstructured CRLF | |
733 | * field-name = 1*ftext | |
734 | * ftext = %d33-57 / %59-126 | |
735 | */ | |
736 | int ch; | |
737 | char *cp = line->buf; | |
738 | ||
739 | /* Count mbox From headers as headers */ | |
740 | if (starts_with(cp, "From ") || starts_with(cp, ">From ")) | |
741 | return 1; | |
742 | ||
743 | while ((ch = *cp++)) { | |
744 | if (ch == ':') | |
745 | return 1; | |
746 | if ((33 <= ch && ch <= 57) || | |
747 | (59 <= ch && ch <= 126)) | |
748 | continue; | |
749 | break; | |
750 | } | |
751 | return 0; | |
752 | } | |
753 | ||
754 | static int read_one_header_line(struct strbuf *line, FILE *in) | |
755 | { | |
756 | struct strbuf continuation = STRBUF_INIT; | |
757 | ||
758 | /* Get the first part of the line. */ | |
759 | if (strbuf_getline(line, in, '\n')) | |
760 | return 0; | |
761 | ||
762 | /* | |
763 | * Is it an empty line or not a valid rfc2822 header? | |
764 | * If so, stop here, and return false ("not a header") | |
765 | */ | |
766 | strbuf_rtrim(line); | |
767 | if (!line->len || !is_rfc2822_header(line)) { | |
768 | /* Re-add the newline */ | |
769 | strbuf_addch(line, '\n'); | |
770 | return 0; | |
771 | } | |
772 | ||
773 | /* | |
774 | * Now we need to eat all the continuation lines.. | |
775 | * Yuck, 2822 header "folding" | |
776 | */ | |
777 | for (;;) { | |
778 | int peek; | |
779 | ||
780 | peek = fgetc(in); ungetc(peek, in); | |
781 | if (peek != ' ' && peek != '\t') | |
782 | break; | |
783 | if (strbuf_getline(&continuation, in, '\n')) | |
784 | break; | |
785 | continuation.buf[0] = ' '; | |
786 | strbuf_rtrim(&continuation); | |
787 | strbuf_addbuf(line, &continuation); | |
788 | } | |
789 | strbuf_release(&continuation); | |
790 | ||
791 | return 1; | |
792 | } | |
793 | ||
173aef7c | 794 | static int find_boundary(struct mailinfo *mi, struct strbuf *line) |
39afcd38 | 795 | { |
173aef7c | 796 | while (!strbuf_getline(line, mi->input, '\n')) { |
8e919277 | 797 | if (*(mi->content_top) && is_multipart_boundary(mi, line)) |
39afcd38 JH |
798 | return 1; |
799 | } | |
800 | return 0; | |
801 | } | |
802 | ||
13c6df26 | 803 | static int handle_boundary(struct mailinfo *mi, struct strbuf *line) |
39afcd38 JH |
804 | { |
805 | struct strbuf newline = STRBUF_INIT; | |
806 | ||
807 | strbuf_addch(&newline, '\n'); | |
808 | again: | |
8e919277 JH |
809 | if (line->len >= (*(mi->content_top))->len + 2 && |
810 | !memcmp(line->buf + (*(mi->content_top))->len, "--", 2)) { | |
39afcd38 JH |
811 | /* we hit an end boundary */ |
812 | /* pop the current boundary off the stack */ | |
8e919277 JH |
813 | strbuf_release(*(mi->content_top)); |
814 | free(*(mi->content_top)); | |
815 | *(mi->content_top) = NULL; | |
39afcd38 JH |
816 | |
817 | /* technically won't happen as is_multipart_boundary() | |
818 | will fail first. But just in case.. | |
819 | */ | |
8e919277 | 820 | if (--mi->content_top < mi->content) { |
39afcd38 JH |
821 | fprintf(stderr, "Detected mismatched boundaries, " |
822 | "can't recover\n"); | |
823 | exit(1); | |
824 | } | |
13c6df26 | 825 | handle_filter(mi, &newline); |
39afcd38 JH |
826 | strbuf_release(&newline); |
827 | ||
828 | /* skip to the next boundary */ | |
173aef7c | 829 | if (!find_boundary(mi, line)) |
39afcd38 JH |
830 | return 0; |
831 | goto again; | |
832 | } | |
833 | ||
834 | /* set some defaults */ | |
ab50e38b | 835 | mi->transfer_encoding = TE_DONTCARE; |
f1e037b9 | 836 | strbuf_reset(&mi->charset); |
39afcd38 JH |
837 | |
838 | /* slurp in this section's info */ | |
173aef7c | 839 | while (read_one_header_line(line, mi->input)) |
d895bf0f | 840 | check_header(mi, line, mi->p_hdr_data, 0); |
39afcd38 JH |
841 | |
842 | strbuf_release(&newline); | |
843 | /* replenish line */ | |
173aef7c | 844 | if (strbuf_getline(line, mi->input, '\n')) |
39afcd38 | 845 | return 0; |
69e24def | 846 | strbuf_addch(line, '\n'); |
39afcd38 JH |
847 | return 1; |
848 | } | |
849 | ||
173aef7c | 850 | static void handle_body(struct mailinfo *mi, struct strbuf *line) |
1d8fa411 | 851 | { |
3b6121f6 | 852 | struct strbuf prev = STRBUF_INIT; |
d4a9ce78 JH |
853 | |
854 | /* Skip up to the first boundary */ | |
8e919277 | 855 | if (*(mi->content_top)) { |
173aef7c | 856 | if (!find_boundary(mi, line)) |
3b6121f6 | 857 | goto handle_body_out; |
87ab7992 DZ |
858 | } |
859 | ||
860 | do { | |
861 | /* process any boundary lines */ | |
8e919277 | 862 | if (*(mi->content_top) && is_multipart_boundary(mi, line)) { |
87ab7992 | 863 | /* flush any leftover */ |
a9fd1383 | 864 | if (prev.len) { |
13c6df26 | 865 | handle_filter(mi, &prev); |
a9fd1383 JH |
866 | strbuf_reset(&prev); |
867 | } | |
13c6df26 | 868 | if (!handle_boundary(mi, line)) |
3b6121f6 | 869 | goto handle_body_out; |
87ab7992 DZ |
870 | } |
871 | ||
86747c13 | 872 | /* Unwrap transfer encoding */ |
ab50e38b | 873 | decode_transfer_encoding(mi, line); |
87ab7992 | 874 | |
ab50e38b | 875 | switch (mi->transfer_encoding) { |
87ab7992 | 876 | case TE_BASE64: |
87f1b884 | 877 | case TE_QP: |
87ab7992 | 878 | { |
3b6121f6 LS |
879 | struct strbuf **lines, **it, *sb; |
880 | ||
881 | /* Prepend any previous partial lines */ | |
fde00d50 | 882 | strbuf_insert(line, 0, prev.buf, prev.len); |
3b6121f6 | 883 | strbuf_reset(&prev); |
87ab7992 | 884 | |
9aa23094 JH |
885 | /* |
886 | * This is a decoded line that may contain | |
87ab7992 DZ |
887 | * multiple new lines. Pass only one chunk |
888 | * at a time to handle_filter() | |
889 | */ | |
fde00d50 | 890 | lines = strbuf_split(line, '\n'); |
3b6121f6 LS |
891 | for (it = lines; (sb = *it); it++) { |
892 | if (*(it + 1) == NULL) /* The last line */ | |
893 | if (sb->buf[sb->len - 1] != '\n') { | |
894 | /* Partial line, save it for later. */ | |
895 | strbuf_addbuf(&prev, sb); | |
896 | break; | |
897 | } | |
13c6df26 | 898 | handle_filter(mi, sb); |
3b6121f6 | 899 | } |
9aa23094 | 900 | /* |
3b6121f6 | 901 | * The partial chunk is saved in "prev" and will be |
9aa23094 | 902 | * appended by the next iteration of read_line_with_nul(). |
87ab7992 | 903 | */ |
3b6121f6 | 904 | strbuf_list_free(lines); |
d4a9ce78 | 905 | break; |
1d8fa411 | 906 | } |
87ab7992 | 907 | default: |
13c6df26 | 908 | handle_filter(mi, line); |
d4a9ce78 | 909 | } |
87ab7992 | 910 | |
173aef7c | 911 | } while (!strbuf_getwholeline(line, mi->input, '\n')); |
3b6121f6 LS |
912 | |
913 | handle_body_out: | |
914 | strbuf_release(&prev); | |
1d8fa411 JH |
915 | } |
916 | ||
3b6121f6 | 917 | static void output_header_lines(FILE *fout, const char *hdr, const struct strbuf *data) |
d7f6bae2 | 918 | { |
3b6121f6 | 919 | const char *sp = data->buf; |
d7f6bae2 | 920 | while (1) { |
3b6121f6 | 921 | char *ep = strchr(sp, '\n'); |
d7f6bae2 JH |
922 | int len; |
923 | if (!ep) | |
3b6121f6 | 924 | len = strlen(sp); |
d7f6bae2 | 925 | else |
3b6121f6 LS |
926 | len = ep - sp; |
927 | fprintf(fout, "%s: %.*s\n", hdr, len, sp); | |
d7f6bae2 JH |
928 | if (!ep) |
929 | break; | |
3b6121f6 | 930 | sp = ep + 1; |
d7f6bae2 JH |
931 | } |
932 | } | |
933 | ||
c69f2395 | 934 | static void handle_info(struct mailinfo *mi) |
2744b234 | 935 | { |
3b6121f6 | 936 | struct strbuf *hdr; |
87ab7992 DZ |
937 | int i; |
938 | ||
939 | for (i = 0; header[i]; i++) { | |
87ab7992 | 940 | /* only print inbody headers if we output a patch file */ |
d895bf0f JH |
941 | if (mi->patch_lines && mi->s_hdr_data[i]) |
942 | hdr = mi->s_hdr_data[i]; | |
943 | else if (mi->p_hdr_data[i]) | |
944 | hdr = mi->p_hdr_data[i]; | |
87ab7992 DZ |
945 | else |
946 | continue; | |
947 | ||
b1a013dd | 948 | if (!strcmp(header[i], "Subject")) { |
849106d5 JH |
949 | if (!mi->keep_subject) { |
950 | cleanup_subject(mi, hdr); | |
3b6121f6 | 951 | cleanup_space(hdr); |
d7f6bae2 | 952 | } |
173aef7c | 953 | output_header_lines(mi->output, "Subject", hdr); |
b1a013dd | 954 | } else if (!strcmp(header[i], "From")) { |
ddfb3696 | 955 | cleanup_space(hdr); |
c69f2395 | 956 | handle_from(mi, hdr); |
173aef7c JH |
957 | fprintf(mi->output, "Author: %s\n", mi->name.buf); |
958 | fprintf(mi->output, "Email: %s\n", mi->email.buf); | |
87ab7992 DZ |
959 | } else { |
960 | cleanup_space(hdr); | |
173aef7c | 961 | fprintf(mi->output, "%s: %s\n", header[i], hdr->buf); |
87ab7992 | 962 | } |
d4a9ce78 | 963 | } |
173aef7c | 964 | fprintf(mi->output, "\n"); |
2744b234 LT |
965 | } |
966 | ||
173aef7c | 967 | static int mailinfo(struct mailinfo *mi, const char *msg, const char *patch) |
34488e3c | 968 | { |
05e625e5 | 969 | FILE *cmitmsg; |
f88a545a | 970 | int peek; |
6e21b508 JH |
971 | struct strbuf line = STRBUF_INIT; |
972 | ||
05e625e5 JH |
973 | cmitmsg = fopen(msg, "w"); |
974 | if (!cmitmsg) { | |
34488e3c LS |
975 | perror(msg); |
976 | return -1; | |
977 | } | |
8f63588a JH |
978 | mi->patchfile = fopen(patch, "w"); |
979 | if (!mi->patchfile) { | |
34488e3c | 980 | perror(patch); |
05e625e5 | 981 | fclose(cmitmsg); |
34488e3c LS |
982 | return -1; |
983 | } | |
87ab7992 | 984 | |
d895bf0f JH |
985 | mi->p_hdr_data = xcalloc(MAX_HDR_PARSED, sizeof(*(mi->p_hdr_data))); |
986 | mi->s_hdr_data = xcalloc(MAX_HDR_PARSED, sizeof(*(mi->s_hdr_data))); | |
87ab7992 | 987 | |
f88a545a | 988 | do { |
173aef7c | 989 | peek = fgetc(mi->input); |
f88a545a | 990 | } while (isspace(peek)); |
173aef7c | 991 | ungetc(peek, mi->input); |
f88a545a | 992 | |
87ab7992 | 993 | /* process the email header */ |
173aef7c | 994 | while (read_one_header_line(&line, mi->input)) |
d895bf0f | 995 | check_header(mi, &line, mi->p_hdr_data, 1); |
87ab7992 | 996 | |
173aef7c | 997 | handle_body(mi, &line); |
05e625e5 JH |
998 | fwrite(mi->log_message.buf, 1, mi->log_message.len, cmitmsg); |
999 | fclose(cmitmsg); | |
8f63588a | 1000 | fclose(mi->patchfile); |
e38ee06e | 1001 | |
c69f2395 | 1002 | handle_info(mi); |
6e21b508 | 1003 | strbuf_release(&line); |
34488e3c LS |
1004 | return 0; |
1005 | } | |
1006 | ||
ad57ef9d | 1007 | static int git_mailinfo_config(const char *var, const char *value, void *mi_) |
43485d3d | 1008 | { |
ad57ef9d JH |
1009 | struct mailinfo *mi = mi_; |
1010 | ||
59556548 | 1011 | if (!starts_with(var, "mailinfo.")) |
ad57ef9d | 1012 | return git_default_config(var, value, NULL); |
43485d3d | 1013 | if (!strcmp(var, "mailinfo.scissors")) { |
ad57ef9d | 1014 | mi->use_scissors = git_config_bool(var, value); |
43485d3d JH |
1015 | return 0; |
1016 | } | |
1017 | /* perhaps others here */ | |
1018 | return 0; | |
1019 | } | |
1020 | ||
c69f2395 JH |
1021 | static void setup_mailinfo(struct mailinfo *mi) |
1022 | { | |
1023 | memset(mi, 0, sizeof(*mi)); | |
1024 | strbuf_init(&mi->name, 0); | |
1025 | strbuf_init(&mi->email, 0); | |
f1e037b9 | 1026 | strbuf_init(&mi->charset, 0); |
05e625e5 | 1027 | strbuf_init(&mi->log_message, 0); |
13c6df26 | 1028 | mi->header_stage = 1; |
ad57ef9d | 1029 | mi->use_inbody_headers = 1; |
8e919277 | 1030 | mi->content_top = mi->content; |
c69f2395 JH |
1031 | git_config(git_mailinfo_config, &mi); |
1032 | } | |
1033 | ||
1034 | static void clear_mailinfo(struct mailinfo *mi) | |
1035 | { | |
d895bf0f JH |
1036 | int i; |
1037 | ||
c69f2395 JH |
1038 | strbuf_release(&mi->name); |
1039 | strbuf_release(&mi->email); | |
f1e037b9 | 1040 | strbuf_release(&mi->charset); |
6200b751 | 1041 | free(mi->message_id); |
d895bf0f JH |
1042 | |
1043 | for (i = 0; mi->p_hdr_data[i]; i++) | |
1044 | strbuf_release(mi->p_hdr_data[i]); | |
1045 | free(mi->p_hdr_data); | |
1046 | for (i = 0; mi->s_hdr_data[i]; i++) | |
1047 | strbuf_release(mi->s_hdr_data[i]); | |
1048 | free(mi->s_hdr_data); | |
8e919277 JH |
1049 | |
1050 | while (mi->content < mi->content_top) { | |
1051 | free(*(mi->content_top)); | |
1052 | mi->content_top--; | |
1053 | } | |
05e625e5 JH |
1054 | |
1055 | strbuf_release(&mi->log_message); | |
c69f2395 JH |
1056 | } |
1057 | ||
6bff6a60 | 1058 | static const char mailinfo_usage[] = |
9c9b4f2f | 1059 | "git mailinfo [-k | -b] [-m | --message-id] [-u | --encoding=<encoding> | -n] [--scissors | --no-scissors] <msg> <patch> < mail >info"; |
d4a9ce78 | 1060 | |
a633fca0 | 1061 | int cmd_mailinfo(int argc, const char **argv, const char *prefix) |
2744b234 | 1062 | { |
bb1091a4 | 1063 | const char *def_charset; |
c69f2395 JH |
1064 | struct mailinfo mi; |
1065 | int status; | |
bb1091a4 | 1066 | |
f1f909e3 JH |
1067 | /* NEEDSWORK: might want to do the optional .git/ directory |
1068 | * discovery | |
1069 | */ | |
c69f2395 | 1070 | setup_mailinfo(&mi); |
f1f909e3 | 1071 | |
a6fa5992 | 1072 | def_charset = get_commit_output_encoding(); |
28be2d08 | 1073 | mi.metainfo_charset = def_charset; |
bb1091a4 | 1074 | |
6bff6a60 JH |
1075 | while (1 < argc && argv[1][0] == '-') { |
1076 | if (!strcmp(argv[1], "-k")) | |
849106d5 | 1077 | mi.keep_subject = 1; |
17635fc9 | 1078 | else if (!strcmp(argv[1], "-b")) |
849106d5 | 1079 | mi.keep_non_patch_brackets_in_subject = 1; |
452dfbed | 1080 | else if (!strcmp(argv[1], "-m") || !strcmp(argv[1], "--message-id")) |
6200b751 | 1081 | mi.add_message_id = 1; |
d4a9ce78 | 1082 | else if (!strcmp(argv[1], "-u")) |
28be2d08 | 1083 | mi.metainfo_charset = def_charset; |
bb1091a4 | 1084 | else if (!strcmp(argv[1], "-n")) |
28be2d08 | 1085 | mi.metainfo_charset = NULL; |
59556548 | 1086 | else if (starts_with(argv[1], "--encoding=")) |
28be2d08 | 1087 | mi.metainfo_charset = argv[1] + 11; |
017678b4 | 1088 | else if (!strcmp(argv[1], "--scissors")) |
ad57ef9d | 1089 | mi.use_scissors = 1; |
017678b4 | 1090 | else if (!strcmp(argv[1], "--no-scissors")) |
ad57ef9d | 1091 | mi.use_scissors = 0; |
d25e5159 | 1092 | else if (!strcmp(argv[1], "--no-inbody-headers")) |
ad57ef9d | 1093 | mi.use_inbody_headers = 0; |
d4a9ce78 | 1094 | else |
f1f909e3 | 1095 | usage(mailinfo_usage); |
6bff6a60 JH |
1096 | argc--; argv++; |
1097 | } | |
1098 | ||
a196d8d4 | 1099 | if (argc != 3) |
f1f909e3 | 1100 | usage(mailinfo_usage); |
34488e3c | 1101 | |
173aef7c JH |
1102 | mi.input = stdin; |
1103 | mi.output = stdout; | |
1104 | status = !!mailinfo(&mi, argv[1], argv[2]); | |
c69f2395 JH |
1105 | clear_mailinfo(&mi); |
1106 | ||
1107 | return status; | |
2744b234 | 1108 | } |