]>
Commit | Line | Data |
---|---|---|
d9bae1a1 | 1 | #include "cache.h" |
b2141fc1 | 2 | #include "config.h" |
be58e70d | 3 | #include "userdiff.h" |
be58e70d JK |
4 | #include "attr.h" |
5 | ||
6 | static struct userdiff_driver *drivers; | |
7 | static int ndrivers; | |
8 | static int drivers_alloc; | |
9 | ||
ae3b970a | 10 | #define PATTERNS(name, pattern, word_regex) \ |
664d44ee JN |
11 | { name, NULL, -1, { pattern, REG_EXTENDED }, \ |
12 | word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" } | |
909a5494 | 13 | #define IPATTERN(name, pattern, word_regex) \ |
664d44ee JN |
14 | { name, NULL, -1, { pattern, REG_EXTENDED | REG_ICASE }, \ |
15 | word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" } | |
be58e70d | 16 | static struct userdiff_driver builtin_drivers[] = { |
e90d065e | 17 | IPATTERN("ada", |
39a87a29 | 18 | "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n" |
e90d065e AJ |
19 | "!^[ \t]*with[ \t].*$\n" |
20 | "^[ \t]*((procedure|function)[ \t]+.*)$\n" | |
21 | "^[ \t]*((package|protected|task)[ \t]+.*)$", | |
22 | /* -- */ | |
23 | "[a-zA-Z][a-zA-Z0-9_]*" | |
39a87a29 | 24 | "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?" |
e90d065e | 25 | "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"), |
909a5494 BC |
26 | IPATTERN("fortran", |
27 | "!^([C*]|[ \t]*!)\n" | |
28 | "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n" | |
29 | "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA" | |
30 | "|([^'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$", | |
31 | /* -- */ | |
32 | "[a-zA-Z][a-zA-Z0-9_]*" | |
33 | "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\." | |
34 | /* numbers and format statements like 2E14.4, or ES12.6, 9X. | |
35 | * Don't worry about format statements without leading digits since | |
36 | * they would have been matched above as a variable anyway. */ | |
37 | "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?" | |
664d44ee | 38 | "|//|\\*\\*|::|[/<>=]="), |
69f9c87d ZB |
39 | IPATTERN("fountain", "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$", |
40 | "[^ \t-]+"), | |
1dbf0c0a AG |
41 | PATTERNS("golang", |
42 | /* Functions */ | |
43 | "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n" | |
44 | /* Structs and interfaces */ | |
45 | "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)", | |
46 | /* -- */ | |
47 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
48 | "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?" | |
49 | "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"), | |
9c03caca | 50 | PATTERNS("html", "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$", |
664d44ee | 51 | "[^<>= \t]+"), |
80c49c3d | 52 | PATTERNS("java", |
be58e70d | 53 | "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n" |
959e2e64 PB |
54 | "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$", |
55 | /* -- */ | |
80c49c3d TR |
56 | "[a-zA-Z_][a-zA-Z0-9_]*" |
57 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
58 | "|[-+*/<>%&^|=!]=" | |
664d44ee | 59 | "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"), |
53b10a14 GH |
60 | PATTERNS("matlab", |
61 | "^[[:space:]]*((classdef|function)[[:space:]].*)$|^%%[[:space:]].*$", | |
62 | "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"), | |
80c49c3d | 63 | PATTERNS("objc", |
be58e70d JK |
64 | /* Negate C statements that can look like functions */ |
65 | "!^[ \t]*(do|for|if|else|return|switch|while)\n" | |
66 | /* Objective-C methods */ | |
67 | "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n" | |
68 | /* C functions */ | |
959e2e64 | 69 | "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n" |
be58e70d | 70 | /* Objective-C class/protocol definitions */ |
80c49c3d TR |
71 | "^(@(implementation|interface|protocol)[ \t].*)$", |
72 | /* -- */ | |
73 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
74 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
664d44ee | 75 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"), |
80c49c3d | 76 | PATTERNS("pascal", |
ad5b6942 | 77 | "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|" |
be58e70d JK |
78 | "implementation|initialization|finalization)[ \t]*.*)$" |
79 | "\n" | |
80c49c3d TR |
80 | "^(.*=[ \t]*(class|record).*)$", |
81 | /* -- */ | |
82 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
83 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+" | |
664d44ee | 84 | "|<>|<=|>=|:=|\\.\\."), |
71a5d4bc | 85 | PATTERNS("perl", |
ea2ca449 JN |
86 | "^package .*\n" |
87 | "^sub [[:alnum:]_':]+[ \t]*" | |
88 | "(\\([^)]*\\)[ \t]*)?" /* prototype */ | |
89 | /* | |
90 | * Attributes. A regex can't count nested parentheses, | |
91 | * so just slurp up whatever we see, taking care not | |
92 | * to accept lines like "sub foo; # defined elsewhere". | |
93 | * | |
94 | * An attribute could contain a semicolon, but at that | |
95 | * point it seems reasonable enough to give up. | |
96 | */ | |
97 | "(:[^;#]*)?" | |
98 | "(\\{[ \t]*)?" /* brace can come here or on the next line */ | |
99 | "(#.*)?$\n" /* comment */ | |
f143d9c6 | 100 | "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*" |
ea2ca449 JN |
101 | "(\\{[ \t]*)?" /* brace can come here or on the next line */ |
102 | "(#.*)?$\n" | |
12f0967a | 103 | "^=head[0-9] .*", /* POD */ |
71a5d4bc JN |
104 | /* -- */ |
105 | "[[:alpha:]_'][[:alnum:]_']*" | |
106 | "|0[xb]?[0-9a-fA-F_]*" | |
107 | /* taking care not to interpret 3..5 as (3.)(.5) */ | |
108 | "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?" | |
109 | "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::" | |
110 | "|&&=|\\|\\|=|//=|\\*\\*=" | |
111 | "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?" | |
112 | "|[-+*/%.^&<>=!|]=" | |
113 | "|=~|!~" | |
664d44ee | 114 | "|<<|<>|<=>|>>"), |
6d2f208c BS |
115 | PATTERNS("php", |
116 | "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n" | |
1ab63164 | 117 | "^[\t ]*((((final|abstract)[\t ]+)?class|interface|trait).*)$", |
80c49c3d TR |
118 | /* -- */ |
119 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
120 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+" | |
664d44ee | 121 | "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"), |
80c49c3d TR |
122 | PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$", |
123 | /* -- */ | |
124 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
125 | "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?" | |
664d44ee | 126 | "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"), |
80c49c3d TR |
127 | /* -- */ |
128 | PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$", | |
129 | /* -- */ | |
130 | "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*" | |
131 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?." | |
664d44ee | 132 | "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"), |
80c49c3d TR |
133 | PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$", |
134 | "[={}\"]|[^={}\" \t]+"), | |
135 | PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$", | |
664d44ee | 136 | "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"), |
80c49c3d TR |
137 | PATTERNS("cpp", |
138 | /* Jump targets or access declarations */ | |
8a2e8da3 JS |
139 | "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n" |
140 | /* functions/methods, variables, and compounds at top level */ | |
141 | "^((::[[:space:]]*)?[A-Za-z_].*)$", | |
80c49c3d TR |
142 | /* -- */ |
143 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
abf8f986 | 144 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lLuU]*" |
407e07f2 | 145 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*"), |
b221207d PO |
146 | PATTERNS("csharp", |
147 | /* Keywords */ | |
148 | "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n" | |
149 | /* Methods and constructors */ | |
a12cec99 | 150 | "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n" |
b221207d PO |
151 | /* Properties */ |
152 | "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n" | |
153 | /* Type definitions */ | |
154 | "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n" | |
155 | /* Namespace */ | |
156 | "^[ \t]*(namespace[ \t]+.*)$", | |
157 | /* -- */ | |
158 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
159 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
664d44ee | 160 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"), |
0719f3ee WD |
161 | IPATTERN("css", |
162 | "![:;][[:space:]]*$\n" | |
163 | "^[_a-z0-9].*$", | |
164 | /* -- */ | |
165 | /* | |
166 | * This regex comes from W3C CSS specs. Should theoretically also | |
167 | * allow ISO 10646 characters U+00A0 and higher, | |
168 | * but they are not handled in this regex. | |
169 | */ | |
170 | "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */ | |
171 | "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */ | |
172 | ), | |
122aa6f9 | 173 | { "default", NULL, -1, { NULL, 0 } }, |
be58e70d | 174 | }; |
80c49c3d | 175 | #undef PATTERNS |
909a5494 | 176 | #undef IPATTERN |
be58e70d JK |
177 | |
178 | static struct userdiff_driver driver_true = { | |
179 | "diff=true", | |
180 | NULL, | |
122aa6f9 | 181 | 0, |
be58e70d JK |
182 | { NULL, 0 } |
183 | }; | |
be58e70d JK |
184 | |
185 | static struct userdiff_driver driver_false = { | |
186 | "!diff", | |
187 | NULL, | |
122aa6f9 | 188 | 1, |
be58e70d JK |
189 | { NULL, 0 } |
190 | }; | |
be58e70d JK |
191 | |
192 | static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len) | |
193 | { | |
194 | int i; | |
195 | for (i = 0; i < ndrivers; i++) { | |
196 | struct userdiff_driver *drv = drivers + i; | |
197 | if (!strncmp(drv->name, k, len) && !drv->name[len]) | |
198 | return drv; | |
199 | } | |
200 | for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) { | |
201 | struct userdiff_driver *drv = builtin_drivers + i; | |
202 | if (!strncmp(drv->name, k, len) && !drv->name[len]) | |
203 | return drv; | |
204 | } | |
205 | return NULL; | |
206 | } | |
207 | ||
be58e70d JK |
208 | static int parse_funcname(struct userdiff_funcname *f, const char *k, |
209 | const char *v, int cflags) | |
210 | { | |
211 | if (git_config_string(&f->pattern, k, v) < 0) | |
212 | return -1; | |
213 | f->cflags = cflags; | |
6680a087 | 214 | return 0; |
be58e70d JK |
215 | } |
216 | ||
122aa6f9 JK |
217 | static int parse_tristate(int *b, const char *k, const char *v) |
218 | { | |
219 | if (v && !strcasecmp(v, "auto")) | |
220 | *b = -1; | |
221 | else | |
222 | *b = git_config_bool(k, v); | |
6680a087 | 223 | return 0; |
122aa6f9 JK |
224 | } |
225 | ||
d9bae1a1 JK |
226 | static int parse_bool(int *b, const char *k, const char *v) |
227 | { | |
228 | *b = git_config_bool(k, v); | |
6680a087 | 229 | return 0; |
d9bae1a1 JK |
230 | } |
231 | ||
c7534ef4 | 232 | int userdiff_config(const char *k, const char *v) |
be58e70d JK |
233 | { |
234 | struct userdiff_driver *drv; | |
0a5987fe JK |
235 | const char *name, *type; |
236 | int namelen; | |
237 | ||
238 | if (parse_config_key(k, "diff", &name, &namelen, &type) || !name) | |
239 | return 0; | |
240 | ||
241 | drv = userdiff_find_by_namelen(name, namelen); | |
242 | if (!drv) { | |
243 | ALLOC_GROW(drivers, ndrivers+1, drivers_alloc); | |
244 | drv = &drivers[ndrivers++]; | |
245 | memset(drv, 0, sizeof(*drv)); | |
246 | drv->name = xmemdupz(name, namelen); | |
247 | drv->binary = -1; | |
248 | } | |
be58e70d | 249 | |
0a5987fe | 250 | if (!strcmp(type, "funcname")) |
be58e70d | 251 | return parse_funcname(&drv->funcname, k, v, 0); |
0a5987fe | 252 | if (!strcmp(type, "xfuncname")) |
be58e70d | 253 | return parse_funcname(&drv->funcname, k, v, REG_EXTENDED); |
0a5987fe | 254 | if (!strcmp(type, "binary")) |
122aa6f9 | 255 | return parse_tristate(&drv->binary, k, v); |
0a5987fe | 256 | if (!strcmp(type, "command")) |
6680a087 | 257 | return git_config_string(&drv->external, k, v); |
0a5987fe | 258 | if (!strcmp(type, "textconv")) |
6680a087 | 259 | return git_config_string(&drv->textconv, k, v); |
0a5987fe | 260 | if (!strcmp(type, "cachetextconv")) |
d9bae1a1 | 261 | return parse_bool(&drv->textconv_want_cache, k, v); |
0a5987fe | 262 | if (!strcmp(type, "wordregex")) |
6680a087 | 263 | return git_config_string(&drv->word_regex, k, v); |
be58e70d JK |
264 | |
265 | return 0; | |
266 | } | |
267 | ||
268 | struct userdiff_driver *userdiff_find_by_name(const char *name) { | |
269 | int len = strlen(name); | |
270 | return userdiff_find_by_namelen(name, len); | |
271 | } | |
272 | ||
acd00ea0 NTND |
273 | struct userdiff_driver *userdiff_find_by_path(struct index_state *istate, |
274 | const char *path) | |
be58e70d | 275 | { |
2aef63d3 | 276 | static struct attr_check *check; |
be58e70d | 277 | |
2aef63d3 JH |
278 | if (!check) |
279 | check = attr_check_initl("diff", NULL); | |
be58e70d JK |
280 | if (!path) |
281 | return NULL; | |
11877b9e | 282 | git_check_attr(istate, path, check); |
be58e70d | 283 | |
2aef63d3 | 284 | if (ATTR_TRUE(check->items[0].value)) |
be58e70d | 285 | return &driver_true; |
2aef63d3 | 286 | if (ATTR_FALSE(check->items[0].value)) |
be58e70d | 287 | return &driver_false; |
2aef63d3 | 288 | if (ATTR_UNSET(check->items[0].value)) |
be58e70d | 289 | return NULL; |
2aef63d3 | 290 | return userdiff_find_by_name(check->items[0].value); |
be58e70d | 291 | } |
3813e690 JK |
292 | |
293 | struct userdiff_driver *userdiff_get_textconv(struct userdiff_driver *driver) | |
294 | { | |
295 | if (!driver->textconv) | |
296 | return NULL; | |
297 | ||
298 | if (driver->textconv_want_cache && !driver->textconv_cache) { | |
299 | struct notes_cache *c = xmalloc(sizeof(*c)); | |
300 | struct strbuf name = STRBUF_INIT; | |
301 | ||
302 | strbuf_addf(&name, "textconv/%s", driver->name); | |
303 | notes_cache_init(c, name.buf, driver->textconv); | |
304 | driver->textconv_cache = c; | |
460c7eb2 | 305 | strbuf_release(&name); |
3813e690 JK |
306 | } |
307 | ||
308 | return driver; | |
309 | } |