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