]>
Commit | Line | Data |
---|---|---|
36bf1958 | 1 | #include "git-compat-util.h" |
b2141fc1 | 2 | #include "config.h" |
be58e70d | 3 | #include "userdiff.h" |
be58e70d | 4 | #include "attr.h" |
36bf1958 | 5 | #include "strbuf.h" |
be58e70d JK |
6 | |
7 | static struct userdiff_driver *drivers; | |
8 | static int ndrivers; | |
9 | static int drivers_alloc; | |
10 | ||
2dd75f12 ÆAB |
11 | #define PATTERNS(lang, rx, wrx) { \ |
12 | .name = lang, \ | |
13 | .binary = -1, \ | |
14 | .funcname = { \ | |
15 | .pattern = rx, \ | |
16 | .cflags = REG_EXTENDED, \ | |
17 | }, \ | |
18 | .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \ | |
be391449 | 19 | .word_regex_multi_byte = wrx "|[^[:space:]]", \ |
2dd75f12 ÆAB |
20 | } |
21 | #define IPATTERN(lang, rx, wrx) { \ | |
22 | .name = lang, \ | |
23 | .binary = -1, \ | |
24 | .funcname = { \ | |
25 | .pattern = rx, \ | |
26 | .cflags = REG_EXTENDED | REG_ICASE, \ | |
27 | }, \ | |
28 | .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \ | |
be391449 | 29 | .word_regex_multi_byte = wrx "|[^[:space:]]", \ |
2dd75f12 | 30 | } |
b6029b32 JH |
31 | |
32 | /* | |
33 | * Built-in drivers for various languages, sorted by their names | |
34 | * (except that the "default" is left at the end). | |
35 | * | |
36 | * When writing or updating patterns, assume that the contents these | |
37 | * patterns are applied to are syntactically correct. The patterns | |
38 | * can be simple without implementing all syntactical corner cases, as | |
39 | * long as they are sufficiently permissive. | |
40 | */ | |
be58e70d | 41 | static struct userdiff_driver builtin_drivers[] = { |
e90d065e | 42 | IPATTERN("ada", |
39a87a29 | 43 | "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n" |
e90d065e AJ |
44 | "!^[ \t]*with[ \t].*$\n" |
45 | "^[ \t]*((procedure|function)[ \t]+.*)$\n" | |
46 | "^[ \t]*((package|protected|task)[ \t]+.*)$", | |
47 | /* -- */ | |
48 | "[a-zA-Z][a-zA-Z0-9_]*" | |
39a87a29 | 49 | "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?" |
e90d065e | 50 | "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"), |
2ff6c346 VE |
51 | PATTERNS("bash", |
52 | /* Optional leading indentation */ | |
53 | "^[ \t]*" | |
54 | /* Start of captured text */ | |
55 | "(" | |
56 | "(" | |
57 | /* POSIX identifier with mandatory parentheses */ | |
58 | "[a-zA-Z_][a-zA-Z0-9_]*[ \t]*\\([ \t]*\\))" | |
59 | "|" | |
60 | /* Bashism identifier with optional parentheses */ | |
61 | "(function[ \t]+[a-zA-Z_][a-zA-Z0-9_]*(([ \t]*\\([ \t]*\\))|([ \t]+))" | |
62 | ")" | |
63 | /* Optional whitespace */ | |
64 | "[ \t]*" | |
65 | /* Compound command starting with `{`, `(`, `((` or `[[` */ | |
66 | "(\\{|\\(\\(?|\\[\\[)" | |
67 | /* End of captured text */ | |
68 | ")", | |
69 | /* -- */ | |
70 | /* Characters not in the default $IFS value */ | |
71 | "[^ \t]+"), | |
6d1c9c52 ÆAB |
72 | PATTERNS("bibtex", |
73 | "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$", | |
74 | /* -- */ | |
ddd164d0 ÆAB |
75 | "[={}\"]|[^={}\" \t]+"), |
76 | PATTERNS("cpp", | |
77 | /* Jump targets or access declarations */ | |
78 | "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n" | |
79 | /* functions/methods, variables, and compounds at top level */ | |
80 | "^((::[[:space:]]*)?[A-Za-z_].*)$", | |
81 | /* -- */ | |
350b87cd | 82 | /* identifiers and keywords */ |
ddd164d0 | 83 | "[a-zA-Z_][a-zA-Z0-9_]*" |
350b87cd | 84 | /* decimal and octal integers as well as floatingpoint numbers */ |
386076ec | 85 | "|[0-9][0-9.]*([Ee][-+]?[0-9]+)?[fFlLuU]*" |
350b87cd | 86 | /* hexadecimal and binary integers */ |
386076ec | 87 | "|0[xXbB][0-9a-fA-F]+[lLuU]*" |
350b87cd | 88 | /* floatingpoint numbers that begin with a decimal point */ |
386076ec | 89 | "|\\.[0-9][0-9]*([Ee][-+]?[0-9]+)?[fFlL]?" |
c4fdba33 | 90 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*|<=>"), |
ddd164d0 ÆAB |
91 | PATTERNS("csharp", |
92 | /* Keywords */ | |
93 | "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n" | |
94 | /* Methods and constructors */ | |
95 | "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n" | |
96 | /* Properties */ | |
97 | "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n" | |
98 | /* Type definitions */ | |
c4e31781 | 99 | "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct|record)[ \t]+.*)$\n" |
ddd164d0 ÆAB |
100 | /* Namespace */ |
101 | "^[ \t]*(namespace[ \t]+.*)$", | |
102 | /* -- */ | |
103 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
104 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
105 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"), | |
106 | IPATTERN("css", | |
107 | "![:;][[:space:]]*$\n" | |
108 | "^[:[@.#]?[_a-z0-9].*$", | |
109 | /* -- */ | |
110 | /* | |
111 | * This regex comes from W3C CSS specs. Should theoretically also | |
112 | * allow ISO 10646 characters U+00A0 and higher, | |
113 | * but they are not handled in this regex. | |
114 | */ | |
115 | "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */ | |
116 | "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */ | |
117 | ), | |
3c81760b SB |
118 | PATTERNS("dts", |
119 | "!;\n" | |
8da56a48 | 120 | "!=\n" |
3c81760b | 121 | /* lines beginning with a word optionally preceded by '&' or the root */ |
8da56a48 | 122 | "^[ \t]*((/[ \t]*\\{|&?[a-zA-Z_]).*)", |
3c81760b SB |
123 | /* -- */ |
124 | /* Property names and math operators */ | |
125 | "[a-zA-Z0-9,._+?#-]+" | |
126 | "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"), | |
a807200f ŁN |
127 | PATTERNS("elixir", |
128 | "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$", | |
d1b1384d | 129 | /* -- */ |
a807200f | 130 | /* Atoms, names, and module attributes */ |
d1b1384d | 131 | "[@:]?[a-zA-Z0-9@_?!]+" |
a807200f ŁN |
132 | /* Numbers with specific base */ |
133 | "|[-+]?0[xob][0-9a-fA-F]+" | |
134 | /* Numbers */ | |
135 | "|[-+]?[0-9][0-9_.]*([eE][-+]?[0-9_]+)?" | |
136 | /* Operators and atoms that represent them */ | |
137 | "|:?(\\+\\+|--|\\.\\.|~~~|<>|\\^\\^\\^|<?\\|>|<<<?|>?>>|<<?~|~>?>|<~>|<=|>=|===?|!==?|=~|&&&?|\\|\\|\\|?|=>|<-|\\\\\\\\|->)" | |
138 | /* Not real operators, but should be grouped */ | |
139 | "|:?%[A-Za-z0-9_.]\\{\\}?"), | |
909a5494 | 140 | IPATTERN("fortran", |
b79e6925 | 141 | /* Don't match comment lines */ |
909a5494 | 142 | "!^([C*]|[ \t]*!)\n" |
b79e6925 | 143 | /* Don't match 'module procedure' lines */ |
909a5494 | 144 | "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n" |
b79e6925 | 145 | /* Program, module, block data */ |
909a5494 | 146 | "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA" |
b79e6925 | 147 | /* Subroutines and functions */ |
75c3b6b2 | 148 | "|([^!'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$", |
909a5494 BC |
149 | /* -- */ |
150 | "[a-zA-Z][a-zA-Z0-9_]*" | |
151 | "|\\.([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])\\." | |
152 | /* numbers and format statements like 2E14.4, or ES12.6, 9X. | |
153 | * Don't worry about format statements without leading digits since | |
154 | * they would have been matched above as a variable anyway. */ | |
155 | "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?" | |
664d44ee | 156 | "|//|\\*\\*|::|[/<>=]="), |
6d1c9c52 ÆAB |
157 | IPATTERN("fountain", |
158 | "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$", | |
159 | /* -- */ | |
69f9c87d | 160 | "[^ \t-]+"), |
1dbf0c0a AG |
161 | PATTERNS("golang", |
162 | /* Functions */ | |
163 | "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n" | |
164 | /* Structs and interfaces */ | |
165 | "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)", | |
166 | /* -- */ | |
167 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
168 | "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?" | |
169 | "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"), | |
6d1c9c52 ÆAB |
170 | PATTERNS("html", |
171 | "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$", | |
172 | /* -- */ | |
664d44ee | 173 | "[^<>= \t]+"), |
80c49c3d | 174 | PATTERNS("java", |
be58e70d | 175 | "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n" |
575e6fcf | 176 | /* Class, enum, interface, and record declarations */ |
93d52ed0 | 177 | "^[ \t]*(([a-z-]+[ \t]+)*(class|enum|interface|record)[ \t]+.*)$\n" |
a8cbc895 TH |
178 | /* Method definitions; note that constructor signatures are not */ |
179 | /* matched because they are indistinguishable from method calls. */ | |
180 | "^[ \t]*(([A-Za-z_<>&][][?&<>.,A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$", | |
959e2e64 | 181 | /* -- */ |
80c49c3d TR |
182 | "[a-zA-Z_][a-zA-Z0-9_]*" |
183 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
184 | "|[-+*/<>%&^|=!]=" | |
664d44ee | 185 | "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"), |
09188ed9 JD |
186 | PATTERNS("kotlin", |
187 | "^[ \t]*(([a-z]+[ \t]+)*(fun|class|interface)[ \t]+.*)$", | |
188 | /* -- */ | |
189 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
190 | /* hexadecimal and binary numbers */ | |
191 | "|0[xXbB][0-9a-fA-F_]+[lLuU]*" | |
192 | /* integers and floats */ | |
193 | "|[0-9][0-9_]*([.][0-9_]*)?([Ee][-+]?[0-9]+)?[fFlLuU]*" | |
194 | /* floating point numbers beginning with decimal point */ | |
195 | "|[.][0-9][0-9_]*([Ee][-+]?[0-9]+)?[fFlLuU]?" | |
196 | /* unary and binary operators */ | |
197 | "|[-+*/<>%&^|=!]==?|--|\\+\\+|<<=|>>=|&&|\\|\\||->|\\.\\*|!!|[?:.][.:]"), | |
09dad925 AH |
198 | PATTERNS("markdown", |
199 | "^ {0,3}#{1,6}[ \t].*", | |
6d1c9c52 | 200 | /* -- */ |
09dad925 | 201 | "[^<>= \t]+"), |
53b10a14 | 202 | PATTERNS("matlab", |
2731a784 BL |
203 | /* |
204 | * Octave pattern is mostly the same as matlab, except that '%%%' and | |
91bf382f | 205 | * '##' can also be used to begin code sections, in addition to '%%' |
2731a784 BL |
206 | * that is understood by both. |
207 | */ | |
91bf382f | 208 | "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$", |
6d1c9c52 | 209 | /* -- */ |
53b10a14 | 210 | "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"), |
80c49c3d | 211 | PATTERNS("objc", |
be58e70d JK |
212 | /* Negate C statements that can look like functions */ |
213 | "!^[ \t]*(do|for|if|else|return|switch|while)\n" | |
214 | /* Objective-C methods */ | |
215 | "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n" | |
216 | /* C functions */ | |
959e2e64 | 217 | "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n" |
be58e70d | 218 | /* Objective-C class/protocol definitions */ |
80c49c3d TR |
219 | "^(@(implementation|interface|protocol)[ \t].*)$", |
220 | /* -- */ | |
221 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
222 | "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?" | |
664d44ee | 223 | "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"), |
80c49c3d | 224 | PATTERNS("pascal", |
82512e00 ÆAB |
225 | "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface" |
226 | "|implementation|initialization|finalization)[ \t]*.*)$\n" | |
80c49c3d TR |
227 | "^(.*=[ \t]*(class|record).*)$", |
228 | /* -- */ | |
229 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
230 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+" | |
664d44ee | 231 | "|<>|<=|>=|:=|\\.\\."), |
71a5d4bc | 232 | PATTERNS("perl", |
ea2ca449 JN |
233 | "^package .*\n" |
234 | "^sub [[:alnum:]_':]+[ \t]*" | |
235 | "(\\([^)]*\\)[ \t]*)?" /* prototype */ | |
236 | /* | |
237 | * Attributes. A regex can't count nested parentheses, | |
238 | * so just slurp up whatever we see, taking care not | |
239 | * to accept lines like "sub foo; # defined elsewhere". | |
240 | * | |
241 | * An attribute could contain a semicolon, but at that | |
242 | * point it seems reasonable enough to give up. | |
243 | */ | |
244 | "(:[^;#]*)?" | |
245 | "(\\{[ \t]*)?" /* brace can come here or on the next line */ | |
246 | "(#.*)?$\n" /* comment */ | |
f143d9c6 | 247 | "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*" |
ea2ca449 JN |
248 | "(\\{[ \t]*)?" /* brace can come here or on the next line */ |
249 | "(#.*)?$\n" | |
12f0967a | 250 | "^=head[0-9] .*", /* POD */ |
71a5d4bc JN |
251 | /* -- */ |
252 | "[[:alpha:]_'][[:alnum:]_']*" | |
253 | "|0[xb]?[0-9a-fA-F_]*" | |
254 | /* taking care not to interpret 3..5 as (3.)(.5) */ | |
255 | "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?" | |
256 | "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::" | |
257 | "|&&=|\\|\\|=|//=|\\*\\*=" | |
258 | "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?" | |
259 | "|[-+*/%.^&<>=!|]=" | |
260 | "|=~|!~" | |
664d44ee | 261 | "|<<|<>|<=>|>>"), |
6d2f208c | 262 | PATTERNS("php", |
aff92827 | 263 | "^[\t ]*(((public|protected|private|static|abstract|final)[\t ]+)*function.*)$\n" |
2c7f3aac | 264 | "^[\t ]*((((final|abstract)[\t ]+)?class|enum|interface|trait).*)$", |
80c49c3d TR |
265 | /* -- */ |
266 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
267 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+" | |
664d44ee | 268 | "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"), |
6d1c9c52 ÆAB |
269 | PATTERNS("python", |
270 | "^[ \t]*((class|(async[ \t]+)?def)[ \t].*)$", | |
80c49c3d TR |
271 | /* -- */ |
272 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
273 | "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?" | |
664d44ee | 274 | "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"), |
80c49c3d | 275 | /* -- */ |
6d1c9c52 ÆAB |
276 | PATTERNS("ruby", |
277 | "^[ \t]*((class|module|def)[ \t].*)$", | |
80c49c3d TR |
278 | /* -- */ |
279 | "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*" | |
280 | "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?." | |
664d44ee | 281 | "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"), |
d74e7860 | 282 | PATTERNS("rust", |
a04c7e0f | 283 | "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl|macro_rules!)[< \t]+[^;]*)$", |
d74e7860 MAL |
284 | /* -- */ |
285 | "[a-zA-Z_][a-zA-Z0-9_]*" | |
33be7b38 | 286 | "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?" |
d74e7860 | 287 | "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"), |
a4373903 AR |
288 | PATTERNS("scheme", |
289 | "^[\t ]*(\\(((define|def(struct|syntax|class|method|rules|record|proto|alias)?)[-*/ \t]|(library|module|struct|class)[*+ \t]).*)$", | |
290 | /* | |
291 | * R7RS valid identifiers include any sequence enclosed | |
292 | * within vertical lines having no backslashes | |
293 | */ | |
294 | "\\|([^\\\\]*)\\|" | |
295 | /* All other words should be delimited by spaces or parentheses */ | |
296 | "|([^][)(}{[ \t])+"), | |
80c49c3d | 297 | PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$", |
be391449 | 298 | "\\\\[a-zA-Z@]+|\\\\.|([a-zA-Z0-9]|[^\x01-\x7f])+"), |
a4cf900e | 299 | { "default", NULL, NULL, -1, { NULL, 0 } }, |
be58e70d | 300 | }; |
80c49c3d | 301 | #undef PATTERNS |
909a5494 | 302 | #undef IPATTERN |
be58e70d JK |
303 | |
304 | static struct userdiff_driver driver_true = { | |
2dd75f12 ÆAB |
305 | .name = "diff=true", |
306 | .binary = 0, | |
be58e70d | 307 | }; |
be58e70d JK |
308 | |
309 | static struct userdiff_driver driver_false = { | |
2dd75f12 ÆAB |
310 | .name = "!diff", |
311 | .binary = 1, | |
be58e70d | 312 | }; |
be58e70d | 313 | |
f12fa9ee ÆAB |
314 | struct find_by_namelen_data { |
315 | const char *name; | |
316 | size_t len; | |
317 | struct userdiff_driver *driver; | |
318 | }; | |
319 | ||
320 | static int userdiff_find_by_namelen_cb(struct userdiff_driver *driver, | |
c25d9e52 JK |
321 | enum userdiff_driver_type type UNUSED, |
322 | void *priv) | |
be58e70d | 323 | { |
f12fa9ee ÆAB |
324 | struct find_by_namelen_data *cb_data = priv; |
325 | ||
f0e578c6 | 326 | if (!xstrncmpz(driver->name, cb_data->name, cb_data->len)) { |
f12fa9ee ÆAB |
327 | cb_data->driver = driver; |
328 | return 1; /* tell the caller to stop iterating */ | |
be58e70d | 329 | } |
f12fa9ee ÆAB |
330 | return 0; |
331 | } | |
332 | ||
be391449 RS |
333 | static int regexec_supports_multi_byte_chars(void) |
334 | { | |
335 | static const char not_space[] = "[^[:space:]]"; | |
336 | static const char utf8_multi_byte_char[] = "\xc2\xa3"; | |
337 | regex_t re; | |
338 | regmatch_t match; | |
339 | static int result = -1; | |
340 | ||
341 | if (result != -1) | |
342 | return result; | |
343 | if (regcomp(&re, not_space, REG_EXTENDED)) | |
344 | BUG("invalid regular expression: %s", not_space); | |
345 | result = !regexec(&re, utf8_multi_byte_char, 1, &match, 0) && | |
346 | match.rm_so == 0 && | |
347 | match.rm_eo == strlen(utf8_multi_byte_char); | |
348 | regfree(&re); | |
349 | return result; | |
350 | } | |
351 | ||
f12fa9ee ÆAB |
352 | static struct userdiff_driver *userdiff_find_by_namelen(const char *name, size_t len) |
353 | { | |
354 | struct find_by_namelen_data udcbdata = { | |
355 | .name = name, | |
356 | .len = len, | |
357 | }; | |
358 | for_each_userdiff_driver(userdiff_find_by_namelen_cb, &udcbdata); | |
359 | return udcbdata.driver; | |
be58e70d JK |
360 | } |
361 | ||
be58e70d JK |
362 | static int parse_funcname(struct userdiff_funcname *f, const char *k, |
363 | const char *v, int cflags) | |
364 | { | |
365 | if (git_config_string(&f->pattern, k, v) < 0) | |
366 | return -1; | |
367 | f->cflags = cflags; | |
6680a087 | 368 | return 0; |
be58e70d JK |
369 | } |
370 | ||
122aa6f9 JK |
371 | static int parse_tristate(int *b, const char *k, const char *v) |
372 | { | |
373 | if (v && !strcasecmp(v, "auto")) | |
374 | *b = -1; | |
375 | else | |
376 | *b = git_config_bool(k, v); | |
6680a087 | 377 | return 0; |
122aa6f9 JK |
378 | } |
379 | ||
d9bae1a1 JK |
380 | static int parse_bool(int *b, const char *k, const char *v) |
381 | { | |
382 | *b = git_config_bool(k, v); | |
6680a087 | 383 | return 0; |
d9bae1a1 JK |
384 | } |
385 | ||
c7534ef4 | 386 | int userdiff_config(const char *k, const char *v) |
be58e70d JK |
387 | { |
388 | struct userdiff_driver *drv; | |
0a5987fe | 389 | const char *name, *type; |
f5914f4b | 390 | size_t namelen; |
0a5987fe JK |
391 | |
392 | if (parse_config_key(k, "diff", &name, &namelen, &type) || !name) | |
393 | return 0; | |
394 | ||
395 | drv = userdiff_find_by_namelen(name, namelen); | |
396 | if (!drv) { | |
397 | ALLOC_GROW(drivers, ndrivers+1, drivers_alloc); | |
398 | drv = &drivers[ndrivers++]; | |
399 | memset(drv, 0, sizeof(*drv)); | |
400 | drv->name = xmemdupz(name, namelen); | |
401 | drv->binary = -1; | |
402 | } | |
be58e70d | 403 | |
0a5987fe | 404 | if (!strcmp(type, "funcname")) |
be58e70d | 405 | return parse_funcname(&drv->funcname, k, v, 0); |
0a5987fe | 406 | if (!strcmp(type, "xfuncname")) |
be58e70d | 407 | return parse_funcname(&drv->funcname, k, v, REG_EXTENDED); |
0a5987fe | 408 | if (!strcmp(type, "binary")) |
122aa6f9 | 409 | return parse_tristate(&drv->binary, k, v); |
0a5987fe | 410 | if (!strcmp(type, "command")) |
6680a087 | 411 | return git_config_string(&drv->external, k, v); |
0a5987fe | 412 | if (!strcmp(type, "textconv")) |
6680a087 | 413 | return git_config_string(&drv->textconv, k, v); |
0a5987fe | 414 | if (!strcmp(type, "cachetextconv")) |
d9bae1a1 | 415 | return parse_bool(&drv->textconv_want_cache, k, v); |
0a5987fe | 416 | if (!strcmp(type, "wordregex")) |
6680a087 | 417 | return git_config_string(&drv->word_regex, k, v); |
a4cf900e JC |
418 | if (!strcmp(type, "algorithm")) |
419 | return git_config_string(&drv->algorithm, k, v); | |
be58e70d JK |
420 | |
421 | return 0; | |
422 | } | |
423 | ||
3b335762 NTND |
424 | struct userdiff_driver *userdiff_find_by_name(const char *name) |
425 | { | |
be58e70d | 426 | int len = strlen(name); |
be391449 RS |
427 | struct userdiff_driver *driver = userdiff_find_by_namelen(name, len); |
428 | if (driver && driver->word_regex_multi_byte) { | |
429 | if (regexec_supports_multi_byte_chars()) | |
430 | driver->word_regex = driver->word_regex_multi_byte; | |
431 | driver->word_regex_multi_byte = NULL; | |
432 | } | |
433 | return driver; | |
be58e70d JK |
434 | } |
435 | ||
acd00ea0 NTND |
436 | struct userdiff_driver *userdiff_find_by_path(struct index_state *istate, |
437 | const char *path) | |
be58e70d | 438 | { |
2aef63d3 | 439 | static struct attr_check *check; |
be58e70d | 440 | |
2aef63d3 JH |
441 | if (!check) |
442 | check = attr_check_initl("diff", NULL); | |
be58e70d JK |
443 | if (!path) |
444 | return NULL; | |
44451a2e | 445 | git_check_attr(istate, path, check); |
be58e70d | 446 | |
2aef63d3 | 447 | if (ATTR_TRUE(check->items[0].value)) |
be58e70d | 448 | return &driver_true; |
2aef63d3 | 449 | if (ATTR_FALSE(check->items[0].value)) |
be58e70d | 450 | return &driver_false; |
2aef63d3 | 451 | if (ATTR_UNSET(check->items[0].value)) |
be58e70d | 452 | return NULL; |
2aef63d3 | 453 | return userdiff_find_by_name(check->items[0].value); |
be58e70d | 454 | } |
3813e690 | 455 | |
bd7ad45b NTND |
456 | struct userdiff_driver *userdiff_get_textconv(struct repository *r, |
457 | struct userdiff_driver *driver) | |
3813e690 JK |
458 | { |
459 | if (!driver->textconv) | |
460 | return NULL; | |
461 | ||
462 | if (driver->textconv_want_cache && !driver->textconv_cache) { | |
463 | struct notes_cache *c = xmalloc(sizeof(*c)); | |
464 | struct strbuf name = STRBUF_INIT; | |
465 | ||
466 | strbuf_addf(&name, "textconv/%s", driver->name); | |
bd7ad45b | 467 | notes_cache_init(r, c, name.buf, driver->textconv); |
3813e690 | 468 | driver->textconv_cache = c; |
460c7eb2 | 469 | strbuf_release(&name); |
3813e690 JK |
470 | } |
471 | ||
472 | return driver; | |
473 | } | |
f12fa9ee ÆAB |
474 | |
475 | static int for_each_userdiff_driver_list(each_userdiff_driver_fn fn, | |
476 | enum userdiff_driver_type type, void *cb_data, | |
477 | struct userdiff_driver *drv, | |
478 | int drv_size) | |
479 | { | |
480 | int i; | |
481 | int ret; | |
482 | for (i = 0; i < drv_size; i++) { | |
483 | struct userdiff_driver *item = drv + i; | |
484 | if ((ret = fn(item, type, cb_data))) | |
485 | return ret; | |
486 | } | |
487 | return 0; | |
488 | } | |
489 | ||
490 | int for_each_userdiff_driver(each_userdiff_driver_fn fn, void *cb_data) | |
491 | { | |
492 | int ret; | |
493 | ||
494 | ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_CUSTOM, | |
495 | cb_data, drivers, ndrivers); | |
496 | if (ret) | |
497 | return ret; | |
498 | ||
499 | ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_BUILTIN, | |
500 | cb_data, builtin_drivers, | |
501 | ARRAY_SIZE(builtin_drivers)); | |
502 | if (ret) | |
503 | return ret; | |
504 | ||
505 | return 0; | |
506 | } |