Merge branch 'jk/fetch-pack' into maint
[git.git] / userdiff.c
blob2ccbee50cb69171062e84c9a1b36baf7e31be2f8
1 #include "cache.h"
2 #include "userdiff.h"
3 #include "attr.h"
5 static struct userdiff_driver *drivers;
6 static int ndrivers;
7 static int drivers_alloc;
9 #define PATTERNS(name, pattern, word_regex) \
10 { name, NULL, -1, { pattern, REG_EXTENDED }, \
11 word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
12 #define IPATTERN(name, pattern, word_regex) \
13 { name, NULL, -1, { pattern, REG_EXTENDED | REG_ICASE }, \
14 word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
15 static struct userdiff_driver builtin_drivers[] = {
16 IPATTERN("ada",
17 "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n"
18 "!^[ \t]*with[ \t].*$\n"
19 "^[ \t]*((procedure|function)[ \t]+.*)$\n"
20 "^[ \t]*((package|protected|task)[ \t]+.*)$",
21 /* -- */
22 "[a-zA-Z][a-zA-Z0-9_]*"
23 "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?"
24 "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
25 IPATTERN("fortran",
26 "!^([C*]|[ \t]*!)\n"
27 "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
28 "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
29 "|([^'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
30 /* -- */
31 "[a-zA-Z][a-zA-Z0-9_]*"
32 "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
33 /* numbers and format statements like 2E14.4, or ES12.6, 9X.
34 * Don't worry about format statements without leading digits since
35 * they would have been matched above as a variable anyway. */
36 "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
37 "|//|\\*\\*|::|[/<>=]="),
38 PATTERNS("html", "^[ \t]*(<[Hh][1-6][ \t].*>.*)$",
39 "[^<>= \t]+"),
40 PATTERNS("java",
41 "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
42 "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
43 /* -- */
44 "[a-zA-Z_][a-zA-Z0-9_]*"
45 "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
46 "|[-+*/<>%&^|=!]="
47 "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
48 PATTERNS("matlab",
49 "^[[:space:]]*((classdef|function)[[:space:]].*)$|^%%[[:space:]].*$",
50 "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
51 PATTERNS("objc",
52 /* Negate C statements that can look like functions */
53 "!^[ \t]*(do|for|if|else|return|switch|while)\n"
54 /* Objective-C methods */
55 "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
56 /* C functions */
57 "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
58 /* Objective-C class/protocol definitions */
59 "^(@(implementation|interface|protocol)[ \t].*)$",
60 /* -- */
61 "[a-zA-Z_][a-zA-Z0-9_]*"
62 "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
63 "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
64 PATTERNS("pascal",
65 "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|"
66 "implementation|initialization|finalization)[ \t]*.*)$"
67 "\n"
68 "^(.*=[ \t]*(class|record).*)$",
69 /* -- */
70 "[a-zA-Z_][a-zA-Z0-9_]*"
71 "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
72 "|<>|<=|>=|:=|\\.\\."),
73 PATTERNS("perl",
74 "^package .*\n"
75 "^sub [[:alnum:]_':]+[ \t]*"
76 "(\\([^)]*\\)[ \t]*)?" /* prototype */
78 * Attributes. A regex can't count nested parentheses,
79 * so just slurp up whatever we see, taking care not
80 * to accept lines like "sub foo; # defined elsewhere".
82 * An attribute could contain a semicolon, but at that
83 * point it seems reasonable enough to give up.
85 "(:[^;#]*)?"
86 "(\\{[ \t]*)?" /* brace can come here or on the next line */
87 "(#.*)?$\n" /* comment */
88 "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
89 "(\\{[ \t]*)?" /* brace can come here or on the next line */
90 "(#.*)?$\n"
91 "^=head[0-9] .*", /* POD */
92 /* -- */
93 "[[:alpha:]_'][[:alnum:]_']*"
94 "|0[xb]?[0-9a-fA-F_]*"
95 /* taking care not to interpret 3..5 as (3.)(.5) */
96 "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
97 "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
98 "|&&=|\\|\\|=|//=|\\*\\*="
99 "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
100 "|[-+*/%.^&<>=!|]="
101 "|=~|!~"
102 "|<<|<>|<=>|>>"),
103 PATTERNS("php",
104 "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
105 "^[\t ]*(class.*)$",
106 /* -- */
107 "[a-zA-Z_][a-zA-Z0-9_]*"
108 "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
109 "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
110 PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
111 /* -- */
112 "[a-zA-Z_][a-zA-Z0-9_]*"
113 "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
114 "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
115 /* -- */
116 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
117 /* -- */
118 "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
119 "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
120 "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
121 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
122 "[={}\"]|[^={}\" \t]+"),
123 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
124 "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
125 PATTERNS("cpp",
126 /* Jump targets or access declarations */
127 "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
128 /* functions/methods, variables, and compounds at top level */
129 "^((::[[:space:]]*)?[A-Za-z_].*)$",
130 /* -- */
131 "[a-zA-Z_][a-zA-Z0-9_]*"
132 "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lLuU]*"
133 "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*"),
134 PATTERNS("csharp",
135 /* Keywords */
136 "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
137 /* Methods and constructors */
138 "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
139 /* Properties */
140 "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
141 /* Type definitions */
142 "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n"
143 /* Namespace */
144 "^[ \t]*(namespace[ \t]+.*)$",
145 /* -- */
146 "[a-zA-Z_][a-zA-Z0-9_]*"
147 "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
148 "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
149 { "default", NULL, -1, { NULL, 0 } },
151 #undef PATTERNS
152 #undef IPATTERN
154 static struct userdiff_driver driver_true = {
155 "diff=true",
156 NULL,
158 { NULL, 0 }
161 static struct userdiff_driver driver_false = {
162 "!diff",
163 NULL,
165 { NULL, 0 }
168 static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len)
170 int i;
171 for (i = 0; i < ndrivers; i++) {
172 struct userdiff_driver *drv = drivers + i;
173 if (!strncmp(drv->name, k, len) && !drv->name[len])
174 return drv;
176 for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
177 struct userdiff_driver *drv = builtin_drivers + i;
178 if (!strncmp(drv->name, k, len) && !drv->name[len])
179 return drv;
181 return NULL;
184 static int parse_funcname(struct userdiff_funcname *f, const char *k,
185 const char *v, int cflags)
187 if (git_config_string(&f->pattern, k, v) < 0)
188 return -1;
189 f->cflags = cflags;
190 return 0;
193 static int parse_tristate(int *b, const char *k, const char *v)
195 if (v && !strcasecmp(v, "auto"))
196 *b = -1;
197 else
198 *b = git_config_bool(k, v);
199 return 0;
202 static int parse_bool(int *b, const char *k, const char *v)
204 *b = git_config_bool(k, v);
205 return 0;
208 int userdiff_config(const char *k, const char *v)
210 struct userdiff_driver *drv;
211 const char *name, *type;
212 int namelen;
214 if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
215 return 0;
217 drv = userdiff_find_by_namelen(name, namelen);
218 if (!drv) {
219 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
220 drv = &drivers[ndrivers++];
221 memset(drv, 0, sizeof(*drv));
222 drv->name = xmemdupz(name, namelen);
223 drv->binary = -1;
226 if (!strcmp(type, "funcname"))
227 return parse_funcname(&drv->funcname, k, v, 0);
228 if (!strcmp(type, "xfuncname"))
229 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
230 if (!strcmp(type, "binary"))
231 return parse_tristate(&drv->binary, k, v);
232 if (!strcmp(type, "command"))
233 return git_config_string(&drv->external, k, v);
234 if (!strcmp(type, "textconv"))
235 return git_config_string(&drv->textconv, k, v);
236 if (!strcmp(type, "cachetextconv"))
237 return parse_bool(&drv->textconv_want_cache, k, v);
238 if (!strcmp(type, "wordregex"))
239 return git_config_string(&drv->word_regex, k, v);
241 return 0;
244 struct userdiff_driver *userdiff_find_by_name(const char *name) {
245 int len = strlen(name);
246 return userdiff_find_by_namelen(name, len);
249 struct userdiff_driver *userdiff_find_by_path(const char *path)
251 static struct git_attr *attr;
252 struct git_attr_check check;
254 if (!attr)
255 attr = git_attr("diff");
256 check.attr = attr;
258 if (!path)
259 return NULL;
260 if (git_check_attr(path, 1, &check))
261 return NULL;
263 if (ATTR_TRUE(check.value))
264 return &driver_true;
265 if (ATTR_FALSE(check.value))
266 return &driver_false;
267 if (ATTR_UNSET(check.value))
268 return NULL;
269 return userdiff_find_by_name(check.value);
272 struct userdiff_driver *userdiff_get_textconv(struct userdiff_driver *driver)
274 if (!driver->textconv)
275 return NULL;
277 if (driver->textconv_want_cache && !driver->textconv_cache) {
278 struct notes_cache *c = xmalloc(sizeof(*c));
279 struct strbuf name = STRBUF_INIT;
281 strbuf_addf(&name, "textconv/%s", driver->name);
282 notes_cache_init(c, name.buf, driver->textconv);
283 driver->textconv_cache = c;
286 return driver;