userdiff.c

   1 #include "cache.h"
   2 #include "userdiff.h"
   3 #include "cache.h"
   4 #include "attr.h"
   5
   6 static struct userdiff_driver *drivers;
   7 static int ndrivers;
   8 static int drivers_alloc;
   9
  10 #define PATTERNS(name, pattern, word_regex)                     \
  11         { name, NULL, -1, { pattern, REG_EXTENDED }, word_regex }
  12 #define IPATTERN(name, pattern, word_regex)                     \
  13         { name, NULL, -1, { pattern, REG_EXTENDED | REG_ICASE }, word_regex }
  14 static struct userdiff_driver builtin_drivers[] = {
  15 IPATTERN("fortran",
  16          "!^([C*]|[ \t]*!)\n"
  17          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
  18          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
  19                 "|([^'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
  20          /* -- */
  21          "[a-zA-Z][a-zA-Z0-9_]*"
  22          "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
  23          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
  24           * Don't worry about format statements without leading digits since
  25           * they would have been matched above as a variable anyway. */
  26          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
  27          "|//|\\*\\*|::|[/<>=]="
  28          "|[^[:space:]]|[\x80-\xff]+"),
  29 PATTERNS("html", "^[ \t]*(<[Hh][1-6][ \t].*>.*)$",
  30          "[^<>= \t]+|[^[:space:]]|[\x80-\xff]+"),
  31 PATTERNS("java",
  32          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
  33          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
  34          /* -- */
  35          "[a-zA-Z_][a-zA-Z0-9_]*"
  36          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  37          "|[-+*/<>%&^|=!]="
  38          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"
  39          "|[^[:space:]]|[\x80-\xff]+"),
  40 PATTERNS("objc",
  41          /* Negate C statements that can look like functions */
  42          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
  43          /* Objective-C methods */
  44          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
  45          /* C functions */
  46          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
  47          /* Objective-C class/protocol definitions */
  48          "^(@(implementation|interface|protocol)[ \t].*)$",
  49          /* -- */
  50          "[a-zA-Z_][a-zA-Z0-9_]*"
  51          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  52          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
  53          "|[^[:space:]]|[\x80-\xff]+"),
  54 PATTERNS("pascal",
  55          "^((procedure|function|constructor|destructor|interface|"
  56                 "implementation|initialization|finalization)[ \t]*.*)$"
  57          "\n"
  58          "^(.*=[ \t]*(class|record).*)$",
  59          /* -- */
  60          "[a-zA-Z_][a-zA-Z0-9_]*"
  61          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  62          "|<>|<=|>=|:=|\\.\\."
  63          "|[^[:space:]]|[\x80-\xff]+"),
  64 PATTERNS("php",
  65          "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
  66          "^[\t ]*(class.*)$",
  67          /* -- */
  68          "[a-zA-Z_][a-zA-Z0-9_]*"
  69          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  70          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"
  71          "|[^[:space:]]|[\x80-\xff]+"),
  72 PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
  73          /* -- */
  74          "[a-zA-Z_][a-zA-Z0-9_]*"
  75          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
  76          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"
  77          "|[^[:space:]|[\x80-\xff]+"),
  78          /* -- */
  79 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
  80          /* -- */
  81          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
  82          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
  83          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"
  84          "|[^[:space:]|[\x80-\xff]+"),
  85 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
  86          "[={}\"]|[^={}\" \t]+"),
  87 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
  88          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+|[^[:space:]]"),
  89 PATTERNS("cpp",
  90          /* Jump targets or access declarations */
  91          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:.*$\n"
  92          /* C/++ functions/methods at top level */
  93          "^([A-Za-z_][A-Za-z_0-9]*([ \t]+[A-Za-z_][A-Za-z_0-9]*([ \t]*::[ \t]*[^[:space:]]+)?){1,}[ \t]*\\([^;]*)$\n"
  94          /* compound type at top level */
  95          "^((struct|class|enum)[^;]*)$",
  96          /* -- */
  97          "[a-zA-Z_][a-zA-Z0-9_]*"
  98          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  99          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
 100          "|[^[:space:]]|[\x80-\xff]+"),
 101 PATTERNS("csharp",
 102          /* Keywords */
 103          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
 104          /* Methods and constructors */
 105          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
 106          /* Properties */
 107          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
 108          /* Type definitions */
 109          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n"
 110          /* Namespace */
 111          "^[ \t]*(namespace[ \t]+.*)$",
 112          /* -- */
 113          "[a-zA-Z_][a-zA-Z0-9_]*"
 114          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 115          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
 116          "|[^[:space:]]|[\x80-\xff]+"),
 117 { "default", NULL, -1, { NULL, 0 } },
 118 };
 119 #undef PATTERNS
 120 #undef IPATTERN
 121
 122 static struct userdiff_driver driver_true = {
 123         "diff=true",
 124         NULL,
 125         0,
 126         { NULL, 0 }
 127 };
 128
 129 static struct userdiff_driver driver_false = {
 130         "!diff",
 131         NULL,
 132         1,
 133         { NULL, 0 }
 134 };
 135
 136 static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len)
 137 {
 138         int i;
 139         for (i = 0; i < ndrivers; i++) {
 140                 struct userdiff_driver *drv = drivers + i;
 141                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 142                         return drv;
 143         }
 144         for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
 145                 struct userdiff_driver *drv = builtin_drivers + i;
 146                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 147                         return drv;
 148         }
 149         return NULL;
 150 }
 151
 152 static struct userdiff_driver *parse_driver(const char *var,
 153                 const char *value, const char *type)
 154 {
 155         struct userdiff_driver *drv;
 156         const char *dot;
 157         const char *name;
 158         int namelen;
 159
 160         if (prefixcmp(var, "diff."))
 161                 return NULL;
 162         dot = strrchr(var, '.');
 163         if (dot == var + 4)
 164                 return NULL;
 165         if (strcmp(type, dot+1))
 166                 return NULL;
 167
 168         name = var + 5;
 169         namelen = dot - name;
 170         drv = userdiff_find_by_namelen(name, namelen);
 171         if (!drv) {
 172                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 173                 drv = &drivers[ndrivers++];
 174                 memset(drv, 0, sizeof(*drv));
 175                 drv->name = xmemdupz(name, namelen);
 176                 drv->binary = -1;
 177         }
 178         return drv;
 179 }
 180
 181 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 182                 const char *v, int cflags)
 183 {
 184         if (git_config_string(&f->pattern, k, v) < 0)
 185                 return -1;
 186         f->cflags = cflags;
 187         return 1;
 188 }
 189
 190 static int parse_string(const char **d, const char *k, const char *v)
 191 {
 192         if (git_config_string(d, k, v) < 0)
 193                 return -1;
 194         return 1;
 195 }
 196
 197 static int parse_tristate(int *b, const char *k, const char *v)
 198 {
 199         if (v && !strcasecmp(v, "auto"))
 200                 *b = -1;
 201         else
 202                 *b = git_config_bool(k, v);
 203         return 1;
 204 }
 205
 206 static int parse_bool(int *b, const char *k, const char *v)
 207 {
 208         *b = git_config_bool(k, v);
 209         return 1;
 210 }
 211
 212 int userdiff_config(const char *k, const char *v)
 213 {
 214         struct userdiff_driver *drv;
 215
 216         if ((drv = parse_driver(k, v, "funcname")))
 217                 return parse_funcname(&drv->funcname, k, v, 0);
 218         if ((drv = parse_driver(k, v, "xfuncname")))
 219                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 220         if ((drv = parse_driver(k, v, "binary")))
 221                 return parse_tristate(&drv->binary, k, v);
 222         if ((drv = parse_driver(k, v, "command")))
 223                 return parse_string(&drv->external, k, v);
 224         if ((drv = parse_driver(k, v, "textconv")))
 225                 return parse_string(&drv->textconv, k, v);
 226         if ((drv = parse_driver(k, v, "cachetextconv")))
 227                 return parse_bool(&drv->textconv_want_cache, k, v);
 228         if ((drv = parse_driver(k, v, "wordregex")))
 229                 return parse_string(&drv->word_regex, k, v);
 230
 231         return 0;
 232 }
 233
 234 struct userdiff_driver *userdiff_find_by_name(const char *name) {
 235         int len = strlen(name);
 236         return userdiff_find_by_namelen(name, len);
 237 }
 238
 239 struct userdiff_driver *userdiff_find_by_path(const char *path)
 240 {
 241         static struct git_attr *attr;
 242         struct git_attr_check check;
 243
 244         if (!attr)
 245                 attr = git_attr("diff");
 246         check.attr = attr;
 247
 248         if (!path)
 249                 return NULL;
 250         if (git_checkattr(path, 1, &check))
 251                 return NULL;
 252
 253         if (ATTR_TRUE(check.value))
 254                 return &driver_true;
 255         if (ATTR_FALSE(check.value))
 256                 return &driver_false;
 257         if (ATTR_UNSET(check.value))
 258                 return NULL;
 259         return userdiff_find_by_name(check.value);
 260 }