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 },            \
  12           word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
  13 #define IPATTERN(name, pattern, word_regex)                     \
  14         { name, NULL, -1, { pattern, REG_EXTENDED | REG_ICASE }, \
  15           word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
  16 static struct userdiff_driver builtin_drivers[] = {
  17 IPATTERN("fortran",
  18          "!^([C*]|[ \t]*!)\n"
  19          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
  20          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
  21                 "|([^'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
  22          /* -- */
  23          "[a-zA-Z][a-zA-Z0-9_]*"
  24          "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
  25          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
  26           * Don't worry about format statements without leading digits since
  27           * they would have been matched above as a variable anyway. */
  28          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
  29          "|//|\\*\\*|::|[/<>=]="),
  30 PATTERNS("html", "^[ \t]*(<[Hh][1-6][ \t].*>.*)$",
  31          "[^<>= \t]+"),
  32 PATTERNS("java",
  33          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
  34          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
  35          /* -- */
  36          "[a-zA-Z_][a-zA-Z0-9_]*"
  37          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  38          "|[-+*/<>%&^|=!]="
  39          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
  40 PATTERNS("objc",
  41          /* Negate C statements that can look like functions */
  42          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
  43          /* Objective-C methods */
  44          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\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 PATTERNS("pascal",
  54          "^((procedure|function|constructor|destructor|interface|"
  55                 "implementation|initialization|finalization)[ \t]*.*)$"
  56          "\n"
  57          "^(.*=[ \t]*(class|record).*)$",
  58          /* -- */
  59          "[a-zA-Z_][a-zA-Z0-9_]*"
  60          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  61          "|<>|<=|>=|:=|\\.\\."),
  62 PATTERNS("perl",
  63          "^[ \t]*package .*;\n"
  64          "^[ \t]*sub .* \\{\n"
  65          "^[A-Z]+ \\{\n"        /* BEGIN, END, ... */
  66          "^=head[0-9] ",        /* POD */
  67          /* -- */
  68          "[[:alpha:]_'][[:alnum:]_']*"
  69          "|0[xb]?[0-9a-fA-F_]*"
  70          /* taking care not to interpret 3..5 as (3.)(.5) */
  71          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
  72          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
  73          "|&&=|\\|\\|=|//=|\\*\\*="
  74          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
  75          "|[-+*/%.^&<>=!|]="
  76          "|=~|!~"
  77          "|<<|<>|<=>|>>"),
  78 PATTERNS("php",
  79          "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
  80          "^[\t ]*(class.*)$",
  81          /* -- */
  82          "[a-zA-Z_][a-zA-Z0-9_]*"
  83          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  84          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
  85 PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
  86          /* -- */
  87          "[a-zA-Z_][a-zA-Z0-9_]*"
  88          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
  89          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
  90          /* -- */
  91 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
  92          /* -- */
  93          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
  94          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
  95          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
  96 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
  97          "[={}\"]|[^={}\" \t]+"),
  98 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
  99          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 100 PATTERNS("cpp",
 101          /* Jump targets or access declarations */
 102          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:.*$\n"
 103          /* C/++ functions/methods at top level */
 104          "^([A-Za-z_][A-Za-z_0-9]*([ \t]+[A-Za-z_][A-Za-z_0-9]*([ \t]*::[ \t]*[^[:space:]]+)?){1,}[ \t]*\\([^;]*)$\n"
 105          /* compound type at top level */
 106          "^((struct|class|enum)[^;]*)$",
 107          /* -- */
 108          "[a-zA-Z_][a-zA-Z0-9_]*"
 109          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 110          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 111 PATTERNS("csharp",
 112          /* Keywords */
 113          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
 114          /* Methods and constructors */
 115          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
 116          /* Properties */
 117          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
 118          /* Type definitions */
 119          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n"
 120          /* Namespace */
 121          "^[ \t]*(namespace[ \t]+.*)$",
 122          /* -- */
 123          "[a-zA-Z_][a-zA-Z0-9_]*"
 124          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 125          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 126 { "default", NULL, -1, { NULL, 0 } },
 127 };
 128 #undef PATTERNS
 129 #undef IPATTERN
 130
 131 static struct userdiff_driver driver_true = {
 132         "diff=true",
 133         NULL,
 134         0,
 135         { NULL, 0 }
 136 };
 137
 138 static struct userdiff_driver driver_false = {
 139         "!diff",
 140         NULL,
 141         1,
 142         { NULL, 0 }
 143 };
 144
 145 static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len)
 146 {
 147         int i;
 148         for (i = 0; i < ndrivers; i++) {
 149                 struct userdiff_driver *drv = drivers + i;
 150                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 151                         return drv;
 152         }
 153         for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
 154                 struct userdiff_driver *drv = builtin_drivers + i;
 155                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 156                         return drv;
 157         }
 158         return NULL;
 159 }
 160
 161 static struct userdiff_driver *parse_driver(const char *var,
 162                 const char *value, const char *type)
 163 {
 164         struct userdiff_driver *drv;
 165         const char *dot;
 166         const char *name;
 167         int namelen;
 168
 169         if (prefixcmp(var, "diff."))
 170                 return NULL;
 171         dot = strrchr(var, '.');
 172         if (dot == var + 4)
 173                 return NULL;
 174         if (strcmp(type, dot+1))
 175                 return NULL;
 176
 177         name = var + 5;
 178         namelen = dot - name;
 179         drv = userdiff_find_by_namelen(name, namelen);
 180         if (!drv) {
 181                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 182                 drv = &drivers[ndrivers++];
 183                 memset(drv, 0, sizeof(*drv));
 184                 drv->name = xmemdupz(name, namelen);
 185                 drv->binary = -1;
 186         }
 187         return drv;
 188 }
 189
 190 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 191                 const char *v, int cflags)
 192 {
 193         if (git_config_string(&f->pattern, k, v) < 0)
 194                 return -1;
 195         f->cflags = cflags;
 196         return 1;
 197 }
 198
 199 static int parse_string(const char **d, const char *k, const char *v)
 200 {
 201         if (git_config_string(d, k, v) < 0)
 202                 return -1;
 203         return 1;
 204 }
 205
 206 static int parse_tristate(int *b, const char *k, const char *v)
 207 {
 208         if (v && !strcasecmp(v, "auto"))
 209                 *b = -1;
 210         else
 211                 *b = git_config_bool(k, v);
 212         return 1;
 213 }
 214
 215 static int parse_bool(int *b, const char *k, const char *v)
 216 {
 217         *b = git_config_bool(k, v);
 218         return 1;
 219 }
 220
 221 int userdiff_config(const char *k, const char *v)
 222 {
 223         struct userdiff_driver *drv;
 224
 225         if ((drv = parse_driver(k, v, "funcname")))
 226                 return parse_funcname(&drv->funcname, k, v, 0);
 227         if ((drv = parse_driver(k, v, "xfuncname")))
 228                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 229         if ((drv = parse_driver(k, v, "binary")))
 230                 return parse_tristate(&drv->binary, k, v);
 231         if ((drv = parse_driver(k, v, "command")))
 232                 return parse_string(&drv->external, k, v);
 233         if ((drv = parse_driver(k, v, "textconv")))
 234                 return parse_string(&drv->textconv, k, v);
 235         if ((drv = parse_driver(k, v, "cachetextconv")))
 236                 return parse_bool(&drv->textconv_want_cache, k, v);
 237         if ((drv = parse_driver(k, v, "wordregex")))
 238                 return parse_string(&drv->word_regex, k, v);
 239
 240         return 0;
 241 }
 242
 243 struct userdiff_driver *userdiff_find_by_name(const char *name) {
 244         int len = strlen(name);
 245         return userdiff_find_by_namelen(name, len);
 246 }
 247
 248 struct userdiff_driver *userdiff_find_by_path(const char *path)
 249 {
 250         static struct git_attr *attr;
 251         struct git_attr_check check;
 252
 253         if (!attr)
 254                 attr = git_attr("diff");
 255         check.attr = attr;
 256
 257         if (!path)
 258                 return NULL;
 259         if (git_checkattr(path, 1, &check))
 260                 return NULL;
 261
 262         if (ATTR_TRUE(check.value))
 263                 return &driver_true;
 264         if (ATTR_FALSE(check.value))
 265                 return &driver_false;
 266         if (ATTR_UNSET(check.value))
 267                 return NULL;
 268         return userdiff_find_by_name(check.value);
 269 }