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("ada",
  18          "!^(.*[ \t])?(is new|renames|is separate)([ \t].*)?$\n"
  19          "!^[ \t]*with[ \t].*$\n"
  20          "^[ \t]*((procedure|function)[ \t]+.*)$\n"
  21          "^[ \t]*((package|protected|task)[ \t]+.*)$",
  22          /* -- */
  23          "[a-zA-Z][a-zA-Z0-9_]*"
  24          "|[0-9][-+0-9#_.eE]"
  25          "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  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].*)$",
  31          /* -- */
  32          "[a-zA-Z][a-zA-Z0-9_]*"
  33          "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
  34          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
  35           * Don't worry about format statements without leading digits since
  36           * they would have been matched above as a variable anyway. */
  37          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
  38          "|//|\\*\\*|::|[/<>=]="),
  39 PATTERNS("html", "^[ \t]*(<[Hh][1-6][ \t].*>.*)$",
  40          "[^<>= \t]+"),
  41 PATTERNS("java",
  42          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
  43          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
  44          /* -- */
  45          "[a-zA-Z_][a-zA-Z0-9_]*"
  46          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  47          "|[-+*/<>%&^|=!]="
  48          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
  49 PATTERNS("matlab",
  50          "^[[:space:]]*((classdef|function)[[:space:]].*)$|^%%[[:space:]].*$",
  51          "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
  52 PATTERNS("objc",
  53          /* Negate C statements that can look like functions */
  54          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
  55          /* Objective-C methods */
  56          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
  57          /* C functions */
  58          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
  59          /* Objective-C class/protocol definitions */
  60          "^(@(implementation|interface|protocol)[ \t].*)$",
  61          /* -- */
  62          "[a-zA-Z_][a-zA-Z0-9_]*"
  63          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  64          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
  65 PATTERNS("pascal",
  66          "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|"
  67                 "implementation|initialization|finalization)[ \t]*.*)$"
  68          "\n"
  69          "^(.*=[ \t]*(class|record).*)$",
  70          /* -- */
  71          "[a-zA-Z_][a-zA-Z0-9_]*"
  72          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  73          "|<>|<=|>=|:=|\\.\\."),
  74 PATTERNS("perl",
  75          "^package .*\n"
  76          "^sub [[:alnum:]_':]+[ \t]*"
  77                 "(\\([^)]*\\)[ \t]*)?" /* prototype */
  78                 /*
  79                  * Attributes.  A regex can't count nested parentheses,
  80                  * so just slurp up whatever we see, taking care not
  81                  * to accept lines like "sub foo; # defined elsewhere".
  82                  *
  83                  * An attribute could contain a semicolon, but at that
  84                  * point it seems reasonable enough to give up.
  85                  */
  86                 "(:[^;#]*)?"
  87                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
  88                 "(#.*)?$\n" /* comment */
  89          "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
  90                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
  91                 "(#.*)?$\n"
  92          "^=head[0-9] .*",      /* POD */
  93          /* -- */
  94          "[[:alpha:]_'][[:alnum:]_']*"
  95          "|0[xb]?[0-9a-fA-F_]*"
  96          /* taking care not to interpret 3..5 as (3.)(.5) */
  97          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
  98          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
  99          "|&&=|\\|\\|=|//=|\\*\\*="
 100          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 101          "|[-+*/%.^&<>=!|]="
 102          "|=~|!~"
 103          "|<<|<>|<=>|>>"),
 104 PATTERNS("php",
 105          "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
 106          "^[\t ]*(class.*)$",
 107          /* -- */
 108          "[a-zA-Z_][a-zA-Z0-9_]*"
 109          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 110          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 111 PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
 112          /* -- */
 113          "[a-zA-Z_][a-zA-Z0-9_]*"
 114          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 115          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 116          /* -- */
 117 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
 118          /* -- */
 119          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 120          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 121          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 122 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
 123          "[={}\"]|[^={}\" \t]+"),
 124 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 125          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 126 PATTERNS("cpp",
 127          /* Jump targets or access declarations */
 128          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:.*$\n"
 129          /* C/++ functions/methods at top level */
 130          "^([A-Za-z_][A-Za-z_0-9]*([ \t*]+[A-Za-z_][A-Za-z_0-9]*([ \t]*::[ \t]*[^[:space:]]+)?){1,}[ \t]*\\([^;]*)$\n"
 131          /* compound type at top level */
 132          "^((struct|class|enum)[^;]*)$",
 133          /* -- */
 134          "[a-zA-Z_][a-zA-Z0-9_]*"
 135          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 136          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 137 PATTERNS("csharp",
 138          /* Keywords */
 139          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
 140          /* Methods and constructors */
 141          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\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]?"
 151          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 152 { "default", NULL, -1, { NULL, 0 } },
 153 };
 154 #undef PATTERNS
 155 #undef IPATTERN
 156
 157 static struct userdiff_driver driver_true = {
 158         "diff=true",
 159         NULL,
 160         0,
 161         { NULL, 0 }
 162 };
 163
 164 static struct userdiff_driver driver_false = {
 165         "!diff",
 166         NULL,
 167         1,
 168         { NULL, 0 }
 169 };
 170
 171 static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len)
 172 {
 173         int i;
 174         for (i = 0; i < ndrivers; i++) {
 175                 struct userdiff_driver *drv = drivers + i;
 176                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 177                         return drv;
 178         }
 179         for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
 180                 struct userdiff_driver *drv = builtin_drivers + i;
 181                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 182                         return drv;
 183         }
 184         return NULL;
 185 }
 186
 187 static struct userdiff_driver *parse_driver(const char *var,
 188                 const char *value, const char *type)
 189 {
 190         struct userdiff_driver *drv;
 191         const char *dot;
 192         const char *name;
 193         int namelen;
 194
 195         if (prefixcmp(var, "diff."))
 196                 return NULL;
 197         dot = strrchr(var, '.');
 198         if (dot == var + 4)
 199                 return NULL;
 200         if (strcmp(type, dot+1))
 201                 return NULL;
 202
 203         name = var + 5;
 204         namelen = dot - name;
 205         drv = userdiff_find_by_namelen(name, namelen);
 206         if (!drv) {
 207                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 208                 drv = &drivers[ndrivers++];
 209                 memset(drv, 0, sizeof(*drv));
 210                 drv->name = xmemdupz(name, namelen);
 211                 drv->binary = -1;
 212         }
 213         return drv;
 214 }
 215
 216 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 217                 const char *v, int cflags)
 218 {
 219         if (git_config_string(&f->pattern, k, v) < 0)
 220                 return -1;
 221         f->cflags = cflags;
 222         return 0;
 223 }
 224
 225 static int parse_tristate(int *b, const char *k, const char *v)
 226 {
 227         if (v && !strcasecmp(v, "auto"))
 228                 *b = -1;
 229         else
 230                 *b = git_config_bool(k, v);
 231         return 0;
 232 }
 233
 234 static int parse_bool(int *b, const char *k, const char *v)
 235 {
 236         *b = git_config_bool(k, v);
 237         return 0;
 238 }
 239
 240 int userdiff_config(const char *k, const char *v)
 241 {
 242         struct userdiff_driver *drv;
 243
 244         if ((drv = parse_driver(k, v, "funcname")))
 245                 return parse_funcname(&drv->funcname, k, v, 0);
 246         if ((drv = parse_driver(k, v, "xfuncname")))
 247                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 248         if ((drv = parse_driver(k, v, "binary")))
 249                 return parse_tristate(&drv->binary, k, v);
 250         if ((drv = parse_driver(k, v, "command")))
 251                 return git_config_string(&drv->external, k, v);
 252         if ((drv = parse_driver(k, v, "textconv")))
 253                 return git_config_string(&drv->textconv, k, v);
 254         if ((drv = parse_driver(k, v, "cachetextconv")))
 255                 return parse_bool(&drv->textconv_want_cache, k, v);
 256         if ((drv = parse_driver(k, v, "wordregex")))
 257                 return git_config_string(&drv->word_regex, k, v);
 258
 259         return 0;
 260 }
 261
 262 struct userdiff_driver *userdiff_find_by_name(const char *name) {
 263         int len = strlen(name);
 264         return userdiff_find_by_namelen(name, len);
 265 }
 266
 267 struct userdiff_driver *userdiff_find_by_path(const char *path)
 268 {
 269         static struct git_attr *attr;
 270         struct git_attr_check check;
 271
 272         if (!attr)
 273                 attr = git_attr("diff");
 274         check.attr = attr;
 275
 276         if (!path)
 277                 return NULL;
 278         if (git_check_attr(path, 1, &check))
 279                 return NULL;
 280
 281         if (ATTR_TRUE(check.value))
 282                 return &driver_true;
 283         if (ATTR_FALSE(check.value))
 284                 return &driver_false;
 285         if (ATTR_UNSET(check.value))
 286                 return NULL;
 287         return userdiff_find_by_name(check.value);
 288 }
 289
 290 struct userdiff_driver *userdiff_get_textconv(struct userdiff_driver *driver)
 291 {
 292         if (!driver->textconv)
 293                 return NULL;
 294
 295         if (driver->textconv_want_cache && !driver->textconv_cache) {
 296                 struct notes_cache *c = xmalloc(sizeof(*c));
 297                 struct strbuf name = STRBUF_INIT;
 298
 299                 strbuf_addf(&name, "textconv/%s", driver->name);
 300                 notes_cache_init(c, name.buf, driver->textconv);
 301                 driver->textconv_cache = c;
 302         }
 303
 304         return driver;
 305 }