userdiff.c

   1 #include "cache.h"
   2 #include "config.h"
   3 #include "userdiff.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[ \t]+new|renames|is[ \t]+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#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?"
  25          "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  26 PATTERNS("bash",
  27          /* Optional leading indentation */
  28          "^[ \t]*"
  29          /* Start of captured text */
  30          "("
  31          "("
  32              /* POSIX identifier with mandatory parentheses */
  33              "[a-zA-Z_][a-zA-Z0-9_]*[ \t]*\\([ \t]*\\))"
  34          "|"
  35              /* Bashism identifier with optional parentheses */
  36              "(function[ \t]+[a-zA-Z_][a-zA-Z0-9_]*(([ \t]*\\([ \t]*\\))|([ \t]+))"
  37          ")"
  38          /* Optional whitespace */
  39          "[ \t]*"
  40          /* Compound command starting with `{`, `(`, `((` or `[[` */
  41          "(\\{|\\(\\(?|\\[\\[)"
  42          /* End of captured text */
  43          ")",
  44          /* -- */
  45          /* Characters not in the default $IFS value */
  46          "[^ \t]+"),
  47 PATTERNS("dts",
  48          "!;\n"
  49          "!=\n"
  50          /* lines beginning with a word optionally preceded by '&' or the root */
  51          "^[ \t]*((/[ \t]*\\{|&?[a-zA-Z_]).*)",
  52          /* -- */
  53          /* Property names and math operators */
  54          "[a-zA-Z0-9,._+?#-]+"
  55          "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"),
  56 PATTERNS("elixir",
  57          "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$",
  58          /* -- */
  59          /* Atoms, names, and module attributes */
  60          "[@:]?[a-zA-Z0-9@_?!]+"
  61          /* Numbers with specific base */
  62          "|[-+]?0[xob][0-9a-fA-F]+"
  63          /* Numbers */
  64          "|[-+]?[0-9][0-9_.]*([eE][-+]?[0-9_]+)?"
  65          /* Operators and atoms that represent them */
  66          "|:?(\\+\\+|--|\\.\\.|~~~|<>|\\^\\^\\^|<?\\|>|<<<?|>?>>|<<?~|~>?>|<~>|<=|>=|===?|!==?|=~|&&&?|\\|\\|\\|?|=>|<-|\\\\\\\\|->)"
  67          /* Not real operators, but should be grouped */
  68          "|:?%[A-Za-z0-9_.]\\{\\}?"),
  69 IPATTERN("fortran",
  70          /* Don't match comment lines */
  71          "!^([C*]|[ \t]*!)\n"
  72          /* Don't match 'module procedure' lines */
  73          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
  74          /* Program, module, block data */
  75          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
  76                 /* Subroutines and functions */
  77                 "|([^!'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
  78          /* -- */
  79          "[a-zA-Z][a-zA-Z0-9_]*"
  80          "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
  81          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
  82           * Don't worry about format statements without leading digits since
  83           * they would have been matched above as a variable anyway. */
  84          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
  85          "|//|\\*\\*|::|[/<>=]="),
  86 IPATTERN("fountain", "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$",
  87          "[^ \t-]+"),
  88 PATTERNS("golang",
  89          /* Functions */
  90          "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n"
  91          /* Structs and interfaces */
  92          "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)",
  93          /* -- */
  94          "[a-zA-Z_][a-zA-Z0-9_]*"
  95          "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?"
  96          "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"),
  97 PATTERNS("html", "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$",
  98          "[^<>= \t]+"),
  99 PATTERNS("java",
 100          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
 101          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
 102          /* -- */
 103          "[a-zA-Z_][a-zA-Z0-9_]*"
 104          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 105          "|[-+*/<>%&^|=!]="
 106          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
 107 PATTERNS("markdown",
 108          "^ {0,3}#{1,6}[ \t].*",
 109          "[^<>= \t]+"),
 110 PATTERNS("matlab",
 111          /*
 112           * Octave pattern is mostly the same as matlab, except that '%%%' and
 113           * '##' can also be used to begin code sections, in addition to '%%'
 114           * that is understood by both.
 115           */
 116          "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$",
 117          "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
 118 PATTERNS("objc",
 119          /* Negate C statements that can look like functions */
 120          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
 121          /* Objective-C methods */
 122          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
 123          /* C functions */
 124          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
 125          /* Objective-C class/protocol definitions */
 126          "^(@(implementation|interface|protocol)[ \t].*)$",
 127          /* -- */
 128          "[a-zA-Z_][a-zA-Z0-9_]*"
 129          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 130          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 131 PATTERNS("pascal",
 132          "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|"
 133                 "implementation|initialization|finalization)[ \t]*.*)$"
 134          "\n"
 135          "^(.*=[ \t]*(class|record).*)$",
 136          /* -- */
 137          "[a-zA-Z_][a-zA-Z0-9_]*"
 138          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 139          "|<>|<=|>=|:=|\\.\\."),
 140 PATTERNS("perl",
 141          "^package .*\n"
 142          "^sub [[:alnum:]_':]+[ \t]*"
 143                 "(\\([^)]*\\)[ \t]*)?" /* prototype */
 144                 /*
 145                  * Attributes.  A regex can't count nested parentheses,
 146                  * so just slurp up whatever we see, taking care not
 147                  * to accept lines like "sub foo; # defined elsewhere".
 148                  *
 149                  * An attribute could contain a semicolon, but at that
 150                  * point it seems reasonable enough to give up.
 151                  */
 152                 "(:[^;#]*)?"
 153                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 154                 "(#.*)?$\n" /* comment */
 155          "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
 156                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 157                 "(#.*)?$\n"
 158          "^=head[0-9] .*",      /* POD */
 159          /* -- */
 160          "[[:alpha:]_'][[:alnum:]_']*"
 161          "|0[xb]?[0-9a-fA-F_]*"
 162          /* taking care not to interpret 3..5 as (3.)(.5) */
 163          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
 164          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
 165          "|&&=|\\|\\|=|//=|\\*\\*="
 166          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 167          "|[-+*/%.^&<>=!|]="
 168          "|=~|!~"
 169          "|<<|<>|<=>|>>"),
 170 PATTERNS("php",
 171          "^[\t ]*(((public|protected|private|static|abstract|final)[\t ]+)*function.*)$\n"
 172          "^[\t ]*((((final|abstract)[\t ]+)?class|interface|trait).*)$",
 173          /* -- */
 174          "[a-zA-Z_][a-zA-Z0-9_]*"
 175          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 176          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 177 PATTERNS("python", "^[ \t]*((class|(async[ \t]+)?def)[ \t].*)$",
 178          /* -- */
 179          "[a-zA-Z_][a-zA-Z0-9_]*"
 180          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 181          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 182          /* -- */
 183 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
 184          /* -- */
 185          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 186          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 187          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 188 PATTERNS("rust",
 189          "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl|macro_rules!)[< \t]+[^;]*)$",
 190          /* -- */
 191          "[a-zA-Z_][a-zA-Z0-9_]*"
 192          "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?"
 193          "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"),
 194 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
 195          "[={}\"]|[^={}\" \t]+"),
 196 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 197          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 198 PATTERNS("cpp",
 199          /* Jump targets or access declarations */
 200          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
 201          /* functions/methods, variables, and compounds at top level */
 202          "^((::[[:space:]]*)?[A-Za-z_].*)$",
 203          /* -- */
 204          "[a-zA-Z_][a-zA-Z0-9_]*"
 205          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lLuU]*"
 206          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*"),
 207 PATTERNS("csharp",
 208          /* Keywords */
 209          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
 210          /* Methods and constructors */
 211          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
 212          /* Properties */
 213          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
 214          /* Type definitions */
 215          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n"
 216          /* Namespace */
 217          "^[ \t]*(namespace[ \t]+.*)$",
 218          /* -- */
 219          "[a-zA-Z_][a-zA-Z0-9_]*"
 220          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 221          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 222 IPATTERN("css",
 223          "![:;][[:space:]]*$\n"
 224          "^[:[@.#]?[_a-z0-9].*$",
 225          /* -- */
 226          /*
 227           * This regex comes from W3C CSS specs. Should theoretically also
 228           * allow ISO 10646 characters U+00A0 and higher,
 229           * but they are not handled in this regex.
 230           */
 231          "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */
 232          "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */
 233 ),
 234 { "default", NULL, -1, { NULL, 0 } },
 235 };
 236 #undef PATTERNS
 237 #undef IPATTERN
 238
 239 static struct userdiff_driver driver_true = {
 240         "diff=true",
 241         NULL,
 242         0,
 243         { NULL, 0 }
 244 };
 245
 246 static struct userdiff_driver driver_false = {
 247         "!diff",
 248         NULL,
 249         1,
 250         { NULL, 0 }
 251 };
 252
 253 static struct userdiff_driver *userdiff_find_by_namelen(const char *k, size_t len)
 254 {
 255         int i;
 256         for (i = 0; i < ndrivers; i++) {
 257                 struct userdiff_driver *drv = drivers + i;
 258                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 259                         return drv;
 260         }
 261         for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
 262                 struct userdiff_driver *drv = builtin_drivers + i;
 263                 if (!strncmp(drv->name, k, len) && !drv->name[len])
 264                         return drv;
 265         }
 266         return NULL;
 267 }
 268
 269 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 270                 const char *v, int cflags)
 271 {
 272         if (git_config_string(&f->pattern, k, v) < 0)
 273                 return -1;
 274         f->cflags = cflags;
 275         return 0;
 276 }
 277
 278 static int parse_tristate(int *b, const char *k, const char *v)
 279 {
 280         if (v && !strcasecmp(v, "auto"))
 281                 *b = -1;
 282         else
 283                 *b = git_config_bool(k, v);
 284         return 0;
 285 }
 286
 287 static int parse_bool(int *b, const char *k, const char *v)
 288 {
 289         *b = git_config_bool(k, v);
 290         return 0;
 291 }
 292
 293 int userdiff_config(const char *k, const char *v)
 294 {
 295         struct userdiff_driver *drv;
 296         const char *name, *type;
 297         size_t namelen;
 298
 299         if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
 300                 return 0;
 301
 302         drv = userdiff_find_by_namelen(name, namelen);
 303         if (!drv) {
 304                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 305                 drv = &drivers[ndrivers++];
 306                 memset(drv, 0, sizeof(*drv));
 307                 drv->name = xmemdupz(name, namelen);
 308                 drv->binary = -1;
 309         }
 310
 311         if (!strcmp(type, "funcname"))
 312                 return parse_funcname(&drv->funcname, k, v, 0);
 313         if (!strcmp(type, "xfuncname"))
 314                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 315         if (!strcmp(type, "binary"))
 316                 return parse_tristate(&drv->binary, k, v);
 317         if (!strcmp(type, "command"))
 318                 return git_config_string(&drv->external, k, v);
 319         if (!strcmp(type, "textconv"))
 320                 return git_config_string(&drv->textconv, k, v);
 321         if (!strcmp(type, "cachetextconv"))
 322                 return parse_bool(&drv->textconv_want_cache, k, v);
 323         if (!strcmp(type, "wordregex"))
 324                 return git_config_string(&drv->word_regex, k, v);
 325
 326         return 0;
 327 }
 328
 329 struct userdiff_driver *userdiff_find_by_name(const char *name)
 330 {
 331         int len = strlen(name);
 332         return userdiff_find_by_namelen(name, len);
 333 }
 334
 335 struct userdiff_driver *userdiff_find_by_path(struct index_state *istate,
 336                                               const char *path)
 337 {
 338         static struct attr_check *check;
 339
 340         if (!check)
 341                 check = attr_check_initl("diff", NULL);
 342         if (!path)
 343                 return NULL;
 344         git_check_attr(istate, path, check);
 345
 346         if (ATTR_TRUE(check->items[0].value))
 347                 return &driver_true;
 348         if (ATTR_FALSE(check->items[0].value))
 349                 return &driver_false;
 350         if (ATTR_UNSET(check->items[0].value))
 351                 return NULL;
 352         return userdiff_find_by_name(check->items[0].value);
 353 }
 354
 355 struct userdiff_driver *userdiff_get_textconv(struct repository *r,
 356                                               struct userdiff_driver *driver)
 357 {
 358         if (!driver->textconv)
 359                 return NULL;
 360
 361         if (driver->textconv_want_cache && !driver->textconv_cache) {
 362                 struct notes_cache *c = xmalloc(sizeof(*c));
 363                 struct strbuf name = STRBUF_INIT;
 364
 365                 strbuf_addf(&name, "textconv/%s", driver->name);
 366                 notes_cache_init(r, c, name.buf, driver->textconv);
 367                 driver->textconv_cache = c;
 368                 strbuf_release(&name);
 369         }
 370
 371         return driver;
 372 }