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
  17 /*
  18  * Built-in drivers for various languages, sorted by their names
  19  * (except that the "default" is left at the end).
  20  *
  21  * When writing or updating patterns, assume that the contents these
  22  * patterns are applied to are syntactically correct.  The patterns
  23  * can be simple without implementing all syntactical corner cases, as
  24  * long as they are sufficiently permissive.
  25  */
  26 static struct userdiff_driver builtin_drivers[] = {
  27 IPATTERN("ada",
  28          "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n"
  29          "!^[ \t]*with[ \t].*$\n"
  30          "^[ \t]*((procedure|function)[ \t]+.*)$\n"
  31          "^[ \t]*((package|protected|task)[ \t]+.*)$",
  32          /* -- */
  33          "[a-zA-Z][a-zA-Z0-9_]*"
  34          "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?"
  35          "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  36 PATTERNS("bash",
  37          /* Optional leading indentation */
  38          "^[ \t]*"
  39          /* Start of captured text */
  40          "("
  41          "("
  42              /* POSIX identifier with mandatory parentheses */
  43              "[a-zA-Z_][a-zA-Z0-9_]*[ \t]*\\([ \t]*\\))"
  44          "|"
  45              /* Bashism identifier with optional parentheses */
  46              "(function[ \t]+[a-zA-Z_][a-zA-Z0-9_]*(([ \t]*\\([ \t]*\\))|([ \t]+))"
  47          ")"
  48          /* Optional whitespace */
  49          "[ \t]*"
  50          /* Compound command starting with `{`, `(`, `((` or `[[` */
  51          "(\\{|\\(\\(?|\\[\\[)"
  52          /* End of captured text */
  53          ")",
  54          /* -- */
  55          /* Characters not in the default $IFS value */
  56          "[^ \t]+"),
  57 PATTERNS("bibtex",
  58          "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
  59          /* -- */
  60          "[={}\"]|[^={}\" \t]+"),
  61 PATTERNS("cpp",
  62          /* Jump targets or access declarations */
  63          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
  64          /* functions/methods, variables, and compounds at top level */
  65          "^((::[[:space:]]*)?[A-Za-z_].*)$",
  66          /* -- */
  67          /* identifiers and keywords */
  68          "[a-zA-Z_][a-zA-Z0-9_]*"
  69          /* decimal and octal integers as well as floatingpoint numbers */
  70          "|[0-9][0-9.]*([Ee][-+]?[0-9]+)?[fFlLuU]*"
  71          /* hexadecimal and binary integers */
  72          "|0[xXbB][0-9a-fA-F]+[lLuU]*"
  73          /* floatingpoint numbers that begin with a decimal point */
  74          "|\\.[0-9][0-9]*([Ee][-+]?[0-9]+)?[fFlL]?"
  75          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*|<=>"),
  76 PATTERNS("csharp",
  77          /* Keywords */
  78          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
  79          /* Methods and constructors */
  80          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
  81          /* Properties */
  82          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
  83          /* Type definitions */
  84          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct|record)[ \t]+.*)$\n"
  85          /* Namespace */
  86          "^[ \t]*(namespace[ \t]+.*)$",
  87          /* -- */
  88          "[a-zA-Z_][a-zA-Z0-9_]*"
  89          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  90          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
  91 IPATTERN("css",
  92          "![:;][[:space:]]*$\n"
  93          "^[:[@.#]?[_a-z0-9].*$",
  94          /* -- */
  95          /*
  96           * This regex comes from W3C CSS specs. Should theoretically also
  97           * allow ISO 10646 characters U+00A0 and higher,
  98           * but they are not handled in this regex.
  99           */
 100          "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */
 101          "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */
 102 ),
 103 PATTERNS("dts",
 104          "!;\n"
 105          "!=\n"
 106          /* lines beginning with a word optionally preceded by '&' or the root */
 107          "^[ \t]*((/[ \t]*\\{|&?[a-zA-Z_]).*)",
 108          /* -- */
 109          /* Property names and math operators */
 110          "[a-zA-Z0-9,._+?#-]+"
 111          "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"),
 112 PATTERNS("elixir",
 113          "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$",
 114          /* -- */
 115          /* Atoms, names, and module attributes */
 116          "[@:]?[a-zA-Z0-9@_?!]+"
 117          /* Numbers with specific base */
 118          "|[-+]?0[xob][0-9a-fA-F]+"
 119          /* Numbers */
 120          "|[-+]?[0-9][0-9_.]*([eE][-+]?[0-9_]+)?"
 121          /* Operators and atoms that represent them */
 122          "|:?(\\+\\+|--|\\.\\.|~~~|<>|\\^\\^\\^|<?\\|>|<<<?|>?>>|<<?~|~>?>|<~>|<=|>=|===?|!==?|=~|&&&?|\\|\\|\\|?|=>|<-|\\\\\\\\|->)"
 123          /* Not real operators, but should be grouped */
 124          "|:?%[A-Za-z0-9_.]\\{\\}?"),
 125 IPATTERN("fortran",
 126          /* Don't match comment lines */
 127          "!^([C*]|[ \t]*!)\n"
 128          /* Don't match 'module procedure' lines */
 129          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
 130          /* Program, module, block data */
 131          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
 132                 /* Subroutines and functions */
 133                 "|([^!'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
 134          /* -- */
 135          "[a-zA-Z][a-zA-Z0-9_]*"
 136          "|\\.([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])\\."
 137          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
 138           * Don't worry about format statements without leading digits since
 139           * they would have been matched above as a variable anyway. */
 140          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
 141          "|//|\\*\\*|::|[/<>=]="),
 142 IPATTERN("fountain",
 143          "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$",
 144          /* -- */
 145          "[^ \t-]+"),
 146 PATTERNS("golang",
 147          /* Functions */
 148          "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n"
 149          /* Structs and interfaces */
 150          "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)",
 151          /* -- */
 152          "[a-zA-Z_][a-zA-Z0-9_]*"
 153          "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?"
 154          "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"),
 155 PATTERNS("html",
 156          "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$",
 157          /* -- */
 158          "[^<>= \t]+"),
 159 PATTERNS("java",
 160          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
 161          /* Class, enum, and interface declarations */
 162          "^[ \t]*(([a-z]+[ \t]+)*(class|enum|interface)[ \t]+[A-Za-z][A-Za-z0-9_$]*[ \t]+.*)$\n"
 163          /* Method definitions; note that constructor signatures are not */
 164          /* matched because they are indistinguishable from method calls. */
 165          "^[ \t]*(([A-Za-z_<>&][][?&<>.,A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
 166          /* -- */
 167          "[a-zA-Z_][a-zA-Z0-9_]*"
 168          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 169          "|[-+*/<>%&^|=!]="
 170          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
 171 PATTERNS("markdown",
 172          "^ {0,3}#{1,6}[ \t].*",
 173          /* -- */
 174          "[^<>= \t]+"),
 175 PATTERNS("matlab",
 176          /*
 177           * Octave pattern is mostly the same as matlab, except that '%%%' and
 178           * '##' can also be used to begin code sections, in addition to '%%'
 179           * that is understood by both.
 180           */
 181          "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$",
 182          /* -- */
 183          "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
 184 PATTERNS("objc",
 185          /* Negate C statements that can look like functions */
 186          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
 187          /* Objective-C methods */
 188          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
 189          /* C functions */
 190          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
 191          /* Objective-C class/protocol definitions */
 192          "^(@(implementation|interface|protocol)[ \t].*)$",
 193          /* -- */
 194          "[a-zA-Z_][a-zA-Z0-9_]*"
 195          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 196          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 197 PATTERNS("pascal",
 198          "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface"
 199          "|implementation|initialization|finalization)[ \t]*.*)$\n"
 200          "^(.*=[ \t]*(class|record).*)$",
 201          /* -- */
 202          "[a-zA-Z_][a-zA-Z0-9_]*"
 203          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 204          "|<>|<=|>=|:=|\\.\\."),
 205 PATTERNS("perl",
 206          "^package .*\n"
 207          "^sub [[:alnum:]_':]+[ \t]*"
 208                 "(\\([^)]*\\)[ \t]*)?" /* prototype */
 209                 /*
 210                  * Attributes.  A regex can't count nested parentheses,
 211                  * so just slurp up whatever we see, taking care not
 212                  * to accept lines like "sub foo; # defined elsewhere".
 213                  *
 214                  * An attribute could contain a semicolon, but at that
 215                  * point it seems reasonable enough to give up.
 216                  */
 217                 "(:[^;#]*)?"
 218                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 219                 "(#.*)?$\n" /* comment */
 220          "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
 221                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 222                 "(#.*)?$\n"
 223          "^=head[0-9] .*",      /* POD */
 224          /* -- */
 225          "[[:alpha:]_'][[:alnum:]_']*"
 226          "|0[xb]?[0-9a-fA-F_]*"
 227          /* taking care not to interpret 3..5 as (3.)(.5) */
 228          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
 229          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
 230          "|&&=|\\|\\|=|//=|\\*\\*="
 231          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 232          "|[-+*/%.^&<>=!|]="
 233          "|=~|!~"
 234          "|<<|<>|<=>|>>"),
 235 PATTERNS("php",
 236          "^[\t ]*(((public|protected|private|static|abstract|final)[\t ]+)*function.*)$\n"
 237          "^[\t ]*((((final|abstract)[\t ]+)?class|enum|interface|trait).*)$",
 238          /* -- */
 239          "[a-zA-Z_][a-zA-Z0-9_]*"
 240          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 241          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 242 PATTERNS("python",
 243          "^[ \t]*((class|(async[ \t]+)?def)[ \t].*)$",
 244          /* -- */
 245          "[a-zA-Z_][a-zA-Z0-9_]*"
 246          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 247          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 248          /* -- */
 249 PATTERNS("ruby",
 250          "^[ \t]*((class|module|def)[ \t].*)$",
 251          /* -- */
 252          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 253          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 254          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 255 PATTERNS("rust",
 256          "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl|macro_rules!)[< \t]+[^;]*)$",
 257          /* -- */
 258          "[a-zA-Z_][a-zA-Z0-9_]*"
 259          "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?"
 260          "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"),
 261 PATTERNS("scheme",
 262          "^[\t ]*(\\(((define|def(struct|syntax|class|method|rules|record|proto|alias)?)[-*/ \t]|(library|module|struct|class)[*+ \t]).*)$",
 263          /*
 264           * R7RS valid identifiers include any sequence enclosed
 265           * within vertical lines having no backslashes
 266           */
 267          "\\|([^\\\\]*)\\|"
 268          /* All other words should be delimited by spaces or parentheses */
 269          "|([^][)(}{[ \t])+"),
 270 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 271          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 272 { "default", NULL, -1, { NULL, 0 } },
 273 };
 274 #undef PATTERNS
 275 #undef IPATTERN
 276
 277 static struct userdiff_driver driver_true = {
 278         "diff=true",
 279         NULL,
 280         0,
 281         { NULL, 0 }
 282 };
 283
 284 static struct userdiff_driver driver_false = {
 285         "!diff",
 286         NULL,
 287         1,
 288         { NULL, 0 }
 289 };
 290
 291 struct find_by_namelen_data {
 292         const char *name;
 293         size_t len;
 294         struct userdiff_driver *driver;
 295 };
 296
 297 static int userdiff_find_by_namelen_cb(struct userdiff_driver *driver,
 298                                        enum userdiff_driver_type type, void *priv)
 299 {
 300         struct find_by_namelen_data *cb_data = priv;
 301
 302         if (!strncmp(driver->name, cb_data->name, cb_data->len) &&
 303             !driver->name[cb_data->len]) {
 304                 cb_data->driver = driver;
 305                 return 1; /* tell the caller to stop iterating */
 306         }
 307         return 0;
 308 }
 309
 310 static struct userdiff_driver *userdiff_find_by_namelen(const char *name, size_t len)
 311 {
 312         struct find_by_namelen_data udcbdata = {
 313                 .name = name,
 314                 .len = len,
 315         };
 316         for_each_userdiff_driver(userdiff_find_by_namelen_cb, &udcbdata);
 317         return udcbdata.driver;
 318 }
 319
 320 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 321                 const char *v, int cflags)
 322 {
 323         if (git_config_string(&f->pattern, k, v) < 0)
 324                 return -1;
 325         f->cflags = cflags;
 326         return 0;
 327 }
 328
 329 static int parse_tristate(int *b, const char *k, const char *v)
 330 {
 331         if (v && !strcasecmp(v, "auto"))
 332                 *b = -1;
 333         else
 334                 *b = git_config_bool(k, v);
 335         return 0;
 336 }
 337
 338 static int parse_bool(int *b, const char *k, const char *v)
 339 {
 340         *b = git_config_bool(k, v);
 341         return 0;
 342 }
 343
 344 int userdiff_config(const char *k, const char *v)
 345 {
 346         struct userdiff_driver *drv;
 347         const char *name, *type;
 348         size_t namelen;
 349
 350         if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
 351                 return 0;
 352
 353         drv = userdiff_find_by_namelen(name, namelen);
 354         if (!drv) {
 355                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 356                 drv = &drivers[ndrivers++];
 357                 memset(drv, 0, sizeof(*drv));
 358                 drv->name = xmemdupz(name, namelen);
 359                 drv->binary = -1;
 360         }
 361
 362         if (!strcmp(type, "funcname"))
 363                 return parse_funcname(&drv->funcname, k, v, 0);
 364         if (!strcmp(type, "xfuncname"))
 365                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 366         if (!strcmp(type, "binary"))
 367                 return parse_tristate(&drv->binary, k, v);
 368         if (!strcmp(type, "command"))
 369                 return git_config_string(&drv->external, k, v);
 370         if (!strcmp(type, "textconv"))
 371                 return git_config_string(&drv->textconv, k, v);
 372         if (!strcmp(type, "cachetextconv"))
 373                 return parse_bool(&drv->textconv_want_cache, k, v);
 374         if (!strcmp(type, "wordregex"))
 375                 return git_config_string(&drv->word_regex, k, v);
 376
 377         return 0;
 378 }
 379
 380 struct userdiff_driver *userdiff_find_by_name(const char *name)
 381 {
 382         int len = strlen(name);
 383         return userdiff_find_by_namelen(name, len);
 384 }
 385
 386 struct userdiff_driver *userdiff_find_by_path(struct index_state *istate,
 387                                               const char *path)
 388 {
 389         static struct attr_check *check;
 390
 391         if (!check)
 392                 check = attr_check_initl("diff", NULL);
 393         if (!path)
 394                 return NULL;
 395         git_check_attr(istate, path, check);
 396
 397         if (ATTR_TRUE(check->items[0].value))
 398                 return &driver_true;
 399         if (ATTR_FALSE(check->items[0].value))
 400                 return &driver_false;
 401         if (ATTR_UNSET(check->items[0].value))
 402                 return NULL;
 403         return userdiff_find_by_name(check->items[0].value);
 404 }
 405
 406 struct userdiff_driver *userdiff_get_textconv(struct repository *r,
 407                                               struct userdiff_driver *driver)
 408 {
 409         if (!driver->textconv)
 410                 return NULL;
 411
 412         if (driver->textconv_want_cache && !driver->textconv_cache) {
 413                 struct notes_cache *c = xmalloc(sizeof(*c));
 414                 struct strbuf name = STRBUF_INIT;
 415
 416                 strbuf_addf(&name, "textconv/%s", driver->name);
 417                 notes_cache_init(r, c, name.buf, driver->textconv);
 418                 driver->textconv_cache = c;
 419                 strbuf_release(&name);
 420         }
 421
 422         return driver;
 423 }
 424
 425 static int for_each_userdiff_driver_list(each_userdiff_driver_fn fn,
 426                                          enum userdiff_driver_type type, void *cb_data,
 427                                          struct userdiff_driver *drv,
 428                                          int drv_size)
 429 {
 430         int i;
 431         int ret;
 432         for (i = 0; i < drv_size; i++) {
 433                 struct userdiff_driver *item = drv + i;
 434                 if ((ret = fn(item, type, cb_data)))
 435                         return ret;
 436         }
 437         return 0;
 438 }
 439
 440 int for_each_userdiff_driver(each_userdiff_driver_fn fn, void *cb_data)
 441 {
 442         int ret;
 443
 444         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_CUSTOM,
 445                                             cb_data, drivers, ndrivers);
 446         if (ret)
 447                 return ret;
 448
 449         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_BUILTIN,
 450                                             cb_data, builtin_drivers,
 451                                             ARRAY_SIZE(builtin_drivers));
 452         if (ret)
 453                 return ret;
 454
 455         return 0;
 456 }