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(lang, rx, wrx) { \
  11         .name = lang, \
  12         .binary = -1, \
  13         .funcname = { \
  14                 .pattern = rx, \
  15                 .cflags = REG_EXTENDED, \
  16         }, \
  17         .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \
  18 }
  19 #define IPATTERN(lang, rx, wrx) { \
  20         .name = lang, \
  21         .binary = -1, \
  22         .funcname = { \
  23                 .pattern = rx, \
  24                 .cflags = REG_EXTENDED | REG_ICASE, \
  25         }, \
  26         .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \
  27 }
  28
  29 /*
  30  * Built-in drivers for various languages, sorted by their names
  31  * (except that the "default" is left at the end).
  32  *
  33  * When writing or updating patterns, assume that the contents these
  34  * patterns are applied to are syntactically correct.  The patterns
  35  * can be simple without implementing all syntactical corner cases, as
  36  * long as they are sufficiently permissive.
  37  */
  38 static struct userdiff_driver builtin_drivers[] = {
  39 IPATTERN("ada",
  40          "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n"
  41          "!^[ \t]*with[ \t].*$\n"
  42          "^[ \t]*((procedure|function)[ \t]+.*)$\n"
  43          "^[ \t]*((package|protected|task)[ \t]+.*)$",
  44          /* -- */
  45          "[a-zA-Z][a-zA-Z0-9_]*"
  46          "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?"
  47          "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  48 PATTERNS("bash",
  49          /* Optional leading indentation */
  50          "^[ \t]*"
  51          /* Start of captured text */
  52          "("
  53          "("
  54              /* POSIX identifier with mandatory parentheses */
  55              "[a-zA-Z_][a-zA-Z0-9_]*[ \t]*\\([ \t]*\\))"
  56          "|"
  57              /* Bashism identifier with optional parentheses */
  58              "(function[ \t]+[a-zA-Z_][a-zA-Z0-9_]*(([ \t]*\\([ \t]*\\))|([ \t]+))"
  59          ")"
  60          /* Optional whitespace */
  61          "[ \t]*"
  62          /* Compound command starting with `{`, `(`, `((` or `[[` */
  63          "(\\{|\\(\\(?|\\[\\[)"
  64          /* End of captured text */
  65          ")",
  66          /* -- */
  67          /* Characters not in the default $IFS value */
  68          "[^ \t]+"),
  69 PATTERNS("bibtex",
  70          "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
  71          /* -- */
  72          "[={}\"]|[^={}\" \t]+"),
  73 PATTERNS("cpp",
  74          /* Jump targets or access declarations */
  75          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
  76          /* functions/methods, variables, and compounds at top level */
  77          "^((::[[:space:]]*)?[A-Za-z_].*)$",
  78          /* -- */
  79          /* identifiers and keywords */
  80          "[a-zA-Z_][a-zA-Z0-9_]*"
  81          /* decimal and octal integers as well as floatingpoint numbers */
  82          "|[0-9][0-9.]*([Ee][-+]?[0-9]+)?[fFlLuU]*"
  83          /* hexadecimal and binary integers */
  84          "|0[xXbB][0-9a-fA-F]+[lLuU]*"
  85          /* floatingpoint numbers that begin with a decimal point */
  86          "|\\.[0-9][0-9]*([Ee][-+]?[0-9]+)?[fFlL]?"
  87          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*|<=>"),
  88 PATTERNS("csharp",
  89          /* Keywords */
  90          "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
  91          /* Methods and constructors */
  92          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
  93          /* Properties */
  94          "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
  95          /* Type definitions */
  96          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct|record)[ \t]+.*)$\n"
  97          /* Namespace */
  98          "^[ \t]*(namespace[ \t]+.*)$",
  99          /* -- */
 100          "[a-zA-Z_][a-zA-Z0-9_]*"
 101          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 102          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 103 IPATTERN("css",
 104          "![:;][[:space:]]*$\n"
 105          "^[:[@.#]?[_a-z0-9].*$",
 106          /* -- */
 107          /*
 108           * This regex comes from W3C CSS specs. Should theoretically also
 109           * allow ISO 10646 characters U+00A0 and higher,
 110           * but they are not handled in this regex.
 111           */
 112          "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */
 113          "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */
 114 ),
 115 PATTERNS("dts",
 116          "!;\n"
 117          "!=\n"
 118          /* lines beginning with a word optionally preceded by '&' or the root */
 119          "^[ \t]*((/[ \t]*\\{|&?[a-zA-Z_]).*)",
 120          /* -- */
 121          /* Property names and math operators */
 122          "[a-zA-Z0-9,._+?#-]+"
 123          "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"),
 124 PATTERNS("elixir",
 125          "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$",
 126          /* -- */
 127          /* Atoms, names, and module attributes */
 128          "[@:]?[a-zA-Z0-9@_?!]+"
 129          /* Numbers with specific base */
 130          "|[-+]?0[xob][0-9a-fA-F]+"
 131          /* Numbers */
 132          "|[-+]?[0-9][0-9_.]*([eE][-+]?[0-9_]+)?"
 133          /* Operators and atoms that represent them */
 134          "|:?(\\+\\+|--|\\.\\.|~~~|<>|\\^\\^\\^|<?\\|>|<<<?|>?>>|<<?~|~>?>|<~>|<=|>=|===?|!==?|=~|&&&?|\\|\\|\\|?|=>|<-|\\\\\\\\|->)"
 135          /* Not real operators, but should be grouped */
 136          "|:?%[A-Za-z0-9_.]\\{\\}?"),
 137 IPATTERN("fortran",
 138          /* Don't match comment lines */
 139          "!^([C*]|[ \t]*!)\n"
 140          /* Don't match 'module procedure' lines */
 141          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
 142          /* Program, module, block data */
 143          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
 144                 /* Subroutines and functions */
 145                 "|([^!'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
 146          /* -- */
 147          "[a-zA-Z][a-zA-Z0-9_]*"
 148          "|\\.([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])\\."
 149          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
 150           * Don't worry about format statements without leading digits since
 151           * they would have been matched above as a variable anyway. */
 152          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
 153          "|//|\\*\\*|::|[/<>=]="),
 154 IPATTERN("fountain",
 155          "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$",
 156          /* -- */
 157          "[^ \t-]+"),
 158 PATTERNS("golang",
 159          /* Functions */
 160          "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n"
 161          /* Structs and interfaces */
 162          "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)",
 163          /* -- */
 164          "[a-zA-Z_][a-zA-Z0-9_]*"
 165          "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?"
 166          "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"),
 167 PATTERNS("html",
 168          "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$",
 169          /* -- */
 170          "[^<>= \t]+"),
 171 PATTERNS("java",
 172          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
 173          /* Class, enum, and interface declarations */
 174          "^[ \t]*(([a-z]+[ \t]+)*(class|enum|interface)[ \t]+[A-Za-z][A-Za-z0-9_$]*[ \t]+.*)$\n"
 175          /* Method definitions; note that constructor signatures are not */
 176          /* matched because they are indistinguishable from method calls. */
 177          "^[ \t]*(([A-Za-z_<>&][][?&<>.,A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
 178          /* -- */
 179          "[a-zA-Z_][a-zA-Z0-9_]*"
 180          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 181          "|[-+*/<>%&^|=!]="
 182          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
 183 PATTERNS("markdown",
 184          "^ {0,3}#{1,6}[ \t].*",
 185          /* -- */
 186          "[^<>= \t]+"),
 187 PATTERNS("matlab",
 188          /*
 189           * Octave pattern is mostly the same as matlab, except that '%%%' and
 190           * '##' can also be used to begin code sections, in addition to '%%'
 191           * that is understood by both.
 192           */
 193          "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$",
 194          /* -- */
 195          "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
 196 PATTERNS("objc",
 197          /* Negate C statements that can look like functions */
 198          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
 199          /* Objective-C methods */
 200          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
 201          /* C functions */
 202          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
 203          /* Objective-C class/protocol definitions */
 204          "^(@(implementation|interface|protocol)[ \t].*)$",
 205          /* -- */
 206          "[a-zA-Z_][a-zA-Z0-9_]*"
 207          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 208          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 209 PATTERNS("pascal",
 210          "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface"
 211          "|implementation|initialization|finalization)[ \t]*.*)$\n"
 212          "^(.*=[ \t]*(class|record).*)$",
 213          /* -- */
 214          "[a-zA-Z_][a-zA-Z0-9_]*"
 215          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 216          "|<>|<=|>=|:=|\\.\\."),
 217 PATTERNS("perl",
 218          "^package .*\n"
 219          "^sub [[:alnum:]_':]+[ \t]*"
 220                 "(\\([^)]*\\)[ \t]*)?" /* prototype */
 221                 /*
 222                  * Attributes.  A regex can't count nested parentheses,
 223                  * so just slurp up whatever we see, taking care not
 224                  * to accept lines like "sub foo; # defined elsewhere".
 225                  *
 226                  * An attribute could contain a semicolon, but at that
 227                  * point it seems reasonable enough to give up.
 228                  */
 229                 "(:[^;#]*)?"
 230                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 231                 "(#.*)?$\n" /* comment */
 232          "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
 233                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 234                 "(#.*)?$\n"
 235          "^=head[0-9] .*",      /* POD */
 236          /* -- */
 237          "[[:alpha:]_'][[:alnum:]_']*"
 238          "|0[xb]?[0-9a-fA-F_]*"
 239          /* taking care not to interpret 3..5 as (3.)(.5) */
 240          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
 241          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
 242          "|&&=|\\|\\|=|//=|\\*\\*="
 243          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 244          "|[-+*/%.^&<>=!|]="
 245          "|=~|!~"
 246          "|<<|<>|<=>|>>"),
 247 PATTERNS("php",
 248          "^[\t ]*(((public|protected|private|static|abstract|final)[\t ]+)*function.*)$\n"
 249          "^[\t ]*((((final|abstract)[\t ]+)?class|enum|interface|trait).*)$",
 250          /* -- */
 251          "[a-zA-Z_][a-zA-Z0-9_]*"
 252          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 253          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 254 PATTERNS("python",
 255          "^[ \t]*((class|(async[ \t]+)?def)[ \t].*)$",
 256          /* -- */
 257          "[a-zA-Z_][a-zA-Z0-9_]*"
 258          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 259          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 260          /* -- */
 261 PATTERNS("ruby",
 262          "^[ \t]*((class|module|def)[ \t].*)$",
 263          /* -- */
 264          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 265          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 266          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 267 PATTERNS("rust",
 268          "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl|macro_rules!)[< \t]+[^;]*)$",
 269          /* -- */
 270          "[a-zA-Z_][a-zA-Z0-9_]*"
 271          "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?"
 272          "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"),
 273 PATTERNS("scheme",
 274          "^[\t ]*(\\(((define|def(struct|syntax|class|method|rules|record|proto|alias)?)[-*/ \t]|(library|module|struct|class)[*+ \t]).*)$",
 275          /*
 276           * R7RS valid identifiers include any sequence enclosed
 277           * within vertical lines having no backslashes
 278           */
 279          "\\|([^\\\\]*)\\|"
 280          /* All other words should be delimited by spaces or parentheses */
 281          "|([^][)(}{[ \t])+"),
 282 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 283          "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 284 { "default", NULL, -1, { NULL, 0 } },
 285 };
 286 #undef PATTERNS
 287 #undef IPATTERN
 288
 289 static struct userdiff_driver driver_true = {
 290         .name = "diff=true",
 291         .binary = 0,
 292 };
 293
 294 static struct userdiff_driver driver_false = {
 295         .name = "!diff",
 296         .binary = 1,
 297 };
 298
 299 struct find_by_namelen_data {
 300         const char *name;
 301         size_t len;
 302         struct userdiff_driver *driver;
 303 };
 304
 305 static int userdiff_find_by_namelen_cb(struct userdiff_driver *driver,
 306                                        enum userdiff_driver_type type, void *priv)
 307 {
 308         struct find_by_namelen_data *cb_data = priv;
 309
 310         if (!strncmp(driver->name, cb_data->name, cb_data->len) &&
 311             !driver->name[cb_data->len]) {
 312                 cb_data->driver = driver;
 313                 return 1; /* tell the caller to stop iterating */
 314         }
 315         return 0;
 316 }
 317
 318 static struct userdiff_driver *userdiff_find_by_namelen(const char *name, size_t len)
 319 {
 320         struct find_by_namelen_data udcbdata = {
 321                 .name = name,
 322                 .len = len,
 323         };
 324         for_each_userdiff_driver(userdiff_find_by_namelen_cb, &udcbdata);
 325         return udcbdata.driver;
 326 }
 327
 328 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 329                 const char *v, int cflags)
 330 {
 331         if (git_config_string(&f->pattern, k, v) < 0)
 332                 return -1;
 333         f->cflags = cflags;
 334         return 0;
 335 }
 336
 337 static int parse_tristate(int *b, const char *k, const char *v)
 338 {
 339         if (v && !strcasecmp(v, "auto"))
 340                 *b = -1;
 341         else
 342                 *b = git_config_bool(k, v);
 343         return 0;
 344 }
 345
 346 static int parse_bool(int *b, const char *k, const char *v)
 347 {
 348         *b = git_config_bool(k, v);
 349         return 0;
 350 }
 351
 352 int userdiff_config(const char *k, const char *v)
 353 {
 354         struct userdiff_driver *drv;
 355         const char *name, *type;
 356         size_t namelen;
 357
 358         if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
 359                 return 0;
 360
 361         drv = userdiff_find_by_namelen(name, namelen);
 362         if (!drv) {
 363                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 364                 drv = &drivers[ndrivers++];
 365                 memset(drv, 0, sizeof(*drv));
 366                 drv->name = xmemdupz(name, namelen);
 367                 drv->binary = -1;
 368         }
 369
 370         if (!strcmp(type, "funcname"))
 371                 return parse_funcname(&drv->funcname, k, v, 0);
 372         if (!strcmp(type, "xfuncname"))
 373                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 374         if (!strcmp(type, "binary"))
 375                 return parse_tristate(&drv->binary, k, v);
 376         if (!strcmp(type, "command"))
 377                 return git_config_string(&drv->external, k, v);
 378         if (!strcmp(type, "textconv"))
 379                 return git_config_string(&drv->textconv, k, v);
 380         if (!strcmp(type, "cachetextconv"))
 381                 return parse_bool(&drv->textconv_want_cache, k, v);
 382         if (!strcmp(type, "wordregex"))
 383                 return git_config_string(&drv->word_regex, k, v);
 384
 385         return 0;
 386 }
 387
 388 struct userdiff_driver *userdiff_find_by_name(const char *name)
 389 {
 390         int len = strlen(name);
 391         return userdiff_find_by_namelen(name, len);
 392 }
 393
 394 struct userdiff_driver *userdiff_find_by_path(struct index_state *istate,
 395                                               const char *path)
 396 {
 397         static struct attr_check *check;
 398
 399         if (!check)
 400                 check = attr_check_initl("diff", NULL);
 401         if (!path)
 402                 return NULL;
 403         git_check_attr(istate, path, check);
 404
 405         if (ATTR_TRUE(check->items[0].value))
 406                 return &driver_true;
 407         if (ATTR_FALSE(check->items[0].value))
 408                 return &driver_false;
 409         if (ATTR_UNSET(check->items[0].value))
 410                 return NULL;
 411         return userdiff_find_by_name(check->items[0].value);
 412 }
 413
 414 struct userdiff_driver *userdiff_get_textconv(struct repository *r,
 415                                               struct userdiff_driver *driver)
 416 {
 417         if (!driver->textconv)
 418                 return NULL;
 419
 420         if (driver->textconv_want_cache && !driver->textconv_cache) {
 421                 struct notes_cache *c = xmalloc(sizeof(*c));
 422                 struct strbuf name = STRBUF_INIT;
 423
 424                 strbuf_addf(&name, "textconv/%s", driver->name);
 425                 notes_cache_init(r, c, name.buf, driver->textconv);
 426                 driver->textconv_cache = c;
 427                 strbuf_release(&name);
 428         }
 429
 430         return driver;
 431 }
 432
 433 static int for_each_userdiff_driver_list(each_userdiff_driver_fn fn,
 434                                          enum userdiff_driver_type type, void *cb_data,
 435                                          struct userdiff_driver *drv,
 436                                          int drv_size)
 437 {
 438         int i;
 439         int ret;
 440         for (i = 0; i < drv_size; i++) {
 441                 struct userdiff_driver *item = drv + i;
 442                 if ((ret = fn(item, type, cb_data)))
 443                         return ret;
 444         }
 445         return 0;
 446 }
 447
 448 int for_each_userdiff_driver(each_userdiff_driver_fn fn, void *cb_data)
 449 {
 450         int ret;
 451
 452         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_CUSTOM,
 453                                             cb_data, drivers, ndrivers);
 454         if (ret)
 455                 return ret;
 456
 457         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_BUILTIN,
 458                                             cb_data, builtin_drivers,
 459                                             ARRAY_SIZE(builtin_drivers));
 460         if (ret)
 461                 return ret;
 462
 463         return 0;
 464 }