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