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