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