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