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