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