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