userdiff.c

   1 #define USE_THE_REPOSITORY_VARIABLE
   2 #define DISABLE_SIGN_COMPARE_WARNINGS
   3
   4 #include "git-compat-util.h"
   5 #include "config.h"
   6 #include "userdiff.h"
   7 #include "attr.h"
   8 #include "strbuf.h"
   9 #include "environment.h"
  10
  11 static struct userdiff_driver *drivers;
  12 static int ndrivers;
  13 static int drivers_alloc;
  14
  15 #define PATTERNS(lang, rx, wrx) { \
  16         .name = lang, \
  17         .binary = -1, \
  18         .funcname = { \
  19                 .pattern = rx, \
  20                 .cflags = REG_EXTENDED, \
  21         }, \
  22         .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \
  23         .word_regex_multi_byte = wrx "|[^[:space:]]", \
  24 }
  25 #define IPATTERN(lang, rx, wrx) { \
  26         .name = lang, \
  27         .binary = -1, \
  28         .funcname = { \
  29                 .pattern = rx, \
  30                 .cflags = REG_EXTENDED | REG_ICASE, \
  31         }, \
  32         .word_regex = wrx "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+", \
  33         .word_regex_multi_byte = wrx "|[^[:space:]]", \
  34 }
  35
  36 /*
  37  * Built-in drivers for various languages, sorted by their names
  38  * (except that the "default" is left at the end).
  39  *
  40  * When writing or updating patterns, assume that the contents these
  41  * patterns are applied to are syntactically correct.  The patterns
  42  * can be simple without implementing all syntactical corner cases, as
  43  * long as they are sufficiently permissive.
  44  */
  45 static struct userdiff_driver builtin_drivers[] = {
  46 IPATTERN("ada",
  47          "!^(.*[ \t])?(is[ \t]+new|renames|is[ \t]+separate)([ \t].*)?$\n"
  48          "!^[ \t]*with[ \t].*$\n"
  49          "^[ \t]*((procedure|function)[ \t]+.*)$\n"
  50          "^[ \t]*((package|protected|task)[ \t]+.*)$",
  51          /* -- */
  52          "[a-zA-Z][a-zA-Z0-9_]*"
  53          "|[-+]?[0-9][0-9#_.aAbBcCdDeEfF]*([eE][+-]?[0-9_]+)?"
  54          "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  55 PATTERNS("bash",
  56          /* Optional leading indentation */
  57          "^[ \t]*"
  58          /* Start of captured text */
  59          "("
  60          "("
  61              /* POSIX identifier with mandatory parentheses */
  62              "[a-zA-Z_][a-zA-Z0-9_]*[ \t]*\\([ \t]*\\))"
  63          "|"
  64              /* Bashism identifier with optional parentheses */
  65              "(function[ \t]+[a-zA-Z_][a-zA-Z0-9_]*(([ \t]*\\([ \t]*\\))|([ \t]+))"
  66          ")"
  67          /* Optional whitespace */
  68          "[ \t]*"
  69          /* Compound command starting with `{`, `(`, `((` or `[[` */
  70          "(\\{|\\(\\(?|\\[\\[)"
  71          /* End of captured text */
  72          ")",
  73          /* -- */
  74          /* Characters not in the default $IFS value */
  75          "[^ \t]+"),
  76 PATTERNS("bibtex",
  77          "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
  78          /* -- */
  79          "[={}\"]|[^={}\" \t]+"),
  80 PATTERNS("cpp",
  81          /* Jump targets or access declarations */
  82          "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
  83          /* functions/methods, variables, and compounds at top level */
  84          "^((::[[:space:]]*)?[A-Za-z_].*)$",
  85          /* -- */
  86          /* identifiers and keywords */
  87          "[a-zA-Z_][a-zA-Z0-9_]*"
  88          /* decimal and octal integers as well as floatingpoint numbers */
  89          "|[0-9][0-9.]*([Ee][-+]?[0-9]+)?[fFlLuU]*"
  90          /* hexadecimal and binary integers */
  91          "|0[xXbB][0-9a-fA-F]+[lLuU]*"
  92          /* floatingpoint numbers that begin with a decimal point */
  93          "|\\.[0-9][0-9]*([Ee][-+]?[0-9]+)?[fFlL]?"
  94          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*|<=>"),
  95 PATTERNS("csharp",
  96          /*
  97           * Jump over reserved keywords which are illegal method names, but which
  98           * can be followed by parentheses without special characters in between,
  99           * making them look like methods.
 100           */
 101          "!(^|[ \t]+)" /* Start of line or whitespace. */
 102                 "(do|while|for|foreach|if|else|new|default|return|switch|case|throw"
 103                 "|catch|using|lock|fixed)"
 104                 "([ \t(]+|$)\n" /* Whitespace, "(", or end of line. */
 105          /*
 106           * Methods/constructors:
 107           * The strategy is to identify a minimum of two groups (any combination
 108           * of keywords/type/name) before the opening parenthesis, and without
 109           * final unexpected characters, normally only used in ordinary statements.
 110           */
 111          "^[ \t]*" /* Remove leading whitespace. */
 112                 "(" /* Start chunk header capture. */
 113                 "(" /* First group. */
 114                         "[][[:alnum:]@_.]" /* Name. */
 115                         "(<[][[:alnum:]@_, \t<>]+>)?" /* Optional generic parameters. */
 116                 ")+"
 117                 "([ \t]+" /* Subsequent groups, prepended with space. */
 118                         "([][[:alnum:]@_.](<[][[:alnum:]@_, \t<>]+>)?)+"
 119                 ")+"
 120                 "[ \t]*" /* Optional space before parameters start. */
 121                 "\\(" /* Start of method parameters. */
 122                 "[^;]*" /* Allow complex parameters, but exclude statements (;). */
 123                 ")$\n" /* Close chunk header capture. */
 124          /*
 125           * Properties:
 126           * As with methods, expect a minimum of two groups. But, more trivial than
 127           * methods, the vast majority of properties long enough to be worth
 128           * showing a chunk header for don't include "=:;,()" on the line they are
 129           * defined, since they don't have a parameter list.
 130           */
 131          "^[ \t]*("
 132                 "([][[:alnum:]@_.](<[][[:alnum:]@_, \t<>]+>)?)+"
 133                 "([ \t]+"
 134                         "([][[:alnum:]@_.](<[][[:alnum:]@_, \t<>]+>)?)+"
 135                 ")+" /* Up to here, same as methods regex. */
 136                 "[^;=:,()]*" /* Compared to methods, no parameter list allowed. */
 137                 ")$\n"
 138          /* Type definitions */
 139          "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct|record)[ \t]+.*)$\n"
 140          /* Namespace */
 141          "^[ \t]*(namespace[ \t]+.*)$",
 142          /* -- */
 143          "[a-zA-Z_][a-zA-Z0-9_]*"
 144          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 145          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 146 IPATTERN("css",
 147          "![:;][[:space:]]*$\n"
 148          "^[:[@.#]?[_a-z0-9].*$",
 149          /* -- */
 150          /*
 151           * This regex comes from W3C CSS specs. Should theoretically also
 152           * allow ISO 10646 characters U+00A0 and higher,
 153           * but they are not handled in this regex.
 154           */
 155          "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */
 156          "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */
 157 ),
 158 PATTERNS("dts",
 159          "!;\n"
 160          "!=\n"
 161          /* lines beginning with a word optionally preceded by '&' or the root */
 162          "^[ \t]*((/[ \t]*\\{|&?[a-zA-Z_]).*)",
 163          /* -- */
 164          /* Property names and math operators */
 165          "[a-zA-Z0-9,._+?#-]+"
 166          "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"),
 167 PATTERNS("elixir",
 168          "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$",
 169          /* -- */
 170          /* Atoms, names, and module attributes */
 171          "[@:]?[a-zA-Z0-9@_?!]+"
 172          /* Numbers with specific base */
 173          "|[-+]?0[xob][0-9a-fA-F]+"
 174          /* Numbers */
 175          "|[-+]?[0-9][0-9_.]*([eE][-+]?[0-9_]+)?"
 176          /* Operators and atoms that represent them */
 177          "|:?(\\+\\+|--|\\.\\.|~~~|<>|\\^\\^\\^|<?\\|>|<<<?|>?>>|<<?~|~>?>|<~>|<=|>=|===?|!==?|=~|&&&?|\\|\\|\\|?|=>|<-|\\\\\\\\|->)"
 178          /* Not real operators, but should be grouped */
 179          "|:?%[A-Za-z0-9_.]\\{\\}?"),
 180 IPATTERN("fortran",
 181          /* Don't match comment lines */
 182          "!^([C*]|[ \t]*!)\n"
 183          /* Don't match 'module procedure' lines */
 184          "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
 185          /* Program, module, block data */
 186          "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
 187                 /* Subroutines and functions */
 188                 "|([^!'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
 189          /* -- */
 190          "[a-zA-Z][a-zA-Z0-9_]*"
 191          "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
 192          /* numbers and format statements like 2E14.4, or ES12.6, 9X.
 193           * Don't worry about format statements without leading digits since
 194           * they would have been matched above as a variable anyway. */
 195          "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
 196          "|//|\\*\\*|::|[/<>=]="),
 197 IPATTERN("fountain",
 198          "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$",
 199          /* -- */
 200          "[^ \t-]+"),
 201 PATTERNS("golang",
 202          /* Functions */
 203          "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n"
 204          /* Structs and interfaces */
 205          "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)",
 206          /* -- */
 207          "[a-zA-Z_][a-zA-Z0-9_]*"
 208          "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?"
 209          "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"),
 210 PATTERNS("html",
 211          "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$",
 212          /* -- */
 213          "[^<>= \t]+"),
 214 PATTERNS("java",
 215          "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
 216          /* Class, enum, interface, and record declarations */
 217          "^[ \t]*(([a-z-]+[ \t]+)*(class|enum|interface|record)[ \t]+.*)$\n"
 218          /* Method definitions; note that constructor signatures are not */
 219          /* matched because they are indistinguishable from method calls. */
 220          "^[ \t]*(([A-Za-z_<>&][][?&<>.,A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
 221          /* -- */
 222          "[a-zA-Z_][a-zA-Z0-9_]*"
 223          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 224          "|[-+*/<>%&^|=!]="
 225          "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
 226 PATTERNS("kotlin",
 227          "^[ \t]*(([a-z]+[ \t]+)*(fun|class|interface)[ \t]+.*)$",
 228          /* -- */
 229          "[a-zA-Z_][a-zA-Z0-9_]*"
 230          /* hexadecimal and binary numbers */
 231          "|0[xXbB][0-9a-fA-F_]+[lLuU]*"
 232          /* integers and floats */
 233          "|[0-9][0-9_]*([.][0-9_]*)?([Ee][-+]?[0-9]+)?[fFlLuU]*"
 234          /* floating point numbers beginning with decimal point */
 235          "|[.][0-9][0-9_]*([Ee][-+]?[0-9]+)?[fFlLuU]?"
 236          /* unary and binary operators */
 237          "|[-+*/<>%&^|=!]==?|--|\\+\\+|<<=|>>=|&&|\\|\\||->|\\.\\*|!!|[?:.][.:]"),
 238 PATTERNS("markdown",
 239          "^ {0,3}#{1,6}[ \t].*",
 240          /* -- */
 241          "[^<>= \t]+"),
 242 PATTERNS("matlab",
 243          /*
 244           * Octave pattern is mostly the same as matlab, except that '%%%' and
 245           * '##' can also be used to begin code sections, in addition to '%%'
 246           * that is understood by both.
 247           */
 248          "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$",
 249          /* -- */
 250          "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
 251 PATTERNS("objc",
 252          /* Negate C statements that can look like functions */
 253          "!^[ \t]*(do|for|if|else|return|switch|while)\n"
 254          /* Objective-C methods */
 255          "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
 256          /* C functions */
 257          "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
 258          /* Objective-C class/protocol definitions */
 259          "^(@(implementation|interface|protocol)[ \t].*)$",
 260          /* -- */
 261          "[a-zA-Z_][a-zA-Z0-9_]*"
 262          "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 263          "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 264 PATTERNS("pascal",
 265          "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface"
 266          "|implementation|initialization|finalization)[ \t]*.*)$\n"
 267          "^(.*=[ \t]*(class|record).*)$",
 268          /* -- */
 269          "[a-zA-Z_][a-zA-Z0-9_]*"
 270          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 271          "|<>|<=|>=|:=|\\.\\."),
 272 PATTERNS("perl",
 273          "^package .*\n"
 274          "^sub [[:alnum:]_':]+[ \t]*"
 275                 "(\\([^)]*\\)[ \t]*)?" /* prototype */
 276                 /*
 277                  * Attributes.  A regex can't count nested parentheses,
 278                  * so just slurp up whatever we see, taking care not
 279                  * to accept lines like "sub foo; # defined elsewhere".
 280                  *
 281                  * An attribute could contain a semicolon, but at that
 282                  * point it seems reasonable enough to give up.
 283                  */
 284                 "(:[^;#]*)?"
 285                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 286                 "(#.*)?$\n" /* comment */
 287          "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
 288                 "(\\{[ \t]*)?" /* brace can come here or on the next line */
 289                 "(#.*)?$\n"
 290          "^=head[0-9] .*",      /* POD */
 291          /* -- */
 292          "[[:alpha:]_'][[:alnum:]_']*"
 293          "|0[xb]?[0-9a-fA-F_]*"
 294          /* taking care not to interpret 3..5 as (3.)(.5) */
 295          "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
 296          "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
 297          "|&&=|\\|\\|=|//=|\\*\\*="
 298          "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 299          "|[-+*/%.^&<>=!|]="
 300          "|=~|!~"
 301          "|<<|<>|<=>|>>"),
 302 PATTERNS("php",
 303          "^[\t ]*(((public|protected|private|static|abstract|final)[\t ]+)*function.*)$\n"
 304          "^[\t ]*((((final|abstract)[\t ]+)?class|enum|interface|trait).*)$",
 305          /* -- */
 306          "[a-zA-Z_][a-zA-Z0-9_]*"
 307          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 308          "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 309 PATTERNS("python",
 310          "^[ \t]*((class|(async[ \t]+)?def)[ \t].*)$",
 311          /* -- */
 312          "[a-zA-Z_][a-zA-Z0-9_]*"
 313          "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 314          "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 315          /* -- */
 316 PATTERNS("ruby",
 317          "^[ \t]*((class|module|def)[ \t].*)$",
 318          /* -- */
 319          "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 320          "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 321          "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 322 PATTERNS("rust",
 323          "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl|macro_rules!)[< \t]+[^;]*)$",
 324          /* -- */
 325          "[a-zA-Z_][a-zA-Z0-9_]*"
 326          "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?"
 327          "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"),
 328 PATTERNS("scheme",
 329          "^[\t ]*(\\(((define|def(struct|syntax|class|method|rules|record|proto|alias)?)[-*/ \t]|(library|module|struct|class)[*+ \t]).*)$",
 330          /*
 331           * R7RS valid identifiers include any sequence enclosed
 332           * within vertical lines having no backslashes
 333           */
 334          "\\|([^\\\\]*)\\|"
 335          /* All other words should be delimited by spaces or parentheses */
 336          "|([^][)(}{[ \t])+"),
 337 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 338          "\\\\[a-zA-Z@]+|\\\\.|([a-zA-Z0-9]|[^\x01-\x7f])+"),
 339 { .name = "default", .binary = -1 },
 340 };
 341 #undef PATTERNS
 342 #undef IPATTERN
 343
 344 static struct userdiff_driver driver_true = {
 345         .name = "diff=true",
 346         .binary = 0,
 347 };
 348
 349 static struct userdiff_driver driver_false = {
 350         .name = "!diff",
 351         .binary = 1,
 352 };
 353
 354 struct find_by_namelen_data {
 355         const char *name;
 356         size_t len;
 357         struct userdiff_driver *driver;
 358 };
 359
 360 static int userdiff_find_by_namelen_cb(struct userdiff_driver *driver,
 361                                        enum userdiff_driver_type type UNUSED,
 362                                        void *priv)
 363 {
 364         struct find_by_namelen_data *cb_data = priv;
 365
 366         if (!xstrncmpz(driver->name, cb_data->name, cb_data->len)) {
 367                 cb_data->driver = driver;
 368                 return 1; /* tell the caller to stop iterating */
 369         }
 370         return 0;
 371 }
 372
 373 static int regexec_supports_multi_byte_chars(void)
 374 {
 375         static const char not_space[] = "[^[:space:]]";
 376         static const char utf8_multi_byte_char[] = "\xc2\xa3";
 377         regex_t re;
 378         regmatch_t match;
 379         static int result = -1;
 380
 381         if (result != -1)
 382                 return result;
 383         if (regcomp(&re, not_space, REG_EXTENDED))
 384                 BUG("invalid regular expression: %s", not_space);
 385         result = !regexec(&re, utf8_multi_byte_char, 1, &match, 0) &&
 386                 match.rm_so == 0 &&
 387                 match.rm_eo == strlen(utf8_multi_byte_char);
 388         regfree(&re);
 389         return result;
 390 }
 391
 392 static struct userdiff_driver *userdiff_find_by_namelen(const char *name, size_t len)
 393 {
 394         struct find_by_namelen_data udcbdata = {
 395                 .name = name,
 396                 .len = len,
 397         };
 398         for_each_userdiff_driver(userdiff_find_by_namelen_cb, &udcbdata);
 399         return udcbdata.driver;
 400 }
 401
 402 static int parse_funcname(struct userdiff_funcname *f, const char *k,
 403                 const char *v, int cflags)
 404 {
 405         f->pattern = NULL;
 406         FREE_AND_NULL(f->pattern_owned);
 407         if (git_config_string(&f->pattern_owned, k, v) < 0)
 408                 return -1;
 409         f->pattern = f->pattern_owned;
 410         f->cflags = cflags;
 411         return 0;
 412 }
 413
 414 static int parse_tristate(int *b, const char *k, const char *v)
 415 {
 416         if (v && !strcasecmp(v, "auto"))
 417                 *b = -1;
 418         else
 419                 *b = git_config_bool(k, v);
 420         return 0;
 421 }
 422
 423 static int parse_bool(int *b, const char *k, const char *v)
 424 {
 425         *b = git_config_bool(k, v);
 426         return 0;
 427 }
 428
 429 int userdiff_config(const char *k, const char *v)
 430 {
 431         struct userdiff_driver *drv;
 432         const char *name, *type;
 433         size_t namelen;
 434
 435         if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
 436                 return 0;
 437
 438         drv = userdiff_find_by_namelen(name, namelen);
 439         if (!drv) {
 440                 ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 441                 drv = &drivers[ndrivers++];
 442                 memset(drv, 0, sizeof(*drv));
 443                 drv->name = xmemdupz(name, namelen);
 444                 drv->binary = -1;
 445         }
 446
 447         if (!strcmp(type, "funcname"))
 448                 return parse_funcname(&drv->funcname, k, v, 0);
 449         if (!strcmp(type, "xfuncname"))
 450                 return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 451         if (!strcmp(type, "binary"))
 452                 return parse_tristate(&drv->binary, k, v);
 453         if (!strcmp(type, "command")) {
 454                 FREE_AND_NULL(drv->external.cmd);
 455                 return git_config_string(&drv->external.cmd, k, v);
 456         }
 457         if (!strcmp(type, "trustexitcode")) {
 458                 drv->external.trust_exit_code = git_config_bool(k, v);
 459                 return 0;
 460         }
 461         if (!strcmp(type, "textconv")) {
 462                 int ret;
 463                 FREE_AND_NULL(drv->textconv_owned);
 464                 ret = git_config_string(&drv->textconv_owned, k, v);
 465                 drv->textconv = drv->textconv_owned;
 466                 return ret;
 467         }
 468         if (!strcmp(type, "cachetextconv"))
 469                 return parse_bool(&drv->textconv_want_cache, k, v);
 470         if (!strcmp(type, "wordregex")) {
 471                 int ret;
 472                 FREE_AND_NULL(drv->word_regex_owned);
 473                 ret = git_config_string(&drv->word_regex_owned, k, v);
 474                 drv->word_regex = drv->word_regex_owned;
 475                 return ret;
 476         }
 477         if (!strcmp(type, "algorithm")) {
 478                 int ret;
 479                 FREE_AND_NULL(drv->algorithm_owned);
 480                 ret = git_config_string(&drv->algorithm_owned, k, v);
 481                 drv->algorithm = drv->algorithm_owned;
 482                 return ret;
 483         }
 484
 485         return 0;
 486 }
 487
 488 struct userdiff_driver *userdiff_find_by_name(const char *name)
 489 {
 490         int len = strlen(name);
 491         struct userdiff_driver *driver = userdiff_find_by_namelen(name, len);
 492         if (driver && driver->word_regex_multi_byte) {
 493                 if (regexec_supports_multi_byte_chars())
 494                         driver->word_regex = driver->word_regex_multi_byte;
 495                 driver->word_regex_multi_byte = NULL;
 496         }
 497         return driver;
 498 }
 499
 500 struct userdiff_driver *userdiff_find_by_path(struct index_state *istate,
 501                                               const char *path)
 502 {
 503         static struct attr_check *check;
 504
 505         if (!check)
 506                 check = attr_check_initl("diff", NULL);
 507         if (!path)
 508                 return NULL;
 509         git_check_attr(istate, path, check);
 510
 511         if (ATTR_TRUE(check->items[0].value))
 512                 return &driver_true;
 513         if (ATTR_FALSE(check->items[0].value))
 514                 return &driver_false;
 515         if (ATTR_UNSET(check->items[0].value))
 516                 return NULL;
 517         return userdiff_find_by_name(check->items[0].value);
 518 }
 519
 520 struct userdiff_driver *userdiff_get_textconv(struct repository *r,
 521                                               struct userdiff_driver *driver)
 522 {
 523         if (!driver->textconv)
 524                 return NULL;
 525
 526         if (driver->textconv_want_cache && !driver->textconv_cache &&
 527             have_git_dir()) {
 528                 struct notes_cache *c = xmalloc(sizeof(*c));
 529                 struct strbuf name = STRBUF_INIT;
 530
 531                 strbuf_addf(&name, "textconv/%s", driver->name);
 532                 notes_cache_init(r, c, name.buf, driver->textconv);
 533                 driver->textconv_cache = c;
 534                 strbuf_release(&name);
 535         }
 536
 537         return driver;
 538 }
 539
 540 static int for_each_userdiff_driver_list(each_userdiff_driver_fn fn,
 541                                          enum userdiff_driver_type type, void *cb_data,
 542                                          struct userdiff_driver *drv,
 543                                          int drv_size)
 544 {
 545         int i;
 546         int ret;
 547         for (i = 0; i < drv_size; i++) {
 548                 struct userdiff_driver *item = drv + i;
 549                 if ((ret = fn(item, type, cb_data)))
 550                         return ret;
 551         }
 552         return 0;
 553 }
 554
 555 int for_each_userdiff_driver(each_userdiff_driver_fn fn, void *cb_data)
 556 {
 557         int ret;
 558
 559         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_CUSTOM,
 560                                             cb_data, drivers, ndrivers);
 561         if (ret)
 562                 return ret;
 563
 564         ret = for_each_userdiff_driver_list(fn, USERDIFF_DRIVER_TYPE_BUILTIN,
 565                                             cb_data, builtin_drivers,
 566                                             ARRAY_SIZE(builtin_drivers));
 567         if (ret)
 568                 return ret;
 569
 570         return 0;
 571 }