userdiff.c

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