Andrew's git / gitweb.git / userdiff.c
?
userdiff.con commit ref-filter.c: find disjoint pattern prefixes (b31e268)
   1#include "cache.h"
   2#include "config.h"
   3#include "userdiff.h"
   4#include "attr.h"
   5
   6static struct userdiff_driver *drivers;
   7static int ndrivers;
   8static 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]+" }
  16static struct userdiff_driver builtin_drivers[] = {
  17IPATTERN("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         "|=>|\\.\\.|\\*\\*|:=|/=|>=|<=|<<|>>|<>"),
  26IPATTERN("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].*)$",
  31         /* -- */
  32         "[a-zA-Z][a-zA-Z0-9_]*"
  33         "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
  34         /* numbers and format statements like 2E14.4, or ES12.6, 9X.
  35          * Don't worry about format statements without leading digits since
  36          * they would have been matched above as a variable anyway. */
  37         "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
  38         "|//|\\*\\*|::|[/<>=]="),
  39IPATTERN("fountain", "^((\\.[^.]|(int|ext|est|int\\.?/ext|i/e)[. ]).*)$",
  40         "[^ \t-]+"),
  41PATTERNS("golang",
  42         /* Functions */
  43         "^[ \t]*(func[ \t]*.*(\\{[ \t]*)?)\n"
  44         /* Structs and interfaces */
  45         "^[ \t]*(type[ \t].*(struct|interface)[ \t]*(\\{[ \t]*)?)",
  46         /* -- */
  47         "[a-zA-Z_][a-zA-Z0-9_]*"
  48         "|[-+0-9.eE]+i?|0[xX]?[0-9a-fA-F]+i?"
  49         "|[-+*/<>%&^|=!:]=|--|\\+\\+|<<=?|>>=?|&\\^=?|&&|\\|\\||<-|\\.{3}"),
  50PATTERNS("html", "^[ \t]*(<[Hh][1-6]([ \t].*)?>.*)$",
  51         "[^<>= \t]+"),
  52PATTERNS("java",
  53         "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
  54         "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
  55         /* -- */
  56         "[a-zA-Z_][a-zA-Z0-9_]*"
  57         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  58         "|[-+*/<>%&^|=!]="
  59         "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
  60PATTERNS("matlab",
  61         /*
  62          * Octave pattern is mostly the same as matlab, except that '%%%' and
  63          * '##' can also be used to begin code sections, in addition to '%%'
  64          * that is understood by both.
  65          */
  66         "^[[:space:]]*((classdef|function)[[:space:]].*)$|^(%%%?|##)[[:space:]].*$",
  67         "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
  68PATTERNS("objc",
  69         /* Negate C statements that can look like functions */
  70         "!^[ \t]*(do|for|if|else|return|switch|while)\n"
  71         /* Objective-C methods */
  72         "^[ \t]*([-+][ \t]*\\([ \t]*[A-Za-z_][A-Za-z_0-9* \t]*\\)[ \t]*[A-Za-z_].*)$\n"
  73         /* C functions */
  74         "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$\n"
  75         /* Objective-C class/protocol definitions */
  76         "^(@(implementation|interface|protocol)[ \t].*)$",
  77         /* -- */
  78         "[a-zA-Z_][a-zA-Z0-9_]*"
  79         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
  80         "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
  81PATTERNS("pascal",
  82         "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|"
  83                "implementation|initialization|finalization)[ \t]*.*)$"
  84         "\n"
  85         "^(.*=[ \t]*(class|record).*)$",
  86         /* -- */
  87         "[a-zA-Z_][a-zA-Z0-9_]*"
  88         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
  89         "|<>|<=|>=|:=|\\.\\."),
  90PATTERNS("perl",
  91         "^package .*\n"
  92         "^sub [[:alnum:]_':]+[ \t]*"
  93                "(\\([^)]*\\)[ \t]*)?" /* prototype */
  94                /*
  95                 * Attributes.  A regex can't count nested parentheses,
  96                 * so just slurp up whatever we see, taking care not
  97                 * to accept lines like "sub foo; # defined elsewhere".
  98                 *
  99                 * An attribute could contain a semicolon, but at that
 100                 * point it seems reasonable enough to give up.
 101                 */
 102                "(:[^;#]*)?"
 103                "(\\{[ \t]*)?" /* brace can come here or on the next line */
 104                "(#.*)?$\n" /* comment */
 105         "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
 106                "(\\{[ \t]*)?" /* brace can come here or on the next line */
 107                "(#.*)?$\n"
 108         "^=head[0-9] .*",      /* POD */
 109         /* -- */
 110         "[[:alpha:]_'][[:alnum:]_']*"
 111         "|0[xb]?[0-9a-fA-F_]*"
 112         /* taking care not to interpret 3..5 as (3.)(.5) */
 113         "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
 114         "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
 115         "|&&=|\\|\\|=|//=|\\*\\*="
 116         "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
 117         "|[-+*/%.^&<>=!|]="
 118         "|=~|!~"
 119         "|<<|<>|<=>|>>"),
 120PATTERNS("php",
 121         "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
 122         "^[\t ]*((((final|abstract)[\t ]+)?class|interface|trait).*)$",
 123         /* -- */
 124         "[a-zA-Z_][a-zA-Z0-9_]*"
 125         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
 126         "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 127PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
 128         /* -- */
 129         "[a-zA-Z_][a-zA-Z0-9_]*"
 130         "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
 131         "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
 132         /* -- */
 133PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
 134         /* -- */
 135         "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
 136         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
 137         "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 138PATTERNS("rust",
 139         "^[\t ]*((pub(\\([^\\)]+\\))?[\t ]+)?((async|const|unsafe|extern([\t ]+\"[^\"]+\"))[\t ]+)?(struct|enum|union|mod|trait|fn|impl)[< \t]+[^;]*)$",
 140         /* -- */
 141         "[a-zA-Z_][a-zA-Z0-9_]*"
 142         "|[0-9][0-9_a-fA-Fiosuxz]*(\\.([0-9]*[eE][+-]?)?[0-9_fF]*)?"
 143         "|[-+*\\/<>%&^|=!:]=|<<=?|>>=?|&&|\\|\\||->|=>|\\.{2}=|\\.{3}|::"),
 144PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
 145         "[={}\"]|[^={}\" \t]+"),
 146PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
 147         "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 148PATTERNS("cpp",
 149         /* Jump targets or access declarations */
 150         "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:[[:space:]]*($|/[/*])\n"
 151         /* functions/methods, variables, and compounds at top level */
 152         "^((::[[:space:]]*)?[A-Za-z_].*)$",
 153         /* -- */
 154         "[a-zA-Z_][a-zA-Z0-9_]*"
 155         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lLuU]*"
 156         "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->\\*?|\\.\\*"),
 157PATTERNS("csharp",
 158         /* Keywords */
 159         "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
 160         /* Methods and constructors */
 161         "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe|async)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[<>@._[:alnum:]]+[ \t]*\\(.*\\))[ \t]*$\n"
 162         /* Properties */
 163         "^[ \t]*(((static|public|internal|private|protected|new|virtual|sealed|override|unsafe)[ \t]+)*[][<>@.~_[:alnum:]]+[ \t]+[@._[:alnum:]]+)[ \t]*$\n"
 164         /* Type definitions */
 165         "^[ \t]*(((static|public|internal|private|protected|new|unsafe|sealed|abstract|partial)[ \t]+)*(class|enum|interface|struct)[ \t]+.*)$\n"
 166         /* Namespace */
 167         "^[ \t]*(namespace[ \t]+.*)$",
 168         /* -- */
 169         "[a-zA-Z_][a-zA-Z0-9_]*"
 170         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
 171         "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 172IPATTERN("css",
 173         "![:;][[:space:]]*$\n"
 174         "^[_a-z0-9].*$",
 175         /* -- */
 176         /*
 177          * This regex comes from W3C CSS specs. Should theoretically also
 178          * allow ISO 10646 characters U+00A0 and higher,
 179          * but they are not handled in this regex.
 180          */
 181         "-?[_a-zA-Z][-_a-zA-Z0-9]*" /* identifiers */
 182         "|-?[0-9]+|\\#[0-9a-fA-F]+" /* numbers */
 183),
 184{ "default", NULL, -1, { NULL, 0 } },
 185};
 186#undef PATTERNS
 187#undef IPATTERN
 188
 189static struct userdiff_driver driver_true = {
 190        "diff=true",
 191        NULL,
 192        0,
 193        { NULL, 0 }
 194};
 195
 196static struct userdiff_driver driver_false = {
 197        "!diff",
 198        NULL,
 199        1,
 200        { NULL, 0 }
 201};
 202
 203static struct userdiff_driver *userdiff_find_by_namelen(const char *k, int len)
 204{
 205        int i;
 206        for (i = 0; i < ndrivers; i++) {
 207                struct userdiff_driver *drv = drivers + i;
 208                if (!strncmp(drv->name, k, len) && !drv->name[len])
 209                        return drv;
 210        }
 211        for (i = 0; i < ARRAY_SIZE(builtin_drivers); i++) {
 212                struct userdiff_driver *drv = builtin_drivers + i;
 213                if (!strncmp(drv->name, k, len) && !drv->name[len])
 214                        return drv;
 215        }
 216        return NULL;
 217}
 218
 219static int parse_funcname(struct userdiff_funcname *f, const char *k,
 220                const char *v, int cflags)
 221{
 222        if (git_config_string(&f->pattern, k, v) < 0)
 223                return -1;
 224        f->cflags = cflags;
 225        return 0;
 226}
 227
 228static int parse_tristate(int *b, const char *k, const char *v)
 229{
 230        if (v && !strcasecmp(v, "auto"))
 231                *b = -1;
 232        else
 233                *b = git_config_bool(k, v);
 234        return 0;
 235}
 236
 237static int parse_bool(int *b, const char *k, const char *v)
 238{
 239        *b = git_config_bool(k, v);
 240        return 0;
 241}
 242
 243int userdiff_config(const char *k, const char *v)
 244{
 245        struct userdiff_driver *drv;
 246        const char *name, *type;
 247        int namelen;
 248
 249        if (parse_config_key(k, "diff", &name, &namelen, &type) || !name)
 250                return 0;
 251
 252        drv = userdiff_find_by_namelen(name, namelen);
 253        if (!drv) {
 254                ALLOC_GROW(drivers, ndrivers+1, drivers_alloc);
 255                drv = &drivers[ndrivers++];
 256                memset(drv, 0, sizeof(*drv));
 257                drv->name = xmemdupz(name, namelen);
 258                drv->binary = -1;
 259        }
 260
 261        if (!strcmp(type, "funcname"))
 262                return parse_funcname(&drv->funcname, k, v, 0);
 263        if (!strcmp(type, "xfuncname"))
 264                return parse_funcname(&drv->funcname, k, v, REG_EXTENDED);
 265        if (!strcmp(type, "binary"))
 266                return parse_tristate(&drv->binary, k, v);
 267        if (!strcmp(type, "command"))
 268                return git_config_string(&drv->external, k, v);
 269        if (!strcmp(type, "textconv"))
 270                return git_config_string(&drv->textconv, k, v);
 271        if (!strcmp(type, "cachetextconv"))
 272                return parse_bool(&drv->textconv_want_cache, k, v);
 273        if (!strcmp(type, "wordregex"))
 274                return git_config_string(&drv->word_regex, k, v);
 275
 276        return 0;
 277}
 278
 279struct userdiff_driver *userdiff_find_by_name(const char *name)
 280{
 281        int len = strlen(name);
 282        return userdiff_find_by_namelen(name, len);
 283}
 284
 285struct userdiff_driver *userdiff_find_by_path(struct index_state *istate,
 286                                              const char *path)
 287{
 288        static struct attr_check *check;
 289
 290        if (!check)
 291                check = attr_check_initl("diff", NULL);
 292        if (!path)
 293                return NULL;
 294        git_check_attr(istate, path, check);
 295
 296        if (ATTR_TRUE(check->items[0].value))
 297                return &driver_true;
 298        if (ATTR_FALSE(check->items[0].value))
 299                return &driver_false;
 300        if (ATTR_UNSET(check->items[0].value))
 301                return NULL;
 302        return userdiff_find_by_name(check->items[0].value);
 303}
 304
 305struct userdiff_driver *userdiff_get_textconv(struct repository *r,
 306                                              struct userdiff_driver *driver)
 307{
 308        if (!driver->textconv)
 309                return NULL;
 310
 311        if (driver->textconv_want_cache && !driver->textconv_cache) {
 312                struct notes_cache *c = xmalloc(sizeof(*c));
 313                struct strbuf name = STRBUF_INIT;
 314
 315                strbuf_addf(&name, "textconv/%s", driver->name);
 316                notes_cache_init(r, c, name.buf, driver->textconv);
 317                driver->textconv_cache = c;
 318                strbuf_release(&name);
 319        }
 320
 321        return driver;
 322}