Merge branch 'tr/userdiff-c-returns-pointer'
[git/git.git] / userdiff.c
index 16b5119..76109da 100644 (file)
@@ -8,10 +8,27 @@ static int ndrivers;
 static int drivers_alloc;
 
 #define PATTERNS(name, pattern, word_regex)                    \
-       { name, NULL, -1, { pattern, REG_EXTENDED }, word_regex }
+       { name, NULL, -1, { pattern, REG_EXTENDED },            \
+         word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
+#define IPATTERN(name, pattern, word_regex)                    \
+       { name, NULL, -1, { pattern, REG_EXTENDED | REG_ICASE }, \
+         word_regex "|[^[:space:]]|[\xc0-\xff][\x80-\xbf]+" }
 static struct userdiff_driver builtin_drivers[] = {
+IPATTERN("fortran",
+        "!^([C*]|[ \t]*!)\n"
+        "!^[ \t]*MODULE[ \t]+PROCEDURE[ \t]\n"
+        "^[ \t]*((END[ \t]+)?(PROGRAM|MODULE|BLOCK[ \t]+DATA"
+               "|([^'\" \t]+[ \t]+)*(SUBROUTINE|FUNCTION))[ \t]+[A-Z].*)$",
+        /* -- */
+        "[a-zA-Z][a-zA-Z0-9_]*"
+        "|\\.([Ee][Qq]|[Nn][Ee]|[Gg][TtEe]|[Ll][TtEe]|[Tt][Rr][Uu][Ee]|[Ff][Aa][Ll][Ss][Ee]|[Aa][Nn][Dd]|[Oo][Rr]|[Nn]?[Ee][Qq][Vv]|[Nn][Oo][Tt])\\."
+        /* numbers and format statements like 2E14.4, or ES12.6, 9X.
+         * Don't worry about format statements without leading digits since
+         * they would have been matched above as a variable anyway. */
+        "|[-+]?[0-9.]+([AaIiDdEeFfLlTtXx][Ss]?[-+]?[0-9.]*)?(_[a-zA-Z0-9][a-zA-Z0-9_]*)?"
+        "|//|\\*\\*|::|[/<>=]="),
 PATTERNS("html", "^[ \t]*(<[Hh][1-6][ \t].*>.*)$",
-        "[^<>= \t]+|[^[:space:]]|[\x80-\xff]+"),
+        "[^<>= \t]+"),
 PATTERNS("java",
         "!^[ \t]*(catch|do|for|if|instanceof|new|return|switch|throw|while)\n"
         "^[ \t]*(([A-Za-z_][A-Za-z_0-9]*[ \t]+)+[A-Za-z_][A-Za-z_0-9]*[ \t]*\\([^;]*)$",
@@ -19,8 +36,10 @@ PATTERNS("java",
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
         "|[-+*/<>%&^|=!]="
-        "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|--|\\+\\+|<<=?|>>>?=?|&&|\\|\\|"),
+PATTERNS("matlab",
+        "^[[:space:]]*((classdef|function)[[:space:]].*)$|^%%[[:space:]].*$",
+        "[a-zA-Z_][a-zA-Z0-9_]*|[-+0-9.e]+|[=~<>]=|\\.[*/\\^']|\\|\\||&&"),
 PATTERNS("objc",
         /* Negate C statements that can look like functions */
         "!^[ \t]*(do|for|if|else|return|switch|while)\n"
@@ -33,43 +52,68 @@ PATTERNS("objc",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
-        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 PATTERNS("pascal",
-        "^((procedure|function|constructor|destructor|interface|"
+        "^(((class[ \t]+)?(procedure|function)|constructor|destructor|interface|"
                "implementation|initialization|finalization)[ \t]*.*)$"
         "\n"
         "^(.*=[ \t]*(class|record).*)$",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
-        "|<>|<=|>=|:=|\\.\\."
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|<>|<=|>=|:=|\\.\\."),
+PATTERNS("perl",
+        "^package .*\n"
+        "^sub [[:alnum:]_':]+[ \t]*"
+               "(\\([^)]*\\)[ \t]*)?" /* prototype */
+               /*
+                * Attributes.  A regex can't count nested parentheses,
+                * so just slurp up whatever we see, taking care not
+                * to accept lines like "sub foo; # defined elsewhere".
+                *
+                * An attribute could contain a semicolon, but at that
+                * point it seems reasonable enough to give up.
+                */
+               "(:[^;#]*)?"
+               "(\\{[ \t]*)?" /* brace can come here or on the next line */
+               "(#.*)?$\n" /* comment */
+        "^(BEGIN|END|INIT|CHECK|UNITCHECK|AUTOLOAD|DESTROY)[ \t]*"
+               "(\\{[ \t]*)?" /* brace can come here or on the next line */
+               "(#.*)?$\n"
+        "^=head[0-9] .*",      /* POD */
+        /* -- */
+        "[[:alpha:]_'][[:alnum:]_']*"
+        "|0[xb]?[0-9a-fA-F_]*"
+        /* taking care not to interpret 3..5 as (3.)(.5) */
+        "|[0-9a-fA-F_]+(\\.[0-9a-fA-F_]+)?([eE][-+]?[0-9_]+)?"
+        "|=>|-[rwxoRWXOezsfdlpSugkbctTBMAC>]|~~|::"
+        "|&&=|\\|\\|=|//=|\\*\\*="
+        "|&&|\\|\\||//|\\+\\+|--|\\*\\*|\\.\\.\\.?"
+        "|[-+*/%.^&<>=!|]="
+        "|=~|!~"
+        "|<<|<>|<=>|>>"),
 PATTERNS("php",
         "^[\t ]*(((public|protected|private|static)[\t ]+)*function.*)$\n"
         "^[\t ]*(class.*)$",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+"
-        "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|[-+*/<>%&^|=!.]=|--|\\+\\+|<<=?|>>=?|===|&&|\\|\\||::|->"),
 PATTERNS("python", "^[ \t]*((class|def)[ \t].*)$",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+[jJlL]?|0[xX]?[0-9a-fA-F]+[lL]?"
-        "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"
-        "|[^[:space:]|[\x80-\xff]+"),
+        "|[-+*/<>%&^|=!]=|//=?|<<=?|>>=?|\\*\\*=?"),
         /* -- */
 PATTERNS("ruby", "^[ \t]*((class|module|def)[ \t].*)$",
         /* -- */
         "(@|@@|\\$)?[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+|0[xXbB]?[0-9a-fA-F]+|\\?(\\\\C-)?(\\\\M-)?."
-        "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"
-        "|[^[:space:]|[\x80-\xff]+"),
+        "|//=?|[-+*/<>%&^|=!]=|<<=?|>>=?|===|\\.{1,3}|::|[!=]~"),
 PATTERNS("bibtex", "(@[a-zA-Z]{1,}[ \t]*\\{{0,1}[ \t]*[^ \t\"@',\\#}{~%]*).*$",
         "[={}\"]|[^={}\" \t]+"),
 PATTERNS("tex", "^(\\\\((sub)*section|chapter|part)\\*{0,1}\\{.*)$",
-        "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+|[^[:space:]]"),
+        "\\\\[a-zA-Z@]+|\\\\.|[a-zA-Z0-9\x80-\xff]+"),
 PATTERNS("cpp",
         /* Jump targets or access declarations */
         "!^[ \t]*[A-Za-z_][A-Za-z_0-9]*:.*$\n"
@@ -80,8 +124,7 @@ PATTERNS("cpp",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
-        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 PATTERNS("csharp",
         /* Keywords */
         "!^[ \t]*(do|while|for|if|else|instanceof|new|return|switch|case|throw|catch|using)\n"
@@ -96,11 +139,11 @@ PATTERNS("csharp",
         /* -- */
         "[a-zA-Z_][a-zA-Z0-9_]*"
         "|[-+0-9.e]+[fFlL]?|0[xXbB]?[0-9a-fA-F]+[lL]?"
-        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"
-        "|[^[:space:]]|[\x80-\xff]+"),
+        "|[-+*/<>%&^|=!]=|--|\\+\\+|<<=?|>>=?|&&|\\|\\||::|->"),
 { "default", NULL, -1, { NULL, 0 } },
 };
 #undef PATTERNS
+#undef IPATTERN
 
 static struct userdiff_driver driver_true = {
        "diff=true",
@@ -230,7 +273,7 @@ struct userdiff_driver *userdiff_find_by_path(const char *path)
 
        if (!path)
                return NULL;
-       if (git_checkattr(path, 1, &check))
+       if (git_check_attr(path, 1, &check))
                return NULL;
 
        if (ATTR_TRUE(check.value))
@@ -241,3 +284,20 @@ struct userdiff_driver *userdiff_find_by_path(const char *path)
                return NULL;
        return userdiff_find_by_name(check.value);
 }
+
+struct userdiff_driver *userdiff_get_textconv(struct userdiff_driver *driver)
+{
+       if (!driver->textconv)
+               return NULL;
+
+       if (driver->textconv_want_cache && !driver->textconv_cache) {
+               struct notes_cache *c = xmalloc(sizeof(*c));
+               struct strbuf name = STRBUF_INIT;
+
+               strbuf_addf(&name, "textconv/%s", driver->name);
+               notes_cache_init(c, name.buf, driver->textconv);
+               driver->textconv_cache = c;
+       }
+
+       return driver;
+}