convert.c

   1 #include "cache.h"
   2 #include "attr.h"
   3
   4 /*
   5  * convert.c - convert a file when checking it out and checking it in.
   6  *
   7  * This should use the pathname to decide on whether it wants to do some
   8  * more interesting conversions (automatic gzip/unzip, general format
   9  * conversions etc etc), but by default it just does automatic CRLF<->LF
  10  * translation when the "auto_crlf" option is set.
  11  */
  12
  13 #define CRLF_GUESS      (-1)
  14 #define CRLF_BINARY     0
  15 #define CRLF_TEXT       1
  16 #define CRLF_INPUT      2
  17
  18 struct text_stat {
  19         /* CR, LF and CRLF counts */
  20         unsigned cr, lf, crlf;
  21
  22         /* These are just approximations! */
  23         unsigned printable, nonprintable;
  24 };
  25
  26 static void gather_stats(const char *buf, unsigned long size, struct text_stat *stats)
  27 {
  28         unsigned long i;
  29
  30         memset(stats, 0, sizeof(*stats));
  31
  32         for (i = 0; i < size; i++) {
  33                 unsigned char c = buf[i];
  34                 if (c == '\r') {
  35                         stats->cr++;
  36                         if (i+1 < size && buf[i+1] == '\n')
  37                                 stats->crlf++;
  38                         continue;
  39                 }
  40                 if (c == '\n') {
  41                         stats->lf++;
  42                         continue;
  43                 }
  44                 if (c == 127)
  45                         /* DEL */
  46                         stats->nonprintable++;
  47                 else if (c < 32) {
  48                         switch (c) {
  49                                 /* BS, HT, ESC and FF */
  50                         case '\b': case '\t': case '\033': case '\014':
  51                                 stats->printable++;
  52                                 break;
  53                         default:
  54                                 stats->nonprintable++;
  55                         }
  56                 }
  57                 else
  58                         stats->printable++;
  59         }
  60 }
  61
  62 /*
  63  * The same heuristics as diff.c::mmfile_is_binary()
  64  */
  65 static int is_binary(unsigned long size, struct text_stat *stats)
  66 {
  67
  68         if ((stats->printable >> 7) < stats->nonprintable)
  69                 return 1;
  70         /*
  71          * Other heuristics? Average line length might be relevant,
  72          * as might LF vs CR vs CRLF counts..
  73          *
  74          * NOTE! It might be normal to have a low ratio of CRLF to LF
  75          * (somebody starts with a LF-only file and edits it with an editor
  76          * that adds CRLF only to lines that are added..). But do  we
  77          * want to support CR-only? Probably not.
  78          */
  79         return 0;
  80 }
  81
  82 static int crlf_to_git(const char *path, char **bufp, unsigned long *sizep, int action)
  83 {
  84         char *buffer, *nbuf;
  85         unsigned long size, nsize;
  86         struct text_stat stats;
  87
  88         if ((action == CRLF_BINARY) || (action == CRLF_GUESS && !auto_crlf))
  89                 return 0;
  90
  91         size = *sizep;
  92         if (!size)
  93                 return 0;
  94         buffer = *bufp;
  95
  96         gather_stats(buffer, size, &stats);
  97
  98         /* No CR? Nothing to convert, regardless. */
  99         if (!stats.cr)
 100                 return 0;
 101
 102         if (action == CRLF_GUESS) {
 103                 /*
 104                  * We're currently not going to even try to convert stuff
 105                  * that has bare CR characters. Does anybody do that crazy
 106                  * stuff?
 107                  */
 108                 if (stats.cr != stats.crlf)
 109                         return 0;
 110
 111                 /*
 112                  * And add some heuristics for binary vs text, of course...
 113                  */
 114                 if (is_binary(size, &stats))
 115                         return 0;
 116         }
 117
 118         /*
 119          * Ok, allocate a new buffer, fill it in, and return true
 120          * to let the caller know that we switched buffers on it.
 121          */
 122         nsize = size - stats.crlf;
 123         nbuf = xmalloc(nsize);
 124         *bufp = nbuf;
 125         *sizep = nsize;
 126
 127         if (action == CRLF_GUESS) {
 128                 /*
 129                  * If we guessed, we already know we rejected a file with
 130                  * lone CR, and we can strip a CR without looking at what
 131                  * follow it.
 132                  */
 133                 do {
 134                         unsigned char c = *buffer++;
 135                         if (c != '\r')
 136                                 *nbuf++ = c;
 137                 } while (--size);
 138         } else {
 139                 do {
 140                         unsigned char c = *buffer++;
 141                         if (! (c == '\r' && (1 < size && *buffer == '\n')))
 142                                 *nbuf++ = c;
 143                 } while (--size);
 144         }
 145
 146         return 1;
 147 }
 148
 149 static int crlf_to_worktree(const char *path, char **bufp, unsigned long *sizep, int action)
 150 {
 151         char *buffer, *nbuf;
 152         unsigned long size, nsize;
 153         struct text_stat stats;
 154         unsigned char last;
 155
 156         if ((action == CRLF_BINARY) || (action == CRLF_INPUT) ||
 157             (action == CRLF_GUESS && auto_crlf <= 0))
 158                 return 0;
 159
 160         size = *sizep;
 161         if (!size)
 162                 return 0;
 163         buffer = *bufp;
 164
 165         gather_stats(buffer, size, &stats);
 166
 167         /* No LF? Nothing to convert, regardless. */
 168         if (!stats.lf)
 169                 return 0;
 170
 171         /* Was it already in CRLF format? */
 172         if (stats.lf == stats.crlf)
 173                 return 0;
 174
 175         if (action == CRLF_GUESS) {
 176                 /* If we have any bare CR characters, we're not going to touch it */
 177                 if (stats.cr != stats.crlf)
 178                         return 0;
 179
 180                 if (is_binary(size, &stats))
 181                         return 0;
 182         }
 183
 184         /*
 185          * Ok, allocate a new buffer, fill it in, and return true
 186          * to let the caller know that we switched buffers on it.
 187          */
 188         nsize = size + stats.lf - stats.crlf;
 189         nbuf = xmalloc(nsize);
 190         *bufp = nbuf;
 191         *sizep = nsize;
 192         last = 0;
 193         do {
 194                 unsigned char c = *buffer++;
 195                 if (c == '\n' && last != '\r')
 196                         *nbuf++ = '\r';
 197                 *nbuf++ = c;
 198                 last = c;
 199         } while (--size);
 200
 201         return 1;
 202 }
 203
 204 static void setup_crlf_check(struct git_attr_check *check)
 205 {
 206         static struct git_attr *attr_crlf;
 207
 208         if (!attr_crlf)
 209                 attr_crlf = git_attr("crlf", 4);
 210         check->attr = attr_crlf;
 211 }
 212
 213 static int git_path_check_crlf(const char *path)
 214 {
 215         struct git_attr_check attr_crlf_check;
 216
 217         setup_crlf_check(&attr_crlf_check);
 218
 219         if (!git_checkattr(path, 1, &attr_crlf_check)) {
 220                 const char *value = attr_crlf_check.value;
 221                 if (ATTR_TRUE(value))
 222                         return CRLF_TEXT;
 223                 else if (ATTR_FALSE(value))
 224                         return CRLF_BINARY;
 225                 else if (ATTR_UNSET(value))
 226                         ;
 227                 else if (!strcmp(value, "input"))
 228                         return CRLF_INPUT;
 229                 /* fallthru */
 230         }
 231         return CRLF_GUESS;
 232 }
 233
 234 int convert_to_git(const char *path, char **bufp, unsigned long *sizep)
 235 {
 236         return crlf_to_git(path, bufp, sizep, git_path_check_crlf(path));
 237 }
 238
 239 int convert_to_working_tree(const char *path, char **bufp, unsigned long *sizep)
 240 {
 241         return crlf_to_worktree(path, bufp, sizep, git_path_check_crlf(path));
 242 }