blame.h

   1 #ifndef BLAME_H
   2 #define BLAME_H
   3
   4 #include "cache.h"
   5 #include "commit.h"
   6 #include "xdiff-interface.h"
   7 #include "revision.h"
   8 #include "prio-queue.h"
   9 #include "diff.h"
  10
  11 #define PICKAXE_BLAME_MOVE              01
  12 #define PICKAXE_BLAME_COPY              02
  13 #define PICKAXE_BLAME_COPY_HARDER       04
  14 #define PICKAXE_BLAME_COPY_HARDEST      010
  15
  16 #define BLAME_DEFAULT_MOVE_SCORE        20
  17 #define BLAME_DEFAULT_COPY_SCORE        40
  18
  19 struct fingerprint;
  20
  21 /*
  22  * One blob in a commit that is being suspected
  23  */
  24 struct blame_origin {
  25         int refcnt;
  26         /* Record preceding blame record for this blob */
  27         struct blame_origin *previous;
  28         /* origins are put in a list linked via `next' hanging off the
  29          * corresponding commit's util field in order to make finding
  30          * them fast.  The presence in this chain does not count
  31          * towards the origin's reference count.  It is tempting to
  32          * let it count as long as the commit is pending examination,
  33          * but even under circumstances where the commit will be
  34          * present multiple times in the priority queue of unexamined
  35          * commits, processing the first instance will not leave any
  36          * work requiring the origin data for the second instance.  An
  37          * interspersed commit changing that would have to be
  38          * preexisting with a different ancestry and with the same
  39          * commit date in order to wedge itself between two instances
  40          * of the same commit in the priority queue _and_ produce
  41          * blame entries relevant for it.  While we don't want to let
  42          * us get tripped up by this case, it certainly does not seem
  43          * worth optimizing for.
  44          */
  45         struct blame_origin *next;
  46         struct commit *commit;
  47         /* `suspects' contains blame entries that may be attributed to
  48          * this origin's commit or to parent commits.  When a commit
  49          * is being processed, all suspects will be moved, either by
  50          * assigning them to an origin in a different commit, or by
  51          * shipping them to the scoreboard's ent list because they
  52          * cannot be attributed to a different commit.
  53          */
  54         struct blame_entry *suspects;
  55         mmfile_t file;
  56         int num_lines;
  57         struct fingerprint *fingerprints;
  58         struct object_id blob_oid;
  59         unsigned short mode;
  60         /* guilty gets set when shipping any suspects to the final
  61          * blame list instead of other commits
  62          */
  63         char guilty;
  64         char path[FLEX_ARRAY];
  65 };
  66
  67 /*
  68  * Each group of lines is described by a blame_entry; it can be split
  69  * as we pass blame to the parents.  They are arranged in linked lists
  70  * kept as `suspects' of some unprocessed origin, or entered (when the
  71  * blame origin has been finalized) into the scoreboard structure.
  72  * While the scoreboard structure is only sorted at the end of
  73  * processing (according to final image line number), the lists
  74  * attached to an origin are sorted by the target line number.
  75  */
  76 struct blame_entry {
  77         struct blame_entry *next;
  78
  79         /* the first line of this group in the final image;
  80          * internally all line numbers are 0 based.
  81          */
  82         int lno;
  83
  84         /* how many lines this group has */
  85         int num_lines;
  86
  87         /* the commit that introduced this group into the final image */
  88         struct blame_origin *suspect;
  89
  90         /* the line number of the first line of this group in the
  91          * suspect's file; internally all line numbers are 0 based.
  92          */
  93         int s_lno;
  94
  95         /* how significant this entry is -- cached to avoid
  96          * scanning the lines over and over.
  97          */
  98         unsigned score;
  99         int ignored;
 100         int unblamable;
 101 };
 102
 103 struct blame_bloom_data;
 104
 105 /*
 106  * The current state of the blame assignment.
 107  */
 108 struct blame_scoreboard {
 109         /* the final commit (i.e. where we started digging from) */
 110         struct commit *final;
 111         /* Priority queue for commits with unassigned blame records */
 112         struct prio_queue commits;
 113         struct repository *repo;
 114         struct rev_info *revs;
 115         const char *path;
 116
 117         /*
 118          * The contents in the final image.
 119          * Used by many functions to obtain contents of the nth line,
 120          * indexed with scoreboard.lineno[blame_entry.lno].
 121          */
 122         const char *final_buf;
 123         unsigned long final_buf_size;
 124
 125         /* linked list of blames */
 126         struct blame_entry *ent;
 127
 128         struct oidset ignore_list;
 129
 130         /* look-up a line in the final buffer */
 131         int num_lines;
 132         int *lineno;
 133
 134         /* stats */
 135         int num_read_blob;
 136         int num_get_patch;
 137         int num_commits;
 138
 139         /*
 140          * blame for a blame_entry with score lower than these thresholds
 141          * is not passed to the parent using move/copy logic.
 142          */
 143         unsigned move_score;
 144         unsigned copy_score;
 145
 146         /* use this file's contents as the final image */
 147         const char *contents_from;
 148
 149         /* flags */
 150         int reverse;
 151         int show_root;
 152         int xdl_opts;
 153         int no_whole_file_rename;
 154         int debug;
 155
 156         /* callbacks */
 157         void(*on_sanity_fail)(struct blame_scoreboard *, int);
 158         void(*found_guilty_entry)(struct blame_entry *, void *);
 159
 160         void *found_guilty_entry_data;
 161         struct blame_bloom_data *bloom_data;
 162 };
 163
 164 /*
 165  * Origin is refcounted and usually we keep the blob contents to be
 166  * reused.
 167  */
 168 static inline struct blame_origin *blame_origin_incref(struct blame_origin *o)
 169 {
 170         if (o)
 171                 o->refcnt++;
 172         return o;
 173 }
 174 void blame_origin_decref(struct blame_origin *o);
 175
 176 void blame_coalesce(struct blame_scoreboard *sb);
 177 void blame_sort_final(struct blame_scoreboard *sb);
 178 unsigned blame_entry_score(struct blame_scoreboard *sb, struct blame_entry *e);
 179 void assign_blame(struct blame_scoreboard *sb, int opt);
 180 const char *blame_nth_line(struct blame_scoreboard *sb, long lno);
 181
 182 void init_scoreboard(struct blame_scoreboard *sb);
 183 void setup_scoreboard(struct blame_scoreboard *sb,
 184                       const char *path,
 185                       struct blame_origin **orig);
 186 void setup_blame_bloom_data(struct blame_scoreboard *sb,
 187                             const char *path);
 188 void cleanup_scoreboard(struct blame_scoreboard *sb);
 189
 190 struct blame_entry *blame_entry_prepend(struct blame_entry *head,
 191                                         long start, long end,
 192                                         struct blame_origin *o);
 193
 194 struct blame_origin *get_blame_suspects(struct commit *commit);
 195
 196 #endif /* BLAME_H */