blame.h

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