sequencer.c: rework search for start of footer to improve clarity
[gitweb.git] / sha1_name.c
index 5d81ea0564c8d90b417eb8ec5ed8c32baa2c3ea3..95003c77ea9a0ee240d60bc7958bb48245b14ad4 100644 (file)
@@ -9,14 +9,82 @@
 
 static int get_sha1_oneline(const char *, unsigned char *, struct commit_list *);
 
-static int find_short_object_filename(int len, const char *name, unsigned char *sha1)
+typedef int (*disambiguate_hint_fn)(const unsigned char *, void *);
+
+struct disambiguate_state {
+       disambiguate_hint_fn fn;
+       void *cb_data;
+       unsigned char candidate[20];
+       unsigned candidate_exists:1;
+       unsigned candidate_checked:1;
+       unsigned candidate_ok:1;
+       unsigned disambiguate_fn_used:1;
+       unsigned ambiguous:1;
+       unsigned always_call_fn:1;
+};
+
+static void update_candidates(struct disambiguate_state *ds, const unsigned char *current)
+{
+       if (ds->always_call_fn) {
+               ds->ambiguous = ds->fn(current, ds->cb_data) ? 1 : 0;
+               return;
+       }
+       if (!ds->candidate_exists) {
+               /* this is the first candidate */
+               hashcpy(ds->candidate, current);
+               ds->candidate_exists = 1;
+               return;
+       } else if (!hashcmp(ds->candidate, current)) {
+               /* the same as what we already have seen */
+               return;
+       }
+
+       if (!ds->fn) {
+               /* cannot disambiguate between ds->candidate and current */
+               ds->ambiguous = 1;
+               return;
+       }
+
+       if (!ds->candidate_checked) {
+               ds->candidate_ok = ds->fn(ds->candidate, ds->cb_data);
+               ds->disambiguate_fn_used = 1;
+               ds->candidate_checked = 1;
+       }
+
+       if (!ds->candidate_ok) {
+               /* discard the candidate; we know it does not satisify fn */
+               hashcpy(ds->candidate, current);
+               ds->candidate_checked = 0;
+               return;
+       }
+
+       /* if we reach this point, we know ds->candidate satisfies fn */
+       if (ds->fn(current, ds->cb_data)) {
+               /*
+                * if both current and candidate satisfy fn, we cannot
+                * disambiguate.
+                */
+               ds->candidate_ok = 0;
+               ds->ambiguous = 1;
+       }
+
+       /* otherwise, current can be discarded and candidate is still good */
+}
+
+static void find_short_object_filename(int len, const char *hex_pfx, struct disambiguate_state *ds)
 {
        struct alternate_object_database *alt;
        char hex[40];
-       int found = 0;
        static struct alternate_object_database *fakeent;
 
        if (!fakeent) {
+               /*
+                * Create a "fake" alternate object database that
+                * points to our own object database, to make it
+                * easier to get a temporary working space in
+                * alt->name/alt->base while iterating over the
+                * object databases including our own.
+                */
                const char *objdir = get_object_directory();
                int objdir_len = strlen(objdir);
                int entlen = objdir_len + 43;
@@ -27,33 +95,28 @@ static int find_short_object_filename(int len, const char *name, unsigned char *
        }
        fakeent->next = alt_odb_list;
 
-       sprintf(hex, "%.2s", name);
-       for (alt = fakeent; alt && found < 2; alt = alt->next) {
+       sprintf(hex, "%.2s", hex_pfx);
+       for (alt = fakeent; alt && !ds->ambiguous; alt = alt->next) {
                struct dirent *de;
                DIR *dir;
-               sprintf(alt->name, "%.2s/", name);
+               sprintf(alt->name, "%.2s/", hex_pfx);
                dir = opendir(alt->base);
                if (!dir)
                        continue;
-               while ((de = readdir(dir)) != NULL) {
+
+               while (!ds->ambiguous && (de = readdir(dir)) != NULL) {
+                       unsigned char sha1[20];
+
                        if (strlen(de->d_name) != 38)
                                continue;
-                       if (memcmp(de->d_name, name + 2, len - 2))
+                       if (memcmp(de->d_name, hex_pfx + 2, len - 2))
                                continue;
-                       if (!found) {
-                               memcpy(hex + 2, de->d_name, 38);
-                               found++;
-                       }
-                       else if (memcmp(hex + 2, de->d_name, 38)) {
-                               found = 2;
-                               break;
-                       }
+                       memcpy(hex + 2, de->d_name, 38);
+                       if (!get_sha1_hex(hex, sha1))
+                               update_candidates(ds, sha1);
                }
                closedir(dir);
        }
-       if (found == 1)
-               return get_sha1_hex(hex, sha1) == 0;
-       return found;
 }
 
 static int match_sha(unsigned len, const unsigned char *a, const unsigned char *b)
@@ -71,103 +134,157 @@ static int match_sha(unsigned len, const unsigned char *a, const unsigned char *
        return 1;
 }
 
-static int find_short_packed_object(int len, const unsigned char *match, unsigned char *sha1)
+static void unique_in_pack(int len,
+                         const unsigned char *bin_pfx,
+                          struct packed_git *p,
+                          struct disambiguate_state *ds)
 {
-       struct packed_git *p;
-       const unsigned char *found_sha1 = NULL;
-       int found = 0;
-
-       prepare_packed_git();
-       for (p = packed_git; p && found < 2; p = p->next) {
-               uint32_t num, last;
-               uint32_t first = 0;
-               open_pack_index(p);
-               num = p->num_objects;
-               last = num;
-               while (first < last) {
-                       uint32_t mid = (first + last) / 2;
-                       const unsigned char *now;
-                       int cmp;
-
-                       now = nth_packed_object_sha1(p, mid);
-                       cmp = hashcmp(match, now);
-                       if (!cmp) {
-                               first = mid;
-                               break;
-                       }
-                       if (cmp > 0) {
-                               first = mid+1;
-                               continue;
-                       }
-                       last = mid;
+       uint32_t num, last, i, first = 0;
+       const unsigned char *current = NULL;
+
+       open_pack_index(p);
+       num = p->num_objects;
+       last = num;
+       while (first < last) {
+               uint32_t mid = (first + last) / 2;
+               const unsigned char *current;
+               int cmp;
+
+               current = nth_packed_object_sha1(p, mid);
+               cmp = hashcmp(bin_pfx, current);
+               if (!cmp) {
+                       first = mid;
+                       break;
                }
-               if (first < num) {
-                       const unsigned char *now, *next;
-                      now = nth_packed_object_sha1(p, first);
-                       if (match_sha(len, match, now)) {
-                               next = nth_packed_object_sha1(p, first+1);
-                              if (!next|| !match_sha(len, match, next)) {
-                                       /* unique within this pack */
-                                       if (!found) {
-                                               found_sha1 = now;
-                                               found++;
-                                       }
-                                       else if (hashcmp(found_sha1, now)) {
-                                               found = 2;
-                                               break;
-                                       }
-                               }
-                               else {
-                                       /* not even unique within this pack */
-                                       found = 2;
-                                       break;
-                               }
-                       }
+               if (cmp > 0) {
+                       first = mid+1;
+                       continue;
                }
+               last = mid;
+       }
+
+       /*
+        * At this point, "first" is the location of the lowest object
+        * with an object name that could match "bin_pfx".  See if we have
+        * 0, 1 or more objects that actually match(es).
+        */
+       for (i = first; i < num && !ds->ambiguous; i++) {
+               current = nth_packed_object_sha1(p, i);
+               if (!match_sha(len, bin_pfx, current))
+                       break;
+               update_candidates(ds, current);
        }
-       if (found == 1)
-               hashcpy(sha1, found_sha1);
-       return found;
+}
+
+static void find_short_packed_object(int len, const unsigned char *bin_pfx,
+                                    struct disambiguate_state *ds)
+{
+       struct packed_git *p;
+
+       prepare_packed_git();
+       for (p = packed_git; p && !ds->ambiguous; p = p->next)
+               unique_in_pack(len, bin_pfx, p, ds);
 }
 
 #define SHORT_NAME_NOT_FOUND (-1)
 #define SHORT_NAME_AMBIGUOUS (-2)
 
-static int find_unique_short_object(int len, char *canonical,
-                                   unsigned char *res, unsigned char *sha1)
+static int finish_object_disambiguation(struct disambiguate_state *ds,
+                                       unsigned char *sha1)
 {
-       int has_unpacked, has_packed;
-       unsigned char unpacked_sha1[20], packed_sha1[20];
+       if (ds->ambiguous)
+               return SHORT_NAME_AMBIGUOUS;
 
-       prepare_alt_odb();
-       has_unpacked = find_short_object_filename(len, canonical, unpacked_sha1);
-       has_packed = find_short_packed_object(len, res, packed_sha1);
-       if (!has_unpacked && !has_packed)
+       if (!ds->candidate_exists)
                return SHORT_NAME_NOT_FOUND;
-       if (1 < has_unpacked || 1 < has_packed)
+
+       if (!ds->candidate_checked)
+               /*
+                * If this is the only candidate, there is no point
+                * calling the disambiguation hint callback.
+                *
+                * On the other hand, if the current candidate
+                * replaced an earlier candidate that did _not_ pass
+                * the disambiguation hint callback, then we do have
+                * more than one objects that match the short name
+                * given, so we should make sure this one matches;
+                * otherwise, if we discovered this one and the one
+                * that we previously discarded in the reverse order,
+                * we would end up showing different results in the
+                * same repository!
+                */
+               ds->candidate_ok = (!ds->disambiguate_fn_used ||
+                                   ds->fn(ds->candidate, ds->cb_data));
+
+       if (!ds->candidate_ok)
                return SHORT_NAME_AMBIGUOUS;
-       if (has_unpacked != has_packed) {
-               hashcpy(sha1, (has_packed ? packed_sha1 : unpacked_sha1));
+
+       hashcpy(sha1, ds->candidate);
+       return 0;
+}
+
+static int disambiguate_commit_only(const unsigned char *sha1, void *cb_data_unused)
+{
+       int kind = sha1_object_info(sha1, NULL);
+       return kind == OBJ_COMMIT;
+}
+
+static int disambiguate_committish_only(const unsigned char *sha1, void *cb_data_unused)
+{
+       struct object *obj;
+       int kind;
+
+       kind = sha1_object_info(sha1, NULL);
+       if (kind == OBJ_COMMIT)
+               return 1;
+       if (kind != OBJ_TAG)
                return 0;
-       }
-       /* Both have unique ones -- do they match? */
-       if (hashcmp(packed_sha1, unpacked_sha1))
-               return SHORT_NAME_AMBIGUOUS;
-       hashcpy(sha1, packed_sha1);
+
+       /* We need to do this the hard way... */
+       obj = deref_tag(lookup_object(sha1), NULL, 0);
+       if (obj && obj->type == OBJ_COMMIT)
+               return 1;
        return 0;
 }
 
-static int get_short_sha1(const char *name, int len, unsigned char *sha1,
-                         int quietly)
+static int disambiguate_tree_only(const unsigned char *sha1, void *cb_data_unused)
 {
-       int i, status;
-       char canonical[40];
-       unsigned char res[20];
+       int kind = sha1_object_info(sha1, NULL);
+       return kind == OBJ_TREE;
+}
 
-       if (len < MINIMUM_ABBREV || len > 40)
-               return -1;
-       hashclr(res);
-       memset(canonical, 'x', 40);
+static int disambiguate_treeish_only(const unsigned char *sha1, void *cb_data_unused)
+{
+       struct object *obj;
+       int kind;
+
+       kind = sha1_object_info(sha1, NULL);
+       if (kind == OBJ_TREE || kind == OBJ_COMMIT)
+               return 1;
+       if (kind != OBJ_TAG)
+               return 0;
+
+       /* We need to do this the hard way... */
+       obj = deref_tag(lookup_object(sha1), NULL, 0);
+       if (obj && (obj->type == OBJ_TREE || obj->type == OBJ_COMMIT))
+               return 1;
+       return 0;
+}
+
+static int disambiguate_blob_only(const unsigned char *sha1, void *cb_data_unused)
+{
+       int kind = sha1_object_info(sha1, NULL);
+       return kind == OBJ_BLOB;
+}
+
+static int prepare_prefixes(const char *name, int len,
+                           unsigned char *bin_pfx,
+                           char *hex_pfx)
+{
+       int i;
+
+       hashclr(bin_pfx);
+       memset(hex_pfx, 'x', 40);
        for (i = 0; i < len ;i++) {
                unsigned char c = name[i];
                unsigned char val;
@@ -181,18 +298,76 @@ static int get_short_sha1(const char *name, int len, unsigned char *sha1,
                }
                else
                        return -1;
-               canonical[i] = c;
+               hex_pfx[i] = c;
                if (!(i & 1))
                        val <<= 4;
-               res[i >> 1] |= val;
+               bin_pfx[i >> 1] |= val;
        }
+       return 0;
+}
+
+static int get_short_sha1(const char *name, int len, unsigned char *sha1,
+                         unsigned flags)
+{
+       int status;
+       char hex_pfx[40];
+       unsigned char bin_pfx[20];
+       struct disambiguate_state ds;
+       int quietly = !!(flags & GET_SHA1_QUIETLY);
+
+       if (len < MINIMUM_ABBREV || len > 40)
+               return -1;
+       if (prepare_prefixes(name, len, bin_pfx, hex_pfx) < 0)
+               return -1;
+
+       prepare_alt_odb();
+
+       memset(&ds, 0, sizeof(ds));
+       if (flags & GET_SHA1_COMMIT)
+               ds.fn = disambiguate_commit_only;
+       else if (flags & GET_SHA1_COMMITTISH)
+               ds.fn = disambiguate_committish_only;
+       else if (flags & GET_SHA1_TREE)
+               ds.fn = disambiguate_tree_only;
+       else if (flags & GET_SHA1_TREEISH)
+               ds.fn = disambiguate_treeish_only;
+       else if (flags & GET_SHA1_BLOB)
+               ds.fn = disambiguate_blob_only;
+
+       find_short_object_filename(len, hex_pfx, &ds);
+       find_short_packed_object(len, bin_pfx, &ds);
+       status = finish_object_disambiguation(&ds, sha1);
 
-       status = find_unique_short_object(i, canonical, res, sha1);
        if (!quietly && (status == SHORT_NAME_AMBIGUOUS))
-               return error("short SHA1 %.*s is ambiguous.", len, canonical);
+               return error("short SHA1 %.*s is ambiguous.", len, hex_pfx);
        return status;
 }
 
+
+int for_each_abbrev(const char *prefix, each_abbrev_fn fn, void *cb_data)
+{
+       char hex_pfx[40];
+       unsigned char bin_pfx[20];
+       struct disambiguate_state ds;
+       int len = strlen(prefix);
+
+       if (len < MINIMUM_ABBREV || len > 40)
+               return -1;
+       if (prepare_prefixes(prefix, len, bin_pfx, hex_pfx) < 0)
+               return -1;
+
+       prepare_alt_odb();
+
+       memset(&ds, 0, sizeof(ds));
+       ds.always_call_fn = 1;
+       ds.cb_data = cb_data;
+       ds.fn = fn;
+
+       find_short_object_filename(len, hex_pfx, &ds);
+       find_short_packed_object(len, bin_pfx, &ds);
+       return ds.ambiguous;
+}
+
 const char *find_unique_abbrev(const unsigned char *sha1, int len)
 {
        int status, exists;
@@ -204,7 +379,7 @@ const char *find_unique_abbrev(const unsigned char *sha1, int len)
                return hex;
        while (len < 40) {
                unsigned char sha1_ret[20];
-               status = get_short_sha1(hex, len, sha1_ret, 1);
+               status = get_short_sha1(hex, len, sha1_ret, GET_SHA1_QUIETLY);
                if (exists
                    ? !status
                    : status == SHORT_NAME_NOT_FOUND) {
@@ -255,7 +430,7 @@ static inline int upstream_mark(const char *string, int len)
        return 0;
 }
 
-static int get_sha1_1(const char *name, int len, unsigned char *sha1);
+static int get_sha1_1(const char *name, int len, unsigned char *sha1, unsigned lookup_flags);
 
 static int get_sha1_basic(const char *str, int len, unsigned char *sha1)
 {
@@ -292,7 +467,7 @@ static int get_sha1_basic(const char *str, int len, unsigned char *sha1)
                ret = interpret_branch_name(str+at, &buf);
                if (ret > 0) {
                        /* substitute this branch name and restart */
-                       return get_sha1_1(buf.buf, buf.len, sha1);
+                       return get_sha1_1(buf.buf, buf.len, sha1, 0);
                } else if (ret == 0) {
                        return -1;
                }
@@ -362,7 +537,7 @@ static int get_parent(const char *name, int len,
                      unsigned char *result, int idx)
 {
        unsigned char sha1[20];
-       int ret = get_sha1_1(name, len, sha1);
+       int ret = get_sha1_1(name, len, sha1, GET_SHA1_COMMITTISH);
        struct commit *commit;
        struct commit_list *p;
 
@@ -395,7 +570,7 @@ static int get_nth_ancestor(const char *name, int len,
        struct commit *commit;
        int ret;
 
-       ret = get_sha1_1(name, len, sha1);
+       ret = get_sha1_1(name, len, sha1, GET_SHA1_COMMITTISH);
        if (ret)
                return ret;
        commit = lookup_commit_reference(sha1);
@@ -441,6 +616,7 @@ static int peel_onion(const char *name, int len, unsigned char *sha1)
        unsigned char outer[20];
        const char *sp;
        unsigned int expected_type = 0;
+       unsigned lookup_flags = 0;
        struct object *o;
 
        /*
@@ -476,7 +652,10 @@ static int peel_onion(const char *name, int len, unsigned char *sha1)
        else
                return -1;
 
-       if (get_sha1_1(name, sp - name - 2, outer))
+       if (expected_type == OBJ_COMMIT)
+               lookup_flags = GET_SHA1_COMMITTISH;
+
+       if (get_sha1_1(name, sp - name - 2, outer, lookup_flags))
                return -1;
 
        o = parse_object(outer);
@@ -525,6 +704,7 @@ static int peel_onion(const char *name, int len, unsigned char *sha1)
 static int get_describe_name(const char *name, int len, unsigned char *sha1)
 {
        const char *cp;
+       unsigned flags = GET_SHA1_QUIETLY | GET_SHA1_COMMIT;
 
        for (cp = name + len - 1; name + 2 <= cp; cp--) {
                char ch = *cp;
@@ -535,14 +715,14 @@ static int get_describe_name(const char *name, int len, unsigned char *sha1)
                        if (ch == 'g' && cp[-1] == '-') {
                                cp++;
                                len -= cp - name;
-                               return get_short_sha1(cp, len, sha1, 1);
+                               return get_short_sha1(cp, len, sha1, flags);
                        }
                }
        }
        return -1;
 }
 
-static int get_sha1_1(const char *name, int len, unsigned char *sha1)
+static int get_sha1_1(const char *name, int len, unsigned char *sha1, unsigned lookup_flags)
 {
        int ret, has_suffix;
        const char *cp;
@@ -587,7 +767,7 @@ static int get_sha1_1(const char *name, int len, unsigned char *sha1)
        if (!ret)
                return 0;
 
-       return get_short_sha1(name, len, sha1, 0);
+       return get_short_sha1(name, len, sha1, lookup_flags);
 }
 
 /*
@@ -769,7 +949,7 @@ int get_sha1_mb(const char *name, unsigned char *sha1)
                struct strbuf sb;
                strbuf_init(&sb, dots - name);
                strbuf_add(&sb, name, dots - name);
-               st = get_sha1(sb.buf, sha1_tmp);
+               st = get_sha1_committish(sb.buf, sha1_tmp);
                strbuf_release(&sb);
        }
        if (st)
@@ -778,7 +958,7 @@ int get_sha1_mb(const char *name, unsigned char *sha1)
        if (!one)
                return -1;
 
-       if (get_sha1(dots[3] ? (dots + 3) : "HEAD", sha1_tmp))
+       if (get_sha1_committish(dots[3] ? (dots + 3) : "HEAD", sha1_tmp))
                return -1;
        two = lookup_commit_reference_gently(sha1_tmp, 0);
        if (!two)
@@ -905,7 +1085,52 @@ int strbuf_check_branch_ref(struct strbuf *sb, const char *name)
 int get_sha1(const char *name, unsigned char *sha1)
 {
        struct object_context unused;
-       return get_sha1_with_context(name, sha1, &unused);
+       return get_sha1_with_context(name, 0, sha1, &unused);
+}
+
+/*
+ * Many callers know that the user meant to name a committish by
+ * syntactical positions where the object name appears.  Calling this
+ * function allows the machinery to disambiguate shorter-than-unique
+ * abbreviated object names between committish and others.
+ *
+ * Note that this does NOT error out when the named object is not a
+ * committish. It is merely to give a hint to the disambiguation
+ * machinery.
+ */
+int get_sha1_committish(const char *name, unsigned char *sha1)
+{
+       struct object_context unused;
+       return get_sha1_with_context(name, GET_SHA1_COMMITTISH,
+                                    sha1, &unused);
+}
+
+int get_sha1_treeish(const char *name, unsigned char *sha1)
+{
+       struct object_context unused;
+       return get_sha1_with_context(name, GET_SHA1_TREEISH,
+                                    sha1, &unused);
+}
+
+int get_sha1_commit(const char *name, unsigned char *sha1)
+{
+       struct object_context unused;
+       return get_sha1_with_context(name, GET_SHA1_COMMIT,
+                                    sha1, &unused);
+}
+
+int get_sha1_tree(const char *name, unsigned char *sha1)
+{
+       struct object_context unused;
+       return get_sha1_with_context(name, GET_SHA1_TREE,
+                                    sha1, &unused);
+}
+
+int get_sha1_blob(const char *name, unsigned char *sha1)
+{
+       struct object_context unused;
+       return get_sha1_with_context(name, GET_SHA1_BLOB,
+                                    sha1, &unused);
 }
 
 /* Must be called only when object_name:filename doesn't exist. */
@@ -1004,16 +1229,6 @@ static void diagnose_invalid_index_path(int stage,
 }
 
 
-int get_sha1_with_mode_1(const char *name, unsigned char *sha1, unsigned *mode,
-                        int only_to_die, const char *prefix)
-{
-       struct object_context oc;
-       int ret;
-       ret = get_sha1_with_context_1(name, sha1, &oc, only_to_die, prefix);
-       *mode = oc.mode;
-       return ret;
-}
-
 static char *resolve_relative_path(const char *rel)
 {
        if (prefixcmp(rel, "./") && prefixcmp(rel, "../"))
@@ -1031,20 +1246,24 @@ static char *resolve_relative_path(const char *rel)
                           rel);
 }
 
-int get_sha1_with_context_1(const char *name, unsigned char *sha1,
-                           struct object_context *oc,
-                           int only_to_die, const char *prefix)
+static int get_sha1_with_context_1(const char *name,
+                                  unsigned flags,
+                                  const char *prefix,
+                                  unsigned char *sha1,
+                                  struct object_context *oc)
 {
        int ret, bracket_depth;
        int namelen = strlen(name);
        const char *cp;
+       int only_to_die = flags & GET_SHA1_ONLY_TO_DIE;
 
        memset(oc, 0, sizeof(*oc));
        oc->mode = S_IFINVALID;
-       ret = get_sha1_1(name, namelen, sha1);
+       ret = get_sha1_1(name, namelen, sha1, flags);
        if (!ret)
                return ret;
-       /* sha1:path --> object name of path in ent sha1
+       /*
+        * sha1:path --> object name of path in ent sha1
         * :path -> object name of absolute path in index
         * :./path -> object name of path relative to cwd in index
         * :[0-3]:path -> object name of path in index at stage
@@ -1119,7 +1338,7 @@ int get_sha1_with_context_1(const char *name, unsigned char *sha1,
                        strncpy(object_name, name, cp-name);
                        object_name[cp-name] = '\0';
                }
-               if (!get_sha1_1(name, cp-name, tree_sha1)) {
+               if (!get_sha1_1(name, cp-name, tree_sha1, GET_SHA1_TREEISH)) {
                        const char *filename = cp+1;
                        char *new_filename = NULL;
 
@@ -1146,3 +1365,22 @@ int get_sha1_with_context_1(const char *name, unsigned char *sha1,
        }
        return ret;
 }
+
+/*
+ * Call this function when you know "name" given by the end user must
+ * name an object but it doesn't; the function _may_ die with a better
+ * diagnostic message than "no such object 'name'", e.g. "Path 'doc' does not
+ * exist in 'HEAD'" when given "HEAD:doc", or it may return in which case
+ * you have a chance to diagnose the error further.
+ */
+void maybe_die_on_misspelt_object_name(const char *name, const char *prefix)
+{
+       struct object_context oc;
+       unsigned char sha1[20];
+       get_sha1_with_context_1(name, GET_SHA1_ONLY_TO_DIE, prefix, sha1, &oc);
+}
+
+int get_sha1_with_context(const char *str, unsigned flags, unsigned char *sha1, struct object_context *orc)
+{
+       return get_sha1_with_context_1(str, flags, NULL, sha1, orc);
+}