Check mutation parents on tree sequence init

c/tests/test_tables.c

@@ -9011,7 +9011,7 @@ test_sort_tables_mutation_times(void)
     CU_ASSERT_EQUAL_FATAL(ret, 0);
 
     /* Check to make sure we have legal mutations */
-    ret = tsk_treeseq_init(&ts, &tables, 0);
+    ret = tsk_treeseq_init(&ts, &tables, TSK_TS_INIT_COMPUTE_MUTATION_PARENTS);
     CU_ASSERT_EQUAL_FATAL(ret, 0);
 
     ret = tsk_treeseq_copy_tables(&ts, &t1, 0);

c/tests/test_trees.c

@@ -3886,16 +3886,16 @@ test_simplest_mutation_edges(void)
                         "1  2   2   0\n";
     const char *sites = "0.5  0\n"
                         "1.5  0\n";
-    const char *mutations = "0    1     1\n"
+    const char *mutations = "0    2     1\n"
+                            "0    1     1\n"
                             "0    0     1\n"
-                            "0    2     1\n"
-                            "1    0     1\n"
+                            "1    2     1\n"
                             "1    1     1\n"
-                            "1    2     1\n";
+                            "1    0     1\n";
     tsk_treeseq_t ts;
     tsk_tree_t tree;
     /* We have mutations over roots, samples and just isolated nodes */
-    tsk_id_t mutation_edges[] = { -1, 0, -1, 1, -1, -1 };
+    tsk_id_t mutation_edges[] = { -1, -1, 0, -1, -1, 1 };
     tsk_size_t i, j, k, t;
     tsk_mutation_t mut;
     tsk_site_t site;
@@ -4238,7 +4238,7 @@ test_single_tree_bad_mutations(void)
     ret = tsk_treeseq_init(&ts, &tables, load_flags);
     CU_ASSERT_EQUAL(ret, TSK_ERR_MUTATION_PARENT_AFTER_CHILD);
     tsk_treeseq_free(&ts);
-    tables.mutations.parent[3] = TSK_NULL;
+    tables.mutations.parent[3] = 2;
 
     /* time < node time */
     tables.mutations.time[2] = 0;

c/tests/testlib.c

@@ -767,6 +767,8 @@ tsk_treeseq_from_text(tsk_treeseq_t *ts, double sequence_length, const char *nod
     tsk_table_collection_t tables;
     tsk_id_t max_population_id;
     tsk_size_t j;
+    tsk_flags_t ts_flags;
+    bool all_parents_null;
 
     CU_ASSERT_FATAL(ts != NULL);
     CU_ASSERT_FATAL(nodes != NULL);
@@ -807,7 +809,21 @@ tsk_treeseq_from_text(tsk_treeseq_t *ts, double sequence_length, const char *nod
         }
     }
 
-    ret = tsk_treeseq_init(ts, &tables, TSK_TS_INIT_BUILD_INDEXES);
+    /* If all mutation.parent are TSK_NULL, use TSK_TS_COMPUTE_MUTATION_PARENTS flag too
+     */
+    ts_flags = TSK_TS_INIT_BUILD_INDEXES;
+    all_parents_null = true;
+    for (j = 0; j < tables.mutations.num_rows; j++) {
+        if (tables.mutations.parent[j] != TSK_NULL) {
+            all_parents_null = false;
+            break;
+        }
+    }
+    if (all_parents_null) {
+        ts_flags |= TSK_TS_INIT_COMPUTE_MUTATION_PARENTS;
+    }
+
+    ret = tsk_treeseq_init(ts, &tables, ts_flags);
     /* tsk_treeseq_print_state(ts, stdout); */
     if (ret != 0) {
         printf("\nret = %s\n", tsk_strerror(ret));

c/tskit/core.c

@@ -347,6 +347,12 @@ tsk_strerror_internal(int err)
                   "(TSK_ERR_DISALLOWED_UNKNOWN_MUTATION_TIME)";
             break;
 
+        case TSK_ERR_BAD_MUTATION_PARENT:
+            ret = "A mutation's parent is not consistent with the topology of the tree. "
+                  "Use compute_mutation_parents to set the parents correctly."
+                  "(TSK_ERR_BAD_MUTATION_PARENT)";
+            break;
+
         /* Migration errors */
         case TSK_ERR_UNSORTED_MIGRATIONS:
             ret = "Migrations must be sorted by time. (TSK_ERR_UNSORTED_MIGRATIONS)";

c/tskit/core.h

@@ -510,6 +510,12 @@ Some mutations have TSK_UNKNOWN_TIME in an algorithm where that's
 disallowed (use compute_mutation_times?).
 */
 #define TSK_ERR_DISALLOWED_UNKNOWN_MUTATION_TIME                    -510
+
+/** 
+A mutation's parent was not consistent with the topology of the tree.
+ */
+#define TSK_ERR_BAD_MUTATION_PARENT                                 -511
+
 /** @} */
 
 /**