diff --git a/PKG-INFO b/PKG-INFO
index d5b37a98f21d98df7b25d174c5a949f2e628de42..9c87cc02c1f691d197d9c360ffea667dc9a35977 100644
--- a/PKG-INFO
+++ b/PKG-INFO
@@ -1,6 +1,6 @@
Metadata-Version: 1.0
Name: swh.model
-Version: 0.0.21
+Version: 0.0.22
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/docs/persistent-identifiers.rst b/docs/persistent-identifiers.rst
index c796a808ff8670f9a13cc13f32b5018ccdfbc35d..7f41d610561b80a288aa3719eeeca543bc92848d 100644
--- a/docs/persistent-identifiers.rst
+++ b/docs/persistent-identifiers.rst
@@ -68,7 +68,7 @@ captured by ``<object_type>``:
* ``rel`` to **releases**,
* ``rev`` to **revisions**,
* ``dir`` to **directories**,
-* ``cnt`` to **releases**.
+* ``cnt`` to **contents**.
The actual object pointed to is identified by the intrinsic identifier
``<object_id>``, which is a hex-encoded (using lowercase ASCII characters) SHA1
diff --git a/swh.model.egg-info/PKG-INFO b/swh.model.egg-info/PKG-INFO
index d5b37a98f21d98df7b25d174c5a949f2e628de42..9c87cc02c1f691d197d9c360ffea667dc9a35977 100644
--- a/swh.model.egg-info/PKG-INFO
+++ b/swh.model.egg-info/PKG-INFO
@@ -1,6 +1,6 @@
Metadata-Version: 1.0
Name: swh.model
-Version: 0.0.21
+Version: 0.0.22
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/swh.model.egg-info/SOURCES.txt b/swh.model.egg-info/SOURCES.txt
index d8706419ea007d2a2294198e0fc110ad06cfad9b..b45f730dce555fb482c6a30a86be695e9ed496b6 100644
--- a/swh.model.egg-info/SOURCES.txt
+++ b/swh.model.egg-info/SOURCES.txt
@@ -38,6 +38,7 @@ swh/model/from_disk.py
swh/model/hashutil.py
swh/model/identifiers.py
swh/model/merkle.py
+swh/model/toposort.py
swh/model/validators.py
swh/model/fields/__init__.py
swh/model/fields/compound.py
@@ -49,6 +50,7 @@ swh/model/tests/test_from_disk.py
swh/model/tests/test_hashutil.py
swh/model/tests/test_identifiers.py
swh/model/tests/test_merkle.py
+swh/model/tests/test_toposort.py
swh/model/tests/test_validators.py
swh/model/tests/fields/__init__.py
swh/model/tests/fields/test_compound.py
diff --git a/swh/model/tests/test_toposort.py b/swh/model/tests/test_toposort.py
new file mode 100644
index 0000000000000000000000000000000000000000..66a8ee1c8ae9543b6dffdccf3a7fd389f8150630
--- /dev/null
+++ b/swh/model/tests/test_toposort.py
@@ -0,0 +1,99 @@
+# Copyright (C) 2017-2018 The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import unittest
+from swh.model.toposort import toposort
+
+
+def is_toposorted_slow(revision_log):
+ """Check (inefficiently) that the given revision log is in any topological
+ order.
+
+ Complexity: O(n^2).
+ (Note: It's totally possible to write a O(n) is_toposorted function,
+ but it requires computing the transitive closure of the input DAG,
+ which requires computing a topological ordering of that DAG, which
+ kind of defeats the purpose of writing unit tests for toposort().)
+
+ Args:
+ revision_log: Revision log as returned by
+ swh.storage.Storage.revision_log().
+
+ Returns:
+ True if the revision log is topologically sorted.
+ """
+ rev_by_id = {r['id']: r for r in revision_log}
+
+ def all_parents(revision):
+ for parent in revision['parents']:
+ yield parent
+ yield from all_parents(rev_by_id[parent])
+
+ visited = set()
+ for rev in revision_log:
+ visited.add(rev['id'])
+ if not all(parent in visited for parent in all_parents(rev)):
+ return False
+ return True
+
+
+class TestToposort(unittest.TestCase):
+ def generate_log(self, graph):
+ for node_id, parents in graph.items():
+ yield {'id': node_id, 'parents': tuple(parents)}
+
+ def unordered_log(self, log):
+ return {(d['id'], tuple(d['parents'])) for d in log}
+
+ def check(self, graph):
+ log = list(self.generate_log(graph))
+ topolog = list(toposort(log))
+ self.assertEqual(len(topolog), len(graph))
+ self.assertEqual(self.unordered_log(topolog), self.unordered_log(log))
+ self.assertTrue(is_toposorted_slow(toposort(log)))
+
+ def test_linked_list(self):
+ self.check({3: [2],
+ 2: [1],
+ 1: []})
+
+ def test_fork(self):
+ self.check({7: [6],
+ 6: [4],
+ 5: [3],
+ 4: [2],
+ 3: [2],
+ 2: [1],
+ 1: []})
+
+ def test_fork_merge(self):
+ self.check({8: [7, 5],
+ 7: [6],
+ 6: [4],
+ 5: [3],
+ 4: [2],
+ 3: [2],
+ 2: [1],
+ 1: []})
+
+ def test_two_origins(self):
+ self.check({9: [8],
+ 8: [7, 5],
+ 7: [6],
+ 6: [4],
+ 5: [3],
+ 4: [],
+ 3: []})
+
+ def test_three_way(self):
+ self.check({9: [8, 4, 2],
+ 8: [7, 5],
+ 7: [6],
+ 6: [4],
+ 5: [3],
+ 4: [2],
+ 3: [2],
+ 2: [1],
+ 1: []})
diff --git a/swh/model/toposort.py b/swh/model/toposort.py
new file mode 100644
index 0000000000000000000000000000000000000000..b0a7231a5e58a70d3d6c36477fd763717fde06cf
--- /dev/null
+++ b/swh/model/toposort.py
@@ -0,0 +1,43 @@
+# Copyright (C) 2017-2018 The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import collections
+
+
+def toposort(revision_log):
+ """Perform a topological sort on a revision log graph.
+
+ Complexity: O(N) (linear in the length of the revision log)
+
+ Args:
+ revision_log: Revision log as returned by
+ swh.storage.Storage.revision_log().
+
+ Yields:
+ The revision log sorted by a topological order
+ """
+ in_degree = {} # rev_id -> numbers of parents left to compute
+ children = collections.defaultdict(list) # rev_id -> children
+
+ # Compute the in_degrees and the parents of all the revisions.
+ # Add the roots to the processing queue.
+ queue = collections.deque()
+ for rev in revision_log:
+ parents = rev['parents']
+ in_degree[rev['id']] = len(parents)
+ if not parents:
+ queue.append(rev)
+ for parent in parents:
+ children[parent].append(rev)
+
+ # Topological sort: yield the 'ready' nodes, decrease the in degree of
+ # their children and add the 'ready' ones to the queue.
+ while queue:
+ rev = queue.popleft()
+ yield rev
+ for child in children[rev['id']]:
+ in_degree[child['id']] -= 1
+ if in_degree[child['id']] == 0:
+ queue.append(child)
diff --git a/version.txt b/version.txt
index 943a41a2f9b8a76b2fa6d967ea9da976022b3ae6..3a338cb2e577d69dc9a1844e049bfc834a93a7e7 100644
--- a/version.txt
+++ b/version.txt
@@ -1 +1 @@
-v0.0.21-0-gbdf26f5
\ No newline at end of file
+v0.0.22-0-ga06122e
\ No newline at end of file