[PATCH] sparse-revlog: protect C code against delta chain including nullrev

[Boris Feld]

# HG changeset patch
# User Boris Feld <boris.feld at octobus.net>
# Date 1543508919 -3600
#      Thu Nov 29 17:28:39 2018 +0100
# Node ID 5a6a715f4ed0fa4d47bdd6b6413797fe6cb17f88
# Parent  2f14d1bbc9a7a142b421285c0708320b46be7a56
# EXP-Topic sparse-followup
# Available At https://bitbucket.org/octobus/mercurial-devel/
#              hg pull https://bitbucket.org/octobus/mercurial-devel/ -r 5a6a715f4ed0
sparse-revlog: protect C code against delta chain including nullrev

For unclear reasons, some repositories include nullrev (-1). Re-computing
delta for such repo remove nullrev from all chain, so some older versions have
been creating them.

This currently raise an IndexError with the new C code doing chain slicing as
it expect all item to be positive.

Both python and C code for reading delta chain preserve nullrev, and the
Python code for chain slicing does not raise an exception in this case. So we
take the safe route and make the new C code works fine in that case.

diff --git a/mercurial/cext/revlog.c b/mercurial/cext/revlog.c
--- a/mercurial/cext/revlog.c
+++ b/mercurial/cext/revlog.c
@@ -1064,6 +1064,13 @@ index_segment_span(indexObject *self, Py
 	int64_t start_offset;
 	int64_t end_offset;
 	int end_size;
+	if (start_rev == -1) {
+        /* revlog have "harmless" corruption, keep going */
+        if (start_rev == end_rev) {
+            return 0;
+        }
+        start_rev += 1;
+    }
 	start_offset = index_get_start(self, start_rev);
 	if (start_offset < 0) {
 		return -1;
@@ -1188,7 +1195,7 @@ static PyObject *index_slicechunktodensi
 		if (revnum == -1 && PyErr_Occurred()) {
 			goto bail;
 		}
-		if (revnum < 0 || revnum >= idxlen) {
+		if (revnum < -1 || revnum >= idxlen) {
 			PyErr_Format(PyExc_IndexError,
 			             "index out of range: %zd", revnum);
 			goto bail;