extmod/modure: Convert byte offsets to unicode indices when necessary.

And add a test.

Fixes issue #9202.

Signed-off-by: Jeff Epler <jepler@gmail.com>

1 files changed, 32 insertions, 0 deletions

diff --git a/tests/unicode/unicode_ure.py b/tests/unicode/unicode_ure.py
new file mode 100644
index 000000000..5a5dc6005
--- /dev/null
+++ b/tests/unicode/unicode_ure.py
@@ -0,0 +1,32 @@
+# test match.span() for unicode strings
+
+try:
+    import ure as re
+except ImportError:
+    try:
+        import re
+    except ImportError:
+        print("SKIP")
+        raise SystemExit
+
+try:
+    m = re.match(".", "a")
+    m.span
+except AttributeError:
+    print("SKIP")
+    raise SystemExit
+
+
+def print_spans(match):
+    print("----")
+    try:
+        i = 0
+        while True:
+            print(match.span(i), match.start(i), match.end(i))
+            i += 1
+    except IndexError:
+        pass
+
+
+m = re.match(r"([0-9]*)(([a-z]*)([0-9]*))", "1234\u2764567")
+print_spans(m)