File 0001-use-editdistance-not-ukkonen.patch of Package python-identify
From: Matthias Fehring <buschmann23@opensuse.org>
Date: 2023-11-01 09:44:00 +0100
Subject: Use editdistance instead of ukkonen
Upstream: never
Upstream switched from editdistance_s to ukkonen that is currently
not packaged for openSUSE.
---
identify/identify.py | 10 ++++------
setup.cfg | 2 +-
2 files changed, 5 insertions(+), 7 deletions(-)
--- a/identify/identify.py 2023-10-28 19:19:41.000000000 +0200
+++ b/identify/identify.py 2023-11-01 09:39:16.942021416 +0100
@@ -243,7 +243,7 @@
3. check exact text match with existing licenses
4. failing that use edit distance
"""
- import ukkonen # `pip install identify[license]`
+ import editdistance # `pip install identify[license]`
with open(filename, encoding='UTF-8') as f:
contents = f.read()
@@ -253,8 +253,6 @@
min_edit_dist = sys.maxsize
min_edit_dist_spdx = ''
- cutoff = math.ceil(.05 * len(norm))
-
# try exact matches
for spdx, text in licenses.LICENSES:
norm_license = _norm_license(text)
@@ -265,13 +263,13 @@
if norm and abs(len(norm) - len(norm_license)) / len(norm) > .05:
continue
- edit_dist = ukkonen.distance(norm, norm_license, cutoff)
- if edit_dist < cutoff and edit_dist < min_edit_dist:
+ edit_dist = editdistance.distance(norm, norm_license)
+ if edit_dist < min_edit_dist:
min_edit_dist = edit_dist
min_edit_dist_spdx = spdx
# if there's less than 5% edited from the license, we found our match
- if norm and min_edit_dist < cutoff:
+ if norm and min_edit_dist / len(norm) < .05:
return min_edit_dist_spdx
else:
# no matches :'(
--- a/setup.cfg 2023-10-28 19:19:41.000000000 +0200
+++ b/setup.cfg 2023-11-01 10:13:07.904157653 +0100
@@ -31,7 +31,7 @@
[options.extras_require]
license =
- ukkonen
+ editdistance
[options.package_data]
identify =