Script 'mail_helper' called by obssrc
Hello community,
here is the log from the commit of package python-identify for openSUSE:Factory
checked in at 2023-11-01 22:10:52
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/python-identify (Old)
and /work/SRC/openSUSE:Factory/.python-identify.new.17445 (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "python-identify"
Wed Nov 1 22:10:52 2023 rev:10 rq:1121649 version:2.5.31
Changes:
--------
--- /work/SRC/openSUSE:Factory/python-identify/python-identify.changes
2021-08-19 10:02:12.427177273 +0200
+++
/work/SRC/openSUSE:Factory/.python-identify.new.17445/python-identify.changes
2023-11-01 22:11:27.042935282 +0100
@@ -1,0 +2,17 @@
+Wed Nov 1 09:17:44 UTC 2023 - Matthias Fehring <buschman...@opensuse.org>
+
+- update to version 2.5.31
+ * no changelog available, alternatively have a look at
+ https://github.com/pre-commit/identify/compare/v2.2.13...v2.5.31
+- use editdistance instead of ukkonen
+ * Upstream changed from editdistance_s to ukkonen for performance
+ reasons. Ukkonen is not now available for openSUSE. So switch
+ back to editdistance for now.
+ * Add 0001-use-editdistance-not-ukkonen.patch
+- spec file changes
+ * remove not required python-setuptools from Requires
+ * upstream URL has changed to https://github.com/pre-commit/identify
+ * change requirement for python-editdistance from Suggests to
+ Recommends
+
+-------------------------------------------------------------------
Old:
----
identify-2.2.13.tar.gz
New:
----
0001-use-editdistance-not-ukkonen.patch
identify-2.5.31.tar.gz
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Other differences:
------------------
++++++ python-identify.spec ++++++
--- /var/tmp/diff_new_pack.dIAZOO/_old 2023-11-01 22:11:27.526953212 +0100
+++ /var/tmp/diff_new_pack.dIAZOO/_new 2023-11-01 22:11:27.526953212 +0100
@@ -1,7 +1,7 @@
#
# spec file for package python-identify
#
-# Copyright (c) 2021 SUSE LLC
+# Copyright (c) 2023 SUSE LLC
#
# All modifications and additions to the file contributed by third parties
# remain the property of their copyright owners, unless otherwise agreed
@@ -16,24 +16,24 @@
#
-%{?!python_module:%define python_module() python-%{**} python3-%{**}}
Name: python-identify
-Version: 2.2.13
+Version: 2.5.31
Release: 0
Summary: File identification library for Python
License: MIT
Group: Development/Languages/Python
-URL: https://github.com/chriskuehl/identify
-Source:
https://github.com/chriskuehl/identify/archive/v%{version}.tar.gz#/identify-%{version}.tar.gz
+URL: https://github.com/pre-commit/identify
+Source:
https://github.com/pre-commit/identify/archive/v%{version}.tar.gz#/identify-%{version}.tar.gz
+# PATCH-FIX-OPENSUSE 0001-use-editdistance-not-ukkonen.patch -- ukkonen not
packaged for opensuse now
+Patch1: 0001-use-editdistance-not-ukkonen.patch
BuildRequires: %{python_module editdistance}
BuildRequires: %{python_module pytest}
BuildRequires: %{python_module setuptools}
BuildRequires: fdupes
BuildRequires: python-rpm-macros
-Requires: python-setuptools
Requires(post): update-alternatives
Requires(postun):update-alternatives
-Suggests: python-editdistance
+Recommends: python-editdistance
BuildArch: noarch
%python_subpackages
@@ -42,9 +42,7 @@
%prep
%setup -q -n identify-%{version}
-# stick with editdistance as it generally has more functionality
-sed -i 's/editdistance_s.distance/editdistance.eval/' identify/identify.py
-sed -i 's/editdistance_s/editdistance/' identify/identify.py
+%autopatch -p1
%build
%python_build
@@ -67,5 +65,6 @@
%doc README.md
%license LICENSE
%python_alternative %{_bindir}/identify-cli
-%{python_sitelib}/*
+%{python_sitelib}/identify
+%{python_sitelib}/identify-%{version}-*-info
++++++ 0001-use-editdistance-not-ukkonen.patch ++++++
From: Matthias Fehring <buschman...@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 =
++++++ identify-2.2.13.tar.gz -> identify-2.5.31.tar.gz ++++++
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/.github/FUNDING.yml
new/identify-2.5.31/.github/FUNDING.yml
--- old/identify-2.2.13/.github/FUNDING.yml 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/.github/FUNDING.yml 1970-01-01 01:00:00.000000000
+0100
@@ -1,2 +0,0 @@
-github: asottile
-open_collective: pre-commit
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/.github/workflows/main.yml
new/identify-2.5.31/.github/workflows/main.yml
--- old/identify-2.2.13/.github/workflows/main.yml 1970-01-01
01:00:00.000000000 +0100
+++ new/identify-2.5.31/.github/workflows/main.yml 2023-10-28
19:19:41.000000000 +0200
@@ -0,0 +1,13 @@
+name: main
+
+on:
+ push:
+ branches: [main, test-me-*]
+ tags: '*'
+ pull_request:
+
+jobs:
+ main:
+ uses: asottile/workflows/.github/workflows/tox.yml@v1.5.0
+ with:
+ env: '["py38", "py39", "py310", "py311"]'
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/.gitignore
new/identify-2.5.31/.gitignore
--- old/identify-2.2.13/.gitignore 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/.gitignore 2023-10-28 19:19:41.000000000 +0200
@@ -1,8 +1,5 @@
*.egg-info
*.py[co]
/.coverage
-/.pytest_cache
/.tox
-/coverage-html
/dist
-/venv
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/.pre-commit-config.yaml
new/identify-2.5.31/.pre-commit-config.yaml
--- old/identify-2.2.13/.pre-commit-config.yaml 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/.pre-commit-config.yaml 2023-10-28 19:19:41.000000000
+0200
@@ -1,45 +1,42 @@
repos:
- repo: https://github.com/pre-commit/pre-commit-hooks
- rev: v4.0.1
+ rev: v4.5.0
hooks:
- - id: check-docstring-first
+ - id: trailing-whitespace
+ - id: end-of-file-fixer
- id: check-yaml
- id: debug-statements
- id: double-quote-string-fixer
- - id: end-of-file-fixer
- id: name-tests-test
- id: requirements-txt-fixer
- - id: trailing-whitespace
- repo: https://github.com/asottile/setup-cfg-fmt
- rev: v1.17.0
+ rev: v2.5.0
hooks:
- id: setup-cfg-fmt
-- repo: https://github.com/PyCQA/flake8
- rev: 3.9.2
- hooks:
- - id: flake8
- exclude: ^identify/vendor/licenses\.py$
- additional_dependencies: [flake8-typing-imports==1.10.1]
-- repo: https://github.com/pre-commit/mirrors-autopep8
- rev: v1.5.7
- hooks:
- - id: autopep8
-- repo: https://github.com/asottile/reorder_python_imports
- rev: v2.6.0
+- repo: https://github.com/asottile/reorder-python-imports
+ rev: v3.12.0
hooks:
- id: reorder-python-imports
- args: [--py3-plus]
+ args: [--py38-plus, --add-import, 'from __future__ import annotations']
- repo: https://github.com/asottile/add-trailing-comma
- rev: v2.1.0
+ rev: v3.1.0
hooks:
- id: add-trailing-comma
- args: [--py36-plus]
- repo: https://github.com/asottile/pyupgrade
- rev: v2.23.1
+ rev: v3.15.0
hooks:
- id: pyupgrade
- args: [--py36-plus]
+ args: [--py38-plus]
+- repo: https://github.com/hhatto/autopep8
+ rev: v2.0.4
+ hooks:
+ - id: autopep8
+- repo: https://github.com/PyCQA/flake8
+ rev: 6.1.0
+ hooks:
+ - id: flake8
+ exclude: ^identify/vendor/licenses\.py$
- repo: https://github.com/pre-commit/mirrors-mypy
- rev: v0.910
+ rev: v1.6.1
hooks:
- id: mypy
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/README.md
new/identify-2.5.31/README.md
--- old/identify-2.2.13/README.md 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/README.md 2023-10-28 19:19:41.000000000 +0200
@@ -1,11 +1,9 @@
+[](https://github.com/pre-commit/identify/actions/workflows/main.yml)
+[](https://results.pre-commit.ci/latest/github/pre-commit/identify/main)
+
identify
========
-[](https://dev.azure.com/asottile/asottile/_build/latest?definitionId=67&branchName=master)
-[](https://dev.azure.com/asottile/asottile/_build/latest?definitionId=67&branchName=master)
-[](https://results.pre-commit.ci/latest/github/pre-commit/identify/master)
-[](https://pypi.python.org/pypi/identify)
-
File identification library for Python.
Given a file (or some information about a file), return a set of standardized
@@ -13,7 +11,9 @@
## Installation
-`pip install identify`
+```bash
+pip install identify
+```
## Usage
### With a file on disk
@@ -80,7 +80,7 @@
$ identify-cli setup.py; echo $?
["file", "non-executable", "python", "text"]
0
-$ identify setup.py --filename-only; echo $?
+$ identify-cli setup.py --filename-only; echo $?
["python", "text"]
0
$ identify-cli wat.wat; echo $?
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/azure-pipelines.yml
new/identify-2.5.31/azure-pipelines.yml
--- old/identify-2.2.13/azure-pipelines.yml 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/azure-pipelines.yml 1970-01-01 01:00:00.000000000
+0100
@@ -1,19 +0,0 @@
-trigger:
- branches:
- include: [master, test-me-*]
- tags:
- include: ['*']
-
-resources:
- repositories:
- - repository: asottile
- type: github
- endpoint: github
- name: asottile/azure-pipeline-templates
- ref: refs/tags/v2.1.0
-
-jobs:
-- template: job--python-tox.yml@asottile
- parameters:
- toxenvs: [pypy3, py36, py37, py38]
- os: linux
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/bin/vendor-licenses
new/identify-2.5.31/bin/vendor-licenses
--- old/identify-2.2.13/bin/vendor-licenses 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/bin/vendor-licenses 2023-10-28 19:19:41.000000000
+0200
@@ -3,6 +3,8 @@
./bin/vendor-licenses > identify/vendor/licenses.py
"""
+from __future__ import annotations
+
import argparse
import os.path
import subprocess
@@ -54,4 +56,4 @@
if __name__ == '__main__':
- exit(main())
+ raise SystemExit(main())
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/identify/cli.py
new/identify-2.5.31/identify/cli.py
--- old/identify-2.2.13/identify/cli.py 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/identify/cli.py 2023-10-28 19:19:41.000000000 +0200
@@ -1,12 +1,13 @@
+from __future__ import annotations
+
import argparse
import json
-from typing import Optional
from typing import Sequence
from identify import identify
-def main(argv: Optional[Sequence[str]] = None) -> int:
+def main(argv: Sequence[str] | None = None) -> int:
parser = argparse.ArgumentParser()
parser.add_argument('--filename-only', action='store_true')
parser.add_argument('path')
@@ -31,4 +32,4 @@
if __name__ == '__main__':
- exit(main())
+ raise SystemExit(main())
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/identify/extensions.py
new/identify-2.5.31/identify/extensions.py
--- old/identify-2.2.13/identify/extensions.py 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/identify/extensions.py 2023-10-28 19:19:41.000000000
+0200
@@ -1,6 +1,8 @@
+from __future__ import annotations
EXTENSIONS = {
'adoc': {'text', 'asciidoc'},
'ai': {'binary', 'adobe-illustrator'},
+ 'aj': {'text', 'aspectj'},
'asciidoc': {'text', 'asciidoc'},
'apinotes': {'text', 'apinotes'},
'asar': {'binary', 'asar'},
@@ -8,14 +10,20 @@
'bash': {'text', 'shell', 'bash'},
'bat': {'text', 'batch'},
'bats': {'text', 'shell', 'bash', 'bats'},
+ 'bazel': {'text', 'bazel'},
+ 'beancount': {'text', 'beancount'},
'bib': {'text', 'bib'},
'bmp': {'binary', 'image', 'bitmap'},
'bz2': {'binary', 'bzip2'},
'bzl': {'text', 'bazel'},
'c': {'text', 'c'},
+ 'c++': {'text', 'c++'},
+ 'c++m': {'text', 'c++'},
'cc': {'text', 'c++'},
+ 'ccm': {'text', 'c++'},
'cfg': {'text'},
'chs': {'text', 'c2hs'},
+ 'cjs': {'text', 'javascript'},
'clj': {'text', 'clojure'},
'cljc': {'text', 'clojure'},
'cljs': {'text', 'clojure', 'clojurescript'},
@@ -24,6 +32,8 @@
'coffee': {'text', 'coffee'},
'conf': {'text'},
'cpp': {'text', 'c++'},
+ 'cppm': {'text', 'c++'},
+ 'cr': {'text', 'crystal'},
'crt': {'text', 'pem'},
'cs': {'text', 'c#'},
'csproj': {'text', 'xml', 'csproj'},
@@ -32,24 +42,39 @@
'css': {'text', 'css'},
'csv': {'text', 'csv'},
'cu': {'text', 'cuda'},
+ 'cue': {'text', 'cue'},
'cuh': {'text', 'cuda'},
'cxx': {'text', 'c++'},
+ 'cxxm': {'text', 'c++'},
+ 'cylc': {'text', 'cylc'},
'dart': {'text', 'dart'},
+ 'dbc': {'text', 'dbc'},
'def': {'text', 'def'},
'dll': {'binary'},
'dtd': {'text', 'dtd'},
'ear': {'binary', 'zip', 'jar'},
'edn': {'text', 'clojure', 'edn'},
'ejs': {'text', 'ejs'},
+ 'env': {'text', 'dotenv'},
'eot': {'binary', 'eot'},
'eps': {'binary', 'eps'},
'erb': {'text', 'erb'},
+ 'erl': {'text', 'erlang'},
+ 'ex': {'text', 'elixir'},
'exe': {'binary'},
+ 'exs': {'text', 'elixir'},
'eyaml': {'text', 'yaml'},
+ 'f03': {'text', 'fortran'},
+ 'f08': {'text', 'fortran'},
+ 'f90': {'text', 'fortran'},
+ 'f95': {'text', 'fortran'},
'feature': {'text', 'gherkin'},
'fish': {'text', 'fish'},
+ 'fits': {'binary', 'fits'},
'gd': {'text', 'gdscript'},
'gemspec': {'text', 'ruby'},
+ 'geojson': {'text', 'geojson', 'json'},
+ 'ggb': {'binary', 'zip', 'ggb'},
'gif': {'binary', 'image', 'gif'},
'go': {'text', 'go'},
'gotmpl': {'text', 'gotmpl'},
@@ -62,8 +87,11 @@
'gypi': {'text', 'gyp', 'python'},
'gz': {'binary', 'gzip'},
'h': {'text', 'header', 'c', 'c++'},
+ 'hbs': {'text', 'handlebars'},
+ 'hcl': {'text', 'hcl'},
'hh': {'text', 'header', 'c++'},
'hpp': {'text', 'header', 'c++'},
+ 'hrl': {'text', 'erlang'},
'hs': {'text', 'haskell'},
'htm': {'text', 'html'},
'html': {'text', 'html'},
@@ -75,8 +103,11 @@
'idr': {'text', 'idris'},
'inc': {'text', 'inc'},
'ini': {'text', 'ini'},
+ 'inl': {'text', 'inl', 'c++'},
+ 'ino': {'text', 'ino', 'c++'},
'inx': {'text', 'xml', 'inx'},
- 'ipynb': {'text', 'jupyter'},
+ 'ipynb': {'text', 'jupyter', 'json'},
+ 'ixx': {'text', 'c++'},
'j2': {'text', 'jinja'},
'jade': {'text', 'jade'},
'jar': {'binary', 'zip', 'jar'},
@@ -85,11 +116,14 @@
'jenkinsfile': {'text', 'groovy', 'jenkins'},
'jinja': {'text', 'jinja'},
'jinja2': {'text', 'jinja'},
+ 'jl': {'text', 'julia'},
'jpeg': {'binary', 'image', 'jpeg'},
'jpg': {'binary', 'image', 'jpeg'},
'js': {'text', 'javascript'},
'json': {'text', 'json'},
+ 'jsonld': {'text', 'json', 'jsonld'},
'jsonnet': {'text', 'jsonnet'},
+ 'json5': {'text', 'json5'},
'jsx': {'text', 'jsx'},
'key': {'text', 'pem'},
'kml': {'text', 'kml', 'xml'},
@@ -98,31 +132,44 @@
'lean': {'text', 'lean'},
'lektorproject': {'text', 'ini', 'lektorproject'},
'less': {'text', 'less'},
+ 'lfm': {'text', 'lazarus', 'lazarus-form'},
'lhs': {'text', 'literate-haskell'},
'libsonnet': {'text', 'jsonnet'},
'lidr': {'text', 'idris'},
+ 'liquid': {'text', 'liquid'},
+ 'lpi': {'text', 'lazarus', 'xml'},
+ 'lpr': {'text', 'lazarus', 'pascal'},
'lr': {'text', 'lektor'},
'lua': {'text', 'lua'},
- 'm': {'text', 'c', 'objective-c'},
+ 'm': {'text', 'objective-c'},
+ 'm4': {'text', 'm4'},
+ 'make': {'text', 'makefile'},
'manifest': {'text', 'manifest'},
'map': {'text', 'map'},
'markdown': {'text', 'markdown'},
'md': {'text', 'markdown'},
'mdx': {'text', 'mdx'},
+ 'meson': {'text', 'meson'},
'mib': {'text', 'mib'},
+ 'mjs': {'text', 'javascript'},
'mk': {'text', 'makefile'},
'ml': {'text', 'ocaml'},
'mli': {'text', 'ocaml'},
'mm': {'text', 'c++', 'objective-c++'},
'modulemap': {'text', 'modulemap'},
+ 'mscx': {'text', 'xml', 'musescore'},
+ 'mscz': {'binary', 'zip', 'musescore'},
+ 'mustache': {'text', 'mustache'},
'myst': {'text', 'myst'},
'ngdoc': {'text', 'ngdoc'},
'nim': {'text', 'nim'},
'nims': {'text', 'nim'},
'nimble': {'text', 'nimble'},
'nix': {'text', 'nix'},
+ 'njk': {'text', 'nunjucks'},
'otf': {'binary', 'otf'},
'p12': {'binary', 'p12'},
+ 'pas': {'text', 'pascal'},
'patch': {'text', 'diff'},
'pdf': {'binary', 'pdf'},
'pem': {'text', 'pem'},
@@ -135,9 +182,13 @@
'pm': {'text', 'perl'},
'png': {'binary', 'image', 'png'},
'po': {'text', 'pofile'},
+ 'pom': {'pom', 'text', 'xml'},
'pp': {'text', 'puppet'},
+ 'prisma': {'text', 'prisma'},
'properties': {'text', 'java-properties'},
'proto': {'text', 'proto'},
+ 'ps1': {'text', 'powershell'},
+ 'pug': {'text', 'pug'},
'puml': {'text', 'plantuml'},
'purs': {'text', 'purescript'},
'pxd': {'text', 'cython'},
@@ -148,9 +199,12 @@
'pyx': {'text', 'cython'},
'pyz': {'binary', 'pyz'},
'pyzw': {'binary', 'pyz'},
+ 'qml': {'text', 'qml'},
'r': {'text', 'r'},
'rake': {'text', 'ruby'},
'rb': {'text', 'ruby'},
+ 'resx': {'text', 'resx', 'xml'},
+ 'rng': {'text', 'xml', 'relax-ng'},
'rs': {'text', 'rust'},
'rst': {'text', 'rst'},
's': {'text', 'asm'},
@@ -168,8 +222,10 @@
'spec': {'text', 'spec'},
'sql': {'text', 'sql'},
'ss': {'text', 'scheme'},
+ 'sty': {'text', 'tex'},
'styl': {'text', 'stylus'},
'sv': {'text', 'system-verilog'},
+ 'svelte': {'text', 'svelte'},
'svg': {'text', 'image', 'svg', 'xml'},
'svh': {'text', 'system-verilog'},
'swf': {'binary', 'swf'},
@@ -178,6 +234,7 @@
'tac': {'text', 'twisted', 'python'},
'tar': {'binary', 'tar'},
'tex': {'text', 'tex'},
+ 'textproto': {'text', 'textproto'},
'tf': {'text', 'terraform'},
'tfvars': {'text', 'terraform'},
'tgz': {'binary', 'gzip'},
@@ -191,6 +248,7 @@
'twig': {'text', 'twig'},
'txsprofile': {'text', 'ini', 'txsprofile'},
'txt': {'text', 'plain-text'},
+ 'txtpb': {'text', 'textproto'},
'urdf': {'text', 'xml', 'urdf'},
'v': {'text', 'verilog'},
'vb': {'text', 'vb'},
@@ -200,6 +258,7 @@
'vh': {'text', 'verilog'},
'vhd': {'text', 'vhdl'},
'vim': {'text', 'vim'},
+ 'vtl': {'text', 'vtl'},
'vue': {'text', 'vue'},
'war': {'binary', 'zip', 'jar'},
'wav': {'binary', 'audio', 'wav'},
@@ -210,6 +269,7 @@
'woff2': {'binary', 'woff2'},
'wsgi': {'text', 'wsgi', 'python'},
'xhtml': {'text', 'xml', 'html', 'xhtml'},
+ 'xacro': {'text', 'xml', 'urdf', 'xacro'},
'xml': {'text', 'xml'},
'xq': {'text', 'xquery'},
'xql': {'text', 'xquery'},
@@ -220,6 +280,7 @@
'xsd': {'text', 'xml', 'xsd'},
'xsl': {'text', 'xml', 'xsl'},
'yaml': {'text', 'yaml'},
+ 'yamlld': {'text', 'yaml', 'yamlld'},
'yang': {'text', 'yang'},
'yin': {'text', 'xml', 'yin'},
'yml': {'text', 'yaml'},
@@ -231,9 +292,11 @@
}
EXTENSIONS_NEED_BINARY_CHECK = {
'plist': {'plist'},
+ 'ppm': {'image', 'ppm'},
}
NAMES = {
+ '.ansible-lint': EXTENSIONS['yaml'],
'.babelrc': EXTENSIONS['json'] | {'babelrc'},
'.bash_aliases': EXTENSIONS['bash'],
'.bash_profile': EXTENSIONS['bash'],
@@ -256,11 +319,13 @@
'.gitlint': EXTENSIONS['ini'] | {'gitlint'},
'.gitmodules': {'text', 'gitmodules'},
'.hgrc': EXTENSIONS['ini'] | {'hgrc'},
+ '.isort.cfg': EXTENSIONS['ini'] | {'isort'},
'.jshintrc': EXTENSIONS['json'] | {'jshintrc'},
'.mailmap': {'text', 'mailmap'},
'.mention-bot': EXTENSIONS['json'] | {'mention-bot'},
'.npmignore': {'text', 'npmignore'},
'.pdbrc': EXTENSIONS['py'] | {'pdbrc'},
+ '.prettierignore': {'text', 'gitignore', 'prettierignore'},
'.pypirc': EXTENSIONS['ini'] | {'pypirc'},
'.rstcheck.cfg': EXTENSIONS['ini'],
'.yamllint': EXTENSIONS['yaml'] | {'yamllint'},
@@ -271,19 +336,25 @@
'.zshenv': EXTENSIONS['zsh'],
'AUTHORS': EXTENSIONS['txt'],
'BUILD': EXTENSIONS['bzl'],
- 'BUILD.bazel': EXTENSIONS['bzl'],
'CMakeLists.txt': EXTENSIONS['cmake'],
'CHANGELOG': EXTENSIONS['txt'],
+ 'config.ru': EXTENSIONS['rb'],
+ 'Containerfile': {'text', 'dockerfile'},
'CONTRIBUTING': EXTENSIONS['txt'],
+ 'copy.bara.sky': EXTENSIONS['bzl'],
'COPYING': EXTENSIONS['txt'],
'Dockerfile': {'text', 'dockerfile'},
'Gemfile': EXTENSIONS['rb'],
'Gemfile.lock': {'text'},
'GNUmakefile': EXTENSIONS['mk'],
+ 'go.mod': {'text', 'go-mod'},
+ 'go.sum': {'text', 'go-sum'},
'Jenkinsfile': EXTENSIONS['jenkins'],
'LICENSE': EXTENSIONS['txt'],
'MAINTAINERS': EXTENSIONS['txt'],
'Makefile': EXTENSIONS['mk'],
+ 'meson.build': EXTENSIONS['meson'],
+ 'meson_options.txt': EXTENSIONS['meson'],
'makefile': EXTENSIONS['mk'],
'NEWS': EXTENSIONS['txt'],
'NOTICE': EXTENSIONS['txt'],
@@ -291,10 +362,15 @@
'Pipfile': EXTENSIONS['toml'],
'Pipfile.lock': EXTENSIONS['json'],
'PKGBUILD': {'text', 'bash', 'pkgbuild', 'alpm'},
+ 'poetry.lock': EXTENSIONS['toml'],
+ 'pom.xml': EXTENSIONS['pom'],
'pylintrc': EXTENSIONS['ini'] | {'pylintrc'},
'README': EXTENSIONS['txt'],
'Rakefile': EXTENSIONS['rb'],
+ 'rebar.config': EXTENSIONS['erl'],
'setup.cfg': EXTENSIONS['ini'],
+ 'sys.config': EXTENSIONS['erl'],
+ 'sys.config.src': EXTENSIONS['erl'],
'WORKSPACE': EXTENSIONS['bzl'],
'wscript': EXTENSIONS['py'],
}
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/identify/identify.py
new/identify-2.5.31/identify/identify.py
--- old/identify-2.2.13/identify/identify.py 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/identify/identify.py 2023-10-28 19:19:41.000000000
+0200
@@ -1,4 +1,7 @@
+from __future__ import annotations
+
import errno
+import math
import os.path
import re
import shlex
@@ -6,10 +9,6 @@
import string
import sys
from typing import IO
-from typing import List
-from typing import Optional
-from typing import Set
-from typing import Tuple
from identify import extensions
from identify import interpreters
@@ -38,7 +37,7 @@
ALL_TAGS = frozenset(_ALL_TAGS)
-def tags_from_path(path: str) -> Set[str]:
+def tags_from_path(path: str) -> set[str]:
try:
sr = os.lstat(path)
except (OSError, ValueError): # same error-handling as `os.lexists()`
@@ -84,7 +83,7 @@
return tags
-def tags_from_filename(path: str) -> Set[str]:
+def tags_from_filename(path: str) -> set[str]:
_, filename = os.path.split(path)
_, ext = os.path.splitext(filename)
@@ -106,7 +105,7 @@
return ret
-def tags_from_interpreter(interpreter: str) -> Set[str]:
+def tags_from_interpreter(interpreter: str) -> set[str]:
_, _, interpreter = interpreter.rpartition('/')
# Try "python3.5.2" => "python3.5" => "python3" until one matches.
@@ -140,7 +139,7 @@
return is_text(f)
-def _shebang_split(line: str) -> List[str]:
+def _shebang_split(line: str) -> list[str]:
try:
# shebangs aren't supposed to be quoted, though some tools such as
# setuptools will write them with quotes so we'll best-guess parse
@@ -154,8 +153,8 @@
def _parse_nix_shebang(
bytesio: IO[bytes],
- cmd: Tuple[str, ...],
-) -> Tuple[str, ...]:
+ cmd: tuple[str, ...],
+) -> tuple[str, ...]:
while bytesio.read(2) == b'#!':
next_line_b = bytesio.readline()
try:
@@ -176,7 +175,7 @@
return cmd
-def parse_shebang(bytesio: IO[bytes]) -> Tuple[str, ...]:
+def parse_shebang(bytesio: IO[bytes]) -> tuple[str, ...]:
"""Parse the shebang from a file opened for reading binary."""
if bytesio.read(2) != b'#!':
return ()
@@ -193,13 +192,17 @@
cmd = tuple(_shebang_split(first_line.strip()))
if cmd and cmd[0] == '/usr/bin/env':
- cmd = cmd[1:]
+ if cmd[1] == '-S':
+ cmd = cmd[2:]
+ else:
+ cmd = cmd[1:]
+
if cmd == ('nix-shell',):
return _parse_nix_shebang(bytesio, cmd)
return cmd
-def parse_shebang_from_file(path: str) -> Tuple[str, ...]:
+def parse_shebang_from_file(path: str) -> tuple[str, ...]:
"""Parse the shebang given a file path."""
if not os.path.lexists(path):
raise ValueError(f'{path} does not exist.')
@@ -226,7 +229,7 @@
return s.strip()
-def license_id(filename: str) -> Optional[str]:
+def license_id(filename: str) -> str | None:
"""Return the spdx id for the license contained in `filename`. If no
license is detected, returns `None`.
@@ -240,7 +243,7 @@
3. check exact text match with existing licenses
4. failing that use edit distance
"""
- import editdistance_s # `pip install identify[license]`
+ import ukkonen # `pip install identify[license]`
with open(filename, encoding='UTF-8') as f:
contents = f.read()
@@ -250,6 +253,8 @@
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)
@@ -260,13 +265,13 @@
if norm and abs(len(norm) - len(norm_license)) / len(norm) > .05:
continue
- edit_dist = editdistance_s.distance(norm, norm_license)
- if edit_dist < min_edit_dist:
+ edit_dist = ukkonen.distance(norm, norm_license, cutoff)
+ if edit_dist < cutoff and 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 / len(norm) < .05:
+ if norm and min_edit_dist < cutoff:
return min_edit_dist_spdx
else:
# no matches :'(
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/identify/interpreters.py
new/identify-2.5.31/identify/interpreters.py
--- old/identify-2.2.13/identify/interpreters.py 2021-08-06
18:07:44.000000000 +0200
+++ new/identify-2.5.31/identify/interpreters.py 2023-10-28
19:19:41.000000000 +0200
@@ -1,8 +1,10 @@
+from __future__ import annotations
INTERPRETERS = {
'ash': {'shell', 'ash'},
'awk': {'awk'},
'bash': {'shell', 'bash'},
'bats': {'shell', 'bash', 'bats'},
+ 'cbsd': {'shell', 'cbsd'},
'csh': {'shell', 'csh'},
'dash': {'shell', 'dash'},
'expect': {'expect'},
@@ -10,6 +12,9 @@
'node': {'javascript'},
'nodejs': {'javascript'},
'perl': {'perl'},
+ 'php': {'php'},
+ 'php7': {'php', 'php7'},
+ 'php8': {'php', 'php8'},
'python': {'python'},
'python2': {'python', 'python2'},
'python3': {'python', 'python3'},
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/identify/vendor/licenses.py
new/identify-2.5.31/identify/vendor/licenses.py
--- old/identify-2.2.13/identify/vendor/licenses.py 2021-08-06
18:07:44.000000000 +0200
+++ new/identify-2.5.31/identify/vendor/licenses.py 2023-10-28
19:19:41.000000000 +0200
@@ -1,3 +1,4 @@
+from __future__ import annotations
LICENSES = (
(
'0BSD',
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/setup.cfg
new/identify-2.5.31/setup.cfg
--- old/identify-2.2.13/setup.cfg 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/setup.cfg 2023-10-28 19:19:41.000000000 +0200
@@ -1,6 +1,6 @@
[metadata]
name = identify
-version = 2.2.13
+version = 2.5.31
description = File identification library for Python
long_description = file: README.md
long_description_content_type = text/markdown
@@ -8,21 +8,17 @@
author = Chris Kuehl
author_email = cku...@ocf.berkeley.edu
license = MIT
-license_file = LICENSE
+license_files = LICENSE
classifiers =
License :: OSI Approved :: MIT License
Programming Language :: Python :: 3
Programming Language :: Python :: 3 :: Only
- Programming Language :: Python :: 3.6
- Programming Language :: Python :: 3.7
- Programming Language :: Python :: 3.8
- Programming Language :: Python :: 3.9
Programming Language :: Python :: Implementation :: CPython
Programming Language :: Python :: Implementation :: PyPy
[options]
packages = find:
-python_requires = >=3.6.1
+python_requires = >=3.8
[options.packages.find]
exclude =
@@ -35,7 +31,11 @@
[options.extras_require]
license =
- editdistance-s
+ ukkonen
+
+[options.package_data]
+identify =
+ py.typed
[bdist_wheel]
universal = True
@@ -48,7 +48,6 @@
disallow_any_generics = true
disallow_incomplete_defs = true
disallow_untyped_defs = true
-no_implicit_optional = true
warn_redundant_casts = true
warn_unused_ignores = true
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/setup.py new/identify-2.5.31/setup.py
--- old/identify-2.2.13/setup.py 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/setup.py 2023-10-28 19:19:41.000000000 +0200
@@ -1,2 +1,4 @@
+from __future__ import annotations
+
from setuptools import setup
setup()
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/tests/cli_test.py
new/identify-2.5.31/tests/cli_test.py
--- old/identify-2.2.13/tests/cli_test.py 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/tests/cli_test.py 2023-10-28 19:19:41.000000000
+0200
@@ -1,3 +1,5 @@
+from __future__ import annotations
+
from identify import cli
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/tests/extensions_test.py
new/identify-2.5.31/tests/extensions_test.py
--- old/identify-2.2.13/tests/extensions_test.py 2021-08-06
18:07:44.000000000 +0200
+++ new/identify-2.5.31/tests/extensions_test.py 2023-10-28
19:19:41.000000000 +0200
@@ -1,3 +1,5 @@
+from __future__ import annotations
+
import pytest
from identify import extensions
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/tests/identify_test.py
new/identify-2.5.31/tests/identify_test.py
--- old/identify-2.2.13/tests/identify_test.py 2021-08-06 18:07:44.000000000
+0200
+++ new/identify-2.5.31/tests/identify_test.py 2023-10-28 19:19:41.000000000
+0200
@@ -1,3 +1,5 @@
+from __future__ import annotations
+
import builtins
import errno
import io
@@ -37,6 +39,7 @@
assert 'plist' in identify.ALL_TAGS # extension, needs binary check
assert 'dockerfile' in identify.ALL_TAGS # by file convention
assert 'python3' in identify.ALL_TAGS # by shebang
+ assert 'php8' in identify.ALL_TAGS # by shebang
def test_tags_from_path_does_not_exist(tmpdir):
@@ -151,6 +154,7 @@
('test.py', {'text', 'python'}),
('test.mk', {'text', 'makefile'}),
('Makefile', {'text', 'makefile'}),
+ ('Containerfile', {'text', 'dockerfile'}),
('Dockerfile', {'text', 'dockerfile'}),
('Dockerfile.xenial', {'text', 'dockerfile'}),
('xenial.Dockerfile', {'text', 'dockerfile'}),
@@ -158,11 +162,14 @@
('Pipfile.lock', {'text', 'json'}),
('mod/test.py', {'text', 'python'}),
('mod/Dockerfile', {'text', 'dockerfile'}),
+ ('config.ru', {'text', 'ruby'}),
('Gemfile', {'text', 'ruby'}),
('Gemfile.lock', {'text'}),
('Jenkinsfile', {'text', 'groovy', 'jenkins'}),
('build.jenkins', {'text', 'groovy', 'jenkins'}),
('build.jenkinsfile', {'text', 'groovy', 'jenkins'}),
+ ('meson.build', {'text', 'meson'}),
+ ('meson_options.txt', {'text', 'plain-text', 'meson'}),
# does not set binary / text
('f.plist', {'plist'}),
@@ -314,6 +321,8 @@
(b'\xf9\x93\x01\x42\xcd', ()),
(b'#!\xf9\x93\x01\x42\xcd', ()),
(b'#!\x00\x00\x00\x00', ()),
+ # shebang lines with multiple arguments
+ (b'#!/usr/bin/env -S python -u', ('python', '-u')),
),
)
def test_parse_shebang(s, expected):
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/identify-2.2.13/tox.ini new/identify-2.5.31/tox.ini
--- old/identify-2.2.13/tox.ini 2021-08-06 18:07:44.000000000 +0200
+++ new/identify-2.5.31/tox.ini 2023-10-28 19:19:41.000000000 +0200
@@ -1,5 +1,5 @@
[tox]
-envlist = py36,pypy3,pre-commit
+envlist = py,pre-commit
[testenv]
deps = -rrequirements-dev.txt
@@ -7,7 +7,7 @@
commands =
coverage erase
coverage run -m pytest {posargs:tests}
- coverage report --fail-under 100
+ coverage report
[testenv:pre-commit]
skip_install = true
