[PATCH v6] devtools: add script to check for non inclusive naming

[Stephen Hemminger <stephen@networkplumber.org>]

Script to find words that should not be used. Really just a wrapper around git grep command.
By default it prints matches.

Uses the word lists from Inclusive Naming Initiative
see https://inclusivenaming.org/word-lists/
Note: the list has extra comma at end of list of elements which is not valid in basic
JSON but allowed in user-friendly JSON5 (https://json5.org/) and therefore needs the
PyPi package for parsing json5 format.

Examples:
$ ./devtools/check-naming-policy.sh -c
app/test-compress-perf/comp_perf_test_cyclecount.c:1
uapp/test-compress-perf/comp_perf_test_throughput.c:1
app/test-compress-perf/comp_perf_test_verify.c:1
app/test/test_common.c:1
...

$ ./devtools/check-naming-policy.py lib/pcapng
lib/pcapng/rte_pcapng.c:                /* sanity check that is really a pcapng mbuf */

$ ./devtools/check-naming-policy.py -l lib/eal
lib/eal/common/eal_common_memory.c
lib/eal/common/eal_common_proc.c
lib/eal/common/eal_common_trace.c
lib/eal/common/eal_memcfg.h
lib/eal/common/rte_malloc.c
lib/eal/freebsd/eal.c
lib/eal/include/generic/rte_power_intrinsics.h
lib/eal/include/generic/rte_rwlock.h
lib/eal/include/generic/rte_spinlock.h
lib/eal/include/rte_debug.h
lib/eal/include/rte_seqlock.h
lib/eal/linux/eal.c
lib/eal/windows/eal.c
lib/eal/x86/include/rte_rtm.h
lib/eal/x86/include/rte_spinlock.h
lib/eal/x86/rte_power_intrinsics.c

$ ./devtools/check-inclusive-naming -h
usage: check-inclusive-naming.py [-h] [-c] [-d] [-l] [-t {0,1,2,3}]
                                 [-x EXCLUDE] [--url URL]
                                 [paths ...]

Identify word usage not aligned with inclusive naming

positional arguments:
  paths                 files and directory to scan

options:
  -h, --help            show this help message and exit
  -c, --count           Show the nuber of lines that match
  -d, --debug           Debug this script
  -l, --files-with-matches
                        Show only names of files with hits
  -t {0,1,2,3}, --tier {0,1,2,3}
                        Show non-conforming words of particular tier
  -x EXCLUDE, --exclude EXCLUDE
                        Exclude path from scan
  --url URL             URL for the non-inclusive naming word list

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
v6 - missed some spelling and python lint in last patch

 MAINTAINERS                        |   1 +
 devtools/check-inclusive-naming.py | 127 +++++++++++++++++++++++++++++
 2 files changed, 128 insertions(+)
 create mode 100755 devtools/check-inclusive-naming.py

diff --git a/MAINTAINERS b/MAINTAINERS
index 4083658697..b53600ff51 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -89,6 +89,7 @@ F: devtools/check-doc-vs-code.sh
 F: devtools/check-dup-includes.sh
 F: devtools/check-maintainers.sh
 F: devtools/check-forbidden-tokens.awk
+F: devtools/check-inclusive-naming.py
 F: devtools/check-git-log.sh
 F: devtools/check-spdx-tag.sh
 F: devtools/check-symbol-change.sh
diff --git a/devtools/check-inclusive-naming.py b/devtools/check-inclusive-naming.py
new file mode 100755
index 0000000000..3d5cebf7b2
--- /dev/null
+++ b/devtools/check-inclusive-naming.py
@@ -0,0 +1,127 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright 2023 Stephen Hemminger
+#
+# This script scans the source tree and creates list of files
+# containing words that are recommended to be avoided by the
+# Inclusive Naming Initiative.
+# See: https://inclusivenaming.org/word-lists/
+
+import argparse
+import subprocess
+from urllib.request import urlopen
+
+# Need JSON5 to be able to handle extra comma
+import json5
+
+naming_url = 'https://inclusivenaming.org/word-lists/index.json'
+
+# These give false positives
+exception_list = [
+    'doc/guides/rel_notes/',
+    'doc/guides/contributing/coding_style.rst'
+    'doc/guides/prog_guide/glossary.rst'
+]
+
+
+def args_parse():
+    "parse arguments and return the argument object back to main"
+
+    parser = argparse.ArgumentParser(
+        description="Identify word usage not aligned with inclusive naming")
+    parser.add_argument("-c",
+                        "--count",
+                        help="Show the number of lines that match",
+                        action='store_true')
+    parser.add_argument("-d",
+                        "--debug",
+                        default=False,
+                        help="Debug this script",
+                        action='store_true')
+    parser.add_argument("-l",
+                        "--files-with-matches",
+                        help="Show only names of files with hits",
+                        action='store_true')
+    # note: tier 0 is "ok to use"
+    parser.add_argument("-t",
+                        "--tier",
+                        type=int,
+                        choices=range(0, 4),
+                        action='append',
+                        help="Show non-conforming words of particular tier")
+    parser.add_argument('-x',
+                        "--exclude",
+                        default=exception_list,
+                        action='append',
+                        help="Exclude path from scan")
+    parser.add_argument('--url',
+                        default=naming_url,
+                        help="URL for the non-inclusive naming word list")
+    parser.add_argument('paths', nargs='*',
+                        help='files and directory to scan')
+
+    return parser.parse_args()
+
+
+def fetch_wordlist(url, tiers):
+    "Read list of words from inclusivenaming.org"
+
+    # The wordlist is returned as JSON like:
+    # {
+    # "data" :
+    #         [
+    #             {
+    #                 "term": "abort",
+    #                 "tier" : "1",
+    #                 "recommendation": "Replace when possible.",
+    # ...
+    with urlopen(url) as response:
+        entries = json5.loads(response.read())['data']
+
+    wordlist = []
+    for item in entries:
+        tier = int(item['tier'])
+        if tiers.count(tier) > 0:
+            # convert minus sign to minus or space regex
+            pattern = item['term'].replace('-', '[- ]')
+            wordlist.append(pattern.lower())
+    return wordlist
+
+
+def process(args):
+    "Find matching words"
+
+    # Default to Tier 1, 2 and 3.
+    if args.tier:
+        tiers = args.tier
+    else:
+        tiers = list(range(1, 4))
+
+    wordlist = fetch_wordlist(args.url, tiers)
+    if args.debug:
+        print(f'Matching on {len(wordlist)} words')
+
+    cmd = ['git', 'grep', '-i']
+    if args.files_with_matches:
+        cmd.append('-l')
+    if args.count:
+        cmd.append('-c')
+    for word in wordlist:
+        cmd.append('-e')
+        cmd.append(word)
+    cmd.append('--')
+    # convert the exception_list to regexp
+    for path in exception_list:
+        cmd.append(f':^{path}')
+    cmd += args.paths
+    if args.debug:
+        print(cmd)
+    subprocess.run(cmd)
+
+
+def main():
+    process(args_parse())
+
+
+if __name__ == "__main__":
+    main()
-- 
2.41.0