Support for Oxford English (rebased) (#3772)

Co-authored-by: Eric Larson <larson.eric.d@gmail.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>

Author: 3 people

.github/workflows/autofix.yml

@@ -0,0 +1,19 @@
+name: autofix.ci
+
+on: # yamllint disable-line rule:truthy
+  pull_request:
+    types: [opened, synchronize, labeled, unlabeled]
+
+permissions:
+  contents: read
+
+jobs:
+  autofix:
+    name: Regenerate autogenerated files
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          persist-credentials: false
+      - run: make codespell_lib/data/dictionary_en_to_en-OX_AUTOGENERATED.txt
+      - uses: autofix-ci/action@635ffb0c9798bd160680f18fd73371e355b85f27

Makefile

@@ -1,16 +1,21 @@
-DICTIONARIES := codespell_lib/data/dictionary*.txt codespell_lib/tests/data/*.wordlist
+DICTIONARIES = codespell_lib/data/dictionary*.txt codespell_lib/tests/data/*.wordlist
 
 PHONY := all check check-dictionaries sort-dictionaries trim-dictionaries check-dist pytest pypi ruff clean
 
-all: check-dictionaries codespell.1
+GENERATED := codespell_lib/data/dictionary_en_to_en-OX_AUTOGENERATED.txt
 
-check: check-dictionaries check-dist pytest ruff
+all: $(GENERATED) check-dictionaries codespell.1
+
+check: $(GENERATED) check-dictionaries check-dist pytest ruff
+
+$(GENERATED): codespell_lib/data/dictionary_en-GB_to_en-US.txt ./tools/gen_OX.sh
+	./tools/gen_OX.sh codespell_lib/data/dictionary_en-GB_to_en-US.txt > codespell_lib/data/dictionary_en_to_en-OX_AUTOGENERATED.txt
 
 codespell.1: codespell.1.include Makefile
 	PYTHONPATH=. help2man codespell --include codespell.1.include --no-info --output codespell.1
 	sed -i '/\.SS \"Usage/,+2d' codespell.1
 
-check-dictionaries:
+check-dictionaries: sort-dictionaries
 	@for dictionary in ${DICTIONARIES}; do \
 		if grep -E -n "^\s*$$|\s$$|^\s" $$dictionary; then \
 			echo "Dictionary $$dictionary contains leading/trailing whitespace and/or blank lines.  Trim with 'make trim-dictionaries'"; \
@@ -24,10 +29,10 @@ check-dictionaries:
 		exit 1; \
 	fi
 
-sort-dictionaries:
+sort-dictionaries: $(GENERATED)
 	pre-commit run --all-files file-contents-sorter
 
-trim-dictionaries:
+trim-dictionaries: $(GENERATED)
 	@for dictionary in ${DICTIONARIES}; do \
 		sed -E -i.bak -e 's/^[[:space:]]+//; s/[[:space:]]+$$//; /^$$/d' $$dictionary && rm $$dictionary.bak; \
 	done
@@ -41,7 +46,7 @@ check-dist:
 ruff:
 	pre-commit run --all-files ruff-check
 
-pytest:
+pytest: sort-dictionaries
 	@if command -v pytest > /dev/null; then \
 		pytest codespell_lib; \
 	else \

codespell_lib/_codespell.py

@@ -123,6 +123,15 @@
         ("en_GB",),
         ("en_US",),
     ),
+    (
+        "en_to_en-OX",
+        "for corrections from en to en-OX",
+        "_en_to_en-OX_AUTOGENERATED",
+        None,
+        None,
+        None,
+        None,
+    ),
 )
 _builtin_default = "clear,rare"
 

codespell_lib/data/dictionary_en-GB_to_en-US.txt

@@ -69,6 +69,7 @@ characterise->characterize
 characterised->characterized
 characterises->characterizes
 characterising->characterizing
+civilisation->civilization
 cognisant->cognizant
 colour->color
 colouration->coloration
@@ -258,6 +259,7 @@ manoeuvres->maneuvers
 manoeuvring->maneuvering
 marshalled->marshaled
 marshalling->marshaling
+maths->math
 maximisation->maximization
 maximise->maximize
 maximised->maximized
@@ -359,7 +361,7 @@ polarises->polarizes
 polarising->polarizing
 practise->practice
 practised->practiced
-practises->practiced
+practises->practices
 practising->practicing
 pretence->pretense
 pretences->pretenses
@@ -368,6 +370,7 @@ prioritise->prioritize
 prioritised->prioritized
 prioritises->prioritizes
 prioritising->prioritizing
+programme->program
 publicise->publicize
 publicised->publicized
 publicises->publicizes
@@ -400,6 +403,8 @@ recognising->recognizing
 refocussed->refocused
 refocusses->refocuses
 refocussing->refocusing
+refuelled->refueled
+refuelling->refueling
 regularisation->regularization
 regularise->regularize
 regularised->regularized
@@ -420,6 +425,8 @@ reparameterised->reparameterized
 reparameterises->reparameterizes
 reparameterising->reparameterizing
 rigour->rigor
+routeing->routing
+saleable->salable
 sanitisation->sanitization
 sanitise->sanitize
 sanitised->sanitized
@@ -489,6 +496,8 @@ synthesiser->synthesizer
 synthesisers->synthesizers
 synthesises->synthesizes
 synthesising->synthesizing
+titbit->tidbit
+titbits->tidbits
 totalled->totaled
 totalling->totaling
 unauthorised->unauthorized