chore: new generator (#32)

Author: ldez

.github/workflows/ci.yml

@@ -36,6 +36,13 @@ jobs:
           git diff --exit-code go.mod
           git diff --exit-code go.sum
 
+      - name: Check generated files
+        run: |
+          make generate
+          git diff --exit-code words.go
+          git diff --exit-code words_uk.go
+          git diff --exit-code words_us.go
+
       - name: golangci-lint
         uses: golangci/golangci-lint-action@v7
         with:

CONTRIBUTING.md

@@ -0,0 +1,29 @@
+# Contributing
+
+The files `words.go`, `works_uk.go`, and `works_us.go` must never be edited by hand.
+
+## Adding a word
+
+Misspell is neither a complete spell-checking program nor a grammar checker.
+It is a tool to correct commonly misspelled English words.
+
+The list of words must contain only common mistakes.
+
+Before adding a word, you should have information about the misspelling frequency.
+
+- [ ] more than 15k inside GitHub (this limit is arbitrary and can involve)
+- [ ] don't exist inside the Wiktionary (as a modern form)
+- [ ] don't exist inside the Cambridge Dictionary (as a modern form)
+- [ ] don't exist inside the Oxford Dictionary (as a modern form)
+
+If all criteria are met, a word can be added to the list of misspellings.
+
+The word should be added to one of the following files.
+
+- `internal/gen/sources/main.json`: common words.
+- `internal/gen/sources/uk.json`: UK only words.
+- `internal/gen/sources/us.json`: US only words.
+
+The target `make generate` will generate the Go files.
+
+The PR description must provide all the information (links) about the misspelling frequency.

Makefile

@@ -1,6 +1,6 @@
 CONTAINER=golangci/misspell
 
-default: lint test build
+default: generate lint test build
 
 install:  ## install misspell into GOPATH/bin
 	go install ./cmd/misspell
@@ -14,6 +14,9 @@ test:  ## run all tests
 lint:  ## run linter
 	golangci-lint run
 
+generate:
+	go run ./internal/gen/
+
 # the grep in line 2 is to remove misspellings in the spelling dictionary
 # that trigger false positives!!
 falsepositives: /scowl-wl

internal/gen/dict.go.tmpl

@@ -0,0 +1,10 @@
+// Code generated by 'internal/gen'. DO NOT EDIT.
+
+package misspell
+
+// Dict{{ .Name }} {{ .Comment }}
+var Dict{{ .Name }} = []string{
+{{- range $index, $tuple := .Tuples }}
+	"{{ $tuple.Typo }}", "{{ $tuple.Correction }}",
+{{- end }}
+}

internal/gen/gen.go

@@ -0,0 +1,147 @@
+package main
+
+import (
+	"bytes"
+	"cmp"
+	_ "embed"
+	"encoding/json"
+	"go/format"
+	"log"
+	"os"
+	"slices"
+	"text/template"
+)
+
+//go:embed dict.go.tmpl
+var dictTemplate string
+
+type Tuple struct {
+	Typo       string
+	Correction string
+}
+
+type Info struct {
+	Name    string
+	Comment string
+	Path    string
+}
+
+// regenerate words Go files from JSON files.
+func main() {
+	dictionaries := map[string]Info{
+		"words.go": {
+			Name:    "Main",
+			Comment: "is the main rule set, not including locale-specific spellings",
+			Path:    "internal/gen/sources/main.json",
+		},
+		"words_uk.go": {
+			Name:    "British",
+			Comment: "converts US spellings to UK spellings",
+			Path:    "internal/gen/sources/uk.json",
+		},
+		"words_us.go": {
+			Name:    "American",
+			Comment: "converts UK spellings to US spellings",
+			Path:    "internal/gen/sources/us.json",
+		},
+	}
+
+	for dest, src := range dictionaries {
+		err := generate(src, dest)
+		if err != nil {
+			log.Fatal(err)
+		}
+	}
+}
+
+func generate(src Info, dest string) error {
+	data, err := read(src.Path)
+	if err != nil {
+		return err
+	}
+
+	tuples := toTuples(data)
+
+	return write(tuples, src, dest)
+}
+
+func toTuples(data map[string][]string) []Tuple {
+	var tuples []Tuple
+
+	for c, typos := range data {
+		for _, typo := range typos {
+			tuples = append(tuples, Tuple{Typo: typo, Correction: c})
+		}
+	}
+
+	return tuples
+}
+
+func read(src string) (map[string][]string, error) {
+	file, err := os.Open(src)
+	if err != nil {
+		return nil, err
+	}
+
+	defer func() {
+		_ = file.Close()
+	}()
+
+	all := make(map[string][]string)
+
+	err = json.NewDecoder(file).Decode(&all)
+	if err != nil {
+		return nil, err
+	}
+
+	return all, nil
+}
+
+func write(tuples []Tuple, src Info, dest string) error {
+	slices.SortStableFunc(tuples, func(a, b Tuple) int {
+		if len(a.Typo) == len(b.Typo) {
+			// if words are same size, then use
+			// normal alphabetical order
+			return cmp.Compare(a.Typo, b.Typo)
+		}
+		// INVERTED  -- biggest words first
+		return cmp.Compare(len(b.Typo), len(a.Typo))
+	})
+
+	tmpl, err := template.New("words").Parse(dictTemplate)
+	if err != nil {
+		return err
+	}
+
+	var buf bytes.Buffer
+
+	err = tmpl.Execute(&buf, map[string]any{
+		"Name":    src.Name,
+		"Comment": src.Comment,
+		"Tuples":  tuples,
+	})
+	if err != nil {
+		return err
+	}
+
+	source, err := format.Source(buf.Bytes())
+	if err != nil {
+		return err
+	}
+
+	words, err := os.Create(dest)
+	if err != nil {
+		return err
+	}
+
+	defer func() {
+		_ = words.Close()
+	}()
+
+	_, err = words.Write(source)
+	if err != nil {
+		return err
+	}
+
+	return nil
+}