finance-duck/internal/journal/codec.go

package journal

import (
	"bytes"
	"encoding/json"
	"fmt"
	"io"
	"reflect"
	"sort"
	"strings"
	"unicode/utf8"

	"finance-duck/internal/domain"
)

// Grammar: kind { on its own line, followed by field: JSON values, then }.
// JSON values may span lines. Blank lines and full-line # or // comments are
// permitted between fields and blocks. Strings use JSON escaping, including \n.
// registryFiles are the non-monthly journal files, in the order they are read
// and written. A block's file is its kind pluralized, so this list and the
// kinds accepted by parseDocument must stay in step.
var registryFiles = []string{"accounts.finance", "categories.finance", "tags.finance", "merchants.finance", "instruments.finance"}

type fieldSpan struct{ start, end int }
type block struct {
	kind, id string
	line     int
	lines    []string
	fields   map[string]fieldSpan
	value    any
}
type piece struct {
	text  string
	block *block
}
type document struct {
	path   string
	pieces []piece
}

func comment(s string) bool {
	s = strings.TrimSpace(s)
	return s == "" || strings.HasPrefix(s, "#") || strings.HasPrefix(s, "//")
}
func decodeStrict(raw []byte, value any) error {
	check := json.NewDecoder(bytes.NewReader(raw))
	check.UseNumber()
	if err := checkJSON(check); err != nil {
		return err
	}
	dec := json.NewDecoder(bytes.NewReader(raw))
	dec.DisallowUnknownFields()
	if err := dec.Decode(value); err != nil {
		return err
	}
	var extra any
	if err := dec.Decode(&extra); err != io.EOF {
		return fmt.Errorf("expected one JSON value")
	}
	return nil
}
func checkJSON(dec *json.Decoder) error {
	token, err := dec.Token()
	if err != nil {
		return err
	}
	delimiter, ok := token.(json.Delim)
	if !ok {
		return nil
	}
	switch delimiter {
	case '{':
		keys := map[string]bool{}
		for dec.More() {
			key, err := dec.Token()
			if err != nil {
				return err
			}
			name, ok := key.(string)
			if !ok {
				return fmt.Errorf("expected JSON object key")
			}
			if keys[name] {
				return fmt.Errorf("duplicate JSON key %q", name)
			}
			keys[name] = true
			if err = checkJSON(dec); err != nil {
				return err
			}
		}
	case '[':
		for dec.More() {
			if err = checkJSON(dec); err != nil {
				return err
			}
		}
	default:
		return fmt.Errorf("unexpected JSON delimiter")
	}
	_, err = dec.Token()
	return err
}
func fieldsOf(value any) (map[string]json.RawMessage, []string, error) {
	raw, err := json.Marshal(value)
	if err != nil {
		return nil, nil, err
	}
	m := map[string]json.RawMessage{}
	if err = json.Unmarshal(raw, &m); err != nil {
		return nil, nil, err
	}
	typ := reflect.TypeOf(value)
	keys := []string{}
	for i := range typ.NumField() {
		key := strings.Split(typ.Field(i).Tag.Get("json"), ",")[0]
		if _, ok := m[key]; ok {
			keys = append(keys, key)
		}
	}
	return m, keys, nil
}
func parseDocument(path string, raw []byte) (*document, error) {
	fail := func(line int, err any) (*document, error) { return nil, fmt.Errorf("%s:%d: %v", path, line, err) }
	if !utf8.Valid(raw) {
		return fail(1, "file is not valid UTF-8")
	}
	lines := strings.SplitAfter(string(raw), "\n")
	doc := &document{path: path}
	pending := ""
	for i := 0; i < len(lines); {
		trimmed := strings.TrimSpace(lines[i])
		if comment(trimmed) {
			pending += lines[i]
			i++
			continue
		}
		if pending != "" {
			doc.pieces = append(doc.pieces, piece{text: pending})
			pending = ""
		}
		header := strings.Fields(trimmed)
		if len(header) != 2 || header[1] != "{" {
			return fail(i+1, "expected 'account|category|tag|merchant|instrument|transaction {'")
		}
		kind := header[0]
		var value any
		switch kind {
		case "account":
			value = &domain.Account{}
		case "category":
			value = &domain.Category{}
		case "tag":
			value = &domain.Tag{}
		case "merchant":
			value = &domain.Merchant{}
		case "instrument":
			value = &domain.Instrument{}
		case "transaction":
			value = &domain.Transaction{}
		default:
			return fail(i+1, "unknown block kind "+kind)
		}
		fieldTypes := map[string]reflect.StructField{}
		typ := reflect.TypeOf(value).Elem()
		for n := range typ.NumField() {
			field := typ.Field(n)
			fieldTypes[strings.Split(field.Tag.Get("json"), ",")[0]] = field
		}
		start := i
		i++
		fields := map[string]fieldSpan{}
		closed := false
		for i < len(lines) {
			s := strings.TrimSpace(lines[i])
			if s == "}" {
				i++
				closed = true
				break
			}
			if comment(s) {
				i++
				continue
			}
			colon := strings.Index(s, ":")
			if colon <= 0 {
				return fail(i+1, "expected field: JSON")
			}
			key := strings.TrimSpace(s[:colon])
			if strings.ContainsAny(key, " \t\"{}") {
				return fail(i+1, "invalid field name")
			}
			if _, exists := fields[key]; exists {
				return fail(i+1, "duplicate field "+key)
			}
			fieldStart := i
			jsonText := strings.TrimSpace(s[colon+1:])
			for !json.Valid([]byte(jsonText)) {
				if jsonText != "" {
					var probe any
					err := json.Unmarshal([]byte(jsonText), &probe)
					if err != nil && !strings.Contains(err.Error(), "unexpected end of JSON input") {
						return fail(fieldStart+1, "field "+key+": "+err.Error())
					}
				}
				i++
				if i >= len(lines) {
					return fail(fieldStart+1, "unterminated JSON value for "+key)
				}
				jsonText += "\n" + strings.TrimSuffix(lines[i], "\n")
			}
			fieldType, known := fieldTypes[key]
			if !known {
				return fail(fieldStart+1, "unknown field "+key)
			}
			fieldValue := reflect.New(fieldType.Type)
			if err := decodeStrict([]byte(jsonText), fieldValue.Interface()); err != nil {
				return fail(fieldStart+1, "field "+key+": "+err.Error())
			}
			reflect.ValueOf(value).Elem().FieldByIndex(fieldType.Index).Set(fieldValue.Elem())
			fields[key] = fieldSpan{start: fieldStart - start, end: i - start}
			i++
		}
		if !closed {
			return fail(start+1, "unterminated block")
		}
		b := &block{kind: kind, line: start + 1, lines: append([]string{}, lines[start:i]...), fields: fields}
		switch v := value.(type) {
		case *domain.Account:
			b.id = v.ID
			b.value = *v
		case *domain.Category:
			b.id = v.ID
			b.value = *v
		case *domain.Tag:
			b.id = v.ID
			b.value = *v
		case *domain.Instrument:
			b.id = v.ID
			b.value = *v
		case *domain.Merchant:
			if v.Aliases == nil {
				v.Aliases = []string{}
			}
			if v.DefaultTagIDs == nil {
				v.DefaultTagIDs = []string{}
			}
			b.id = v.ID
			b.value = *v
		case *domain.Transaction:
			if v.Enrichment.TagIDs == nil {
				v.Enrichment.TagIDs = []string{}
			}
			b.id = v.Facts.ID
			b.value = *v
		}
		doc.pieces = append(doc.pieces, piece{block: b})
	}
	if pending != "" {
		doc.pieces = append(doc.pieces, piece{text: pending})
	}
	return doc, nil
}
func renderNew(kind string, value any) ([]byte, error) {
	fields, keys, err := fieldsOf(value)
	if err != nil {
		return nil, err
	}
	var out strings.Builder
	out.WriteString(kind + " {\n")
	for _, key := range keys {
		out.WriteString("  " + key + ": " + string(fields[key]) + "\n")
	}
	out.WriteString("}\n")
	return []byte(out.String()), nil
}
func (b *block) render(value any) ([]byte, error) {
	if reflect.DeepEqual(b.value, value) {
		return []byte(strings.Join(b.lines, "")), nil
	}
	fields, keys, err := fieldsOf(value)
	if err != nil {
		return nil, err
	}
	oldFields, _, err := fieldsOf(b.value)
	if err != nil {
		return nil, err
	}
	starts := map[int]string{}
	for key, f := range b.fields {
		starts[f.start] = key
	}
	var out strings.Builder
	for i := 0; i < len(b.lines); i++ {
		if i == len(b.lines)-1 {
			for _, key := range keys {
				if _, ok := b.fields[key]; !ok {
					out.WriteString("  " + key + ": " + string(fields[key]) + "\n")
				}
			}
		}
		key, ok := starts[i]
		if !ok {
			out.WriteString(b.lines[i])
			continue
		}
		f := b.fields[key]
		next, exists := fields[key]
		if exists {
			if bytes.Equal(oldFields[key], next) {
				out.WriteString(strings.Join(b.lines[i:f.end+1], ""))
			} else {
				out.WriteString("  " + key + ": " + string(next) + "\n")
			}
		}
		i = f.end
	}
	return []byte(out.String()), nil
}
func datasetFiles(d domain.Dataset) map[string]map[string]piece {
	files := map[string]map[string]piece{}
	for _, p := range registryFiles {
		files[p] = map[string]piece{}
	}
	add := func(path, kind, id string, value any) {
		if files[path] == nil {
			files[path] = map[string]piece{}
		}
		files[path][id] = piece{block: &block{kind: kind, id: id, value: value}}
	}
	for _, v := range d.Accounts {
		add("accounts.finance", "account", v.ID, v)
	}
	for _, v := range d.Categories {
		add("categories.finance", "category", v.ID, v)
	}
	for _, v := range d.Tags {
		add("tags.finance", "tag", v.ID, v)
	}
	for _, v := range d.Merchants {
		add("merchants.finance", "merchant", v.ID, v)
	}
	for _, v := range d.Instruments {
		add("instruments.finance", "instrument", v.ID, v)
	}
	for _, v := range d.Transactions {
		month := v.Facts.BookingDate[:7]
		add("journal/"+month[:4]+"/"+month+".finance", "transaction", v.Facts.ID, v)
	}
	return files
}
func renderFiles(d domain.Dataset, docs map[string]*document) (map[string][]byte, error) {
	wanted := datasetFiles(d)
	for path := range docs {
		if wanted[path] == nil {
			wanted[path] = map[string]piece{}
		}
	}
	out := map[string][]byte{}
	for path, blocks := range wanted {
		var buf bytes.Buffer
		if doc := docs[path]; doc != nil {
			for _, p := range doc.pieces {
				if p.block == nil {
					buf.WriteString(p.text)
					continue
				}
				b := p.block
				if next, ok := blocks[b.id]; ok {
					raw, err := b.render(next.block.value)
					if err != nil {
						return nil, err
					}
					buf.Write(raw)
					delete(blocks, b.id)
				} else {
					for _, line := range b.lines {
						if comment(line) {
							buf.WriteString(line)
						}
					}
				}
			}
		}
		ids := make([]string, 0, len(blocks))
		for id := range blocks {
			ids = append(ids, id)
		}
		sort.Strings(ids)
		for _, id := range ids {
			if buf.Len() > 0 && !bytes.HasSuffix(buf.Bytes(), []byte("\n")) {
				buf.WriteByte('\n')
			}
			p := blocks[id]
			raw, err := renderNew(p.block.kind, p.block.value)
			if err != nil {
				return nil, err
			}
			buf.Write(raw)
		}
		out[path] = buf.Bytes()
	}
	return out, nil
}