@kansei-link/bantou 0.2.0

package/data/tax-rules/jp-tax-rules-v1.json

@@ -0,0 +1,170 @@
+{
+  "version": "1.1.0",
+  "locale": "ja-JP",
+
+  "overseas_saas_providers": [
+    "aws", "amazon web services",
+    "google cloud", "gcp",
+    "azure", "microsoft azure",
+    "openai", "anthropic", "claude",
+    "cursor",
+    "github", "gitlab",
+    "cloudflare",
+    "vercel", "netlify", "heroku",
+    "digitalocean", "linode",
+    "stripe",
+    "figma", "adobe", "canva",
+    "slack", "discord",
+    "zoom",
+    "notion", "asana", "trello", "jira",
+    "dropbox", "datadog", "sentry",
+    "twilio", "sendgrid", "mailgun",
+    "hubspot", "intercom",
+    "1password", "lastpass",
+    "postman", "linear"
+  ],
+
+  "overseas_ad_platforms": [
+    "google ads", "google adwords", "adwords",
+    "meta", "facebook ads", "instagram ads",
+    "twitter ads", "x ads",
+    "linkedin ads",
+    "tiktok ads",
+    "youtube ads",
+    "pinterest ads",
+    "reddit ads",
+    "apple search ads"
+  ],
+
+  "domestic_telecom_providers": [
+    "docomo", "ドコモ",
+    "au", "kddi",
+    "softbank", "ソフトバンク",
+    "ntt", "ntt東日本", "ntt西日本", "nttコミュニケーションズ",
+    "さくらインターネット", "さくら",
+    "ocn",
+    "フレッツ", "flets",
+    "biglobe", "ビッグローブ",
+    "so-net", "ソネット",
+    "お名前.com", "onamae",
+    "iij", "インターネットイニシアティブ",
+    "楽天モバイル", "rakuten mobile",
+    "uqモバイル", "uq mobile",
+    "ワイモバイル", "y!mobile"
+  ],
+
+  "domestic_ad_platforms": [
+    "yahoo!広告", "yahoo広告", "yda", "ydn",
+    "line ads", "line広告",
+    "smartnews広告", "gunosy広告",
+    "indeed"
+  ],
+
+  "jct_indicator_keywords": [
+    "jct",
+    "消費税",
+    "インボイス",
+    "invoice",
+    "適格請求書"
+  ],
+
+  "asset_capitalization": {
+    "expense_max": 99999,
+    "lump_sum_3yr_max": 199999,
+    "sme_immediate_max": 299999
+  },
+
+  "withholding_tax": {
+    "bracket_1_ceiling": 1000000,
+    "rate_bracket_1": 0.1021,
+    "rate_bracket_2": 0.2042
+  },
+
+  "consumption_tax": {
+    "standard_rate": 10,
+    "reduced_rate": 8,
+    "exempt_rate": 0,
+
+    "newspaper_keywords": [
+      "新聞", "日経", "朝日", "読売", "毎日", "産経",
+      "newspaper", "nikkei"
+    ],
+
+    "food_beverage_keywords": [
+      "弁当", "ケータリング", "テイクアウト", "持ち帰り",
+      "出前", "デリバリー", "uber eats", "ubereats",
+      "出前館", "wolt", "menu",
+      "コンビニ", "スーパー"
+    ],
+
+    "takeout_delivery_keywords": [
+      "テイクアウト", "持ち帰り", "出前", "デリバリー",
+      "uber eats", "ubereats", "出前館", "wolt",
+      "menu", "弁当", "仕出し"
+    ],
+
+    "food_purchase_keywords": [
+      "コンビニ", "セブンイレブン", "ローソン", "ファミマ",
+      "スーパー", "イオン", "西友", "ライフ",
+      "成城石井", "紀ノ国屋", "オーケー",
+      "食品", "飲料", "お茶", "コーヒー豆",
+      "菓子", "お菓子", "スイーツ"
+    ],
+
+    "catering_with_service_keywords": [
+      "ケータリング", "出張料理", "パーティー",
+      "配膳", "サービス料"
+    ],
+
+    "residential_rent_keywords": [
+      "住居", "住宅", "マンション", "アパート",
+      "居住用", "住居用", "自宅", "社宅",
+      "residential"
+    ],
+
+    "non_taxable_categories": [
+      "salary",
+      "membership_fee",
+      "insurance",
+      "taxes_dues",
+      "donation",
+      "condolence"
+    ],
+
+    "non_taxable_reasons": {
+      "salary": "給与 — 消費税対象外",
+      "membership_fee": "諸会費 — 原則として消費税対象外",
+      "insurance": "保険料 — 消費税非課税(保険業法に基づく保険契約)",
+      "taxes_dues": "租税公課 — 消費税不課税(国・地方への公的支出)",
+      "donation": "寄付金 — 消費税不課税(対価性なし)",
+      "condolence": "慶弔費 — 消費税不課税(対価性なし)"
+    }
+  },
+
+  "invoice_system": {
+    "registration_prefix": "T",
+    "registration_digits": 13,
+    "transitional_periods": [
+      {
+        "start": "2023-10-01",
+        "end": "2026-09-30",
+        "deduction_rate": 80,
+        "label": "経過措置第1期 — 仕入税額控除80%"
+      },
+      {
+        "start": "2026-10-01",
+        "end": "2029-09-30",
+        "deduction_rate": 50,
+        "label": "経過措置第2期 — 仕入税額控除50%"
+      },
+      {
+        "start": "2029-10-01",
+        "end": "9999-12-31",
+        "deduction_rate": 0,
+        "label": "経過措置終了 — 仕入税額控除不可"
+      }
+    ],
+    "small_business_threshold": 10000,
+    "small_business_note": "1万円未満の課税仕入れは、少額特例により帳簿のみで仕入税額控除可(基準期間の課税売上高1億円以下の事業者、2029年9月まで)"
+  }
+}

package/dist/adapters/csv-parser.d.ts

@@ -0,0 +1,11 @@
+/**
+ * Parse CSV text into array of header-keyed row objects.
+ *
+ * @param csvText - Raw CSV text (UTF-8).
+ * @returns { headers, rows } where each row is Record<headerName, value>.
+ */
+export declare function parseCsv(csvText: string): {
+    headers: string[];
+    rows: Record<string, string>[];
+};
+//# sourceMappingURL=csv-parser.d.ts.map

package/dist/adapters/csv-parser.js

@@ -0,0 +1,133 @@
+// Simple CSV parser for Japanese accounting CSV files.
+//
+// Handles:
+//   - UTF-8 BOM (common in Japanese CSV exports)
+//   - Quoted fields with embedded commas and newlines
+//   - Empty lines
+//   - CRLF / LF line endings
+//
+// Note: Shift-JIS support is NOT included. Users must export as UTF-8
+// (弥生: エクスポート → 文字コード: UTF-8 を選択).
+/**
+ * Parse CSV text into array of header-keyed row objects.
+ *
+ * @param csvText - Raw CSV text (UTF-8).
+ * @returns { headers, rows } where each row is Record<headerName, value>.
+ */
+export function parseCsv(csvText) {
+    // Strip BOM
+    let text = csvText;
+    if (text.charCodeAt(0) === 0xFEFF) {
+        text = text.slice(1);
+    }
+    const lines = splitCsvLines(text);
+    if (lines.length === 0) {
+        return { headers: [], rows: [] };
+    }
+    const headers = parseCsvLine(lines[0]).map(h => h.trim());
+    const rows = [];
+    for (let i = 1; i < lines.length; i++) {
+        const line = lines[i].trim();
+        if (!line)
+            continue; // skip empty lines
+        const values = parseCsvLine(line);
+        const row = {};
+        for (let j = 0; j < headers.length; j++) {
+            row[headers[j]] = (values[j] ?? '').trim();
+        }
+        rows.push(row);
+    }
+    return { headers, rows };
+}
+/**
+ * Split CSV text into lines, respecting quoted fields that span multiple lines.
+ */
+function splitCsvLines(text) {
+    const lines = [];
+    let current = '';
+    let inQuote = false;
+    for (let i = 0; i < text.length; i++) {
+        const ch = text[i];
+        if (ch === '"') {
+            // Check for escaped quote ("")
+            if (inQuote && i + 1 < text.length && text[i + 1] === '"') {
+                current += '""';
+                i++; // skip next quote
+            }
+            else {
+                inQuote = !inQuote;
+                current += ch;
+            }
+        }
+        else if ((ch === '\n' || ch === '\r') && !inQuote) {
+            // Line break outside quotes = end of line
+            if (ch === '\r' && i + 1 < text.length && text[i + 1] === '\n') {
+                i++; // skip \n in \r\n
+            }
+            lines.push(current);
+            current = '';
+        }
+        else {
+            current += ch;
+        }
+    }
+    // Last line (no trailing newline)
+    if (current) {
+        lines.push(current);
+    }
+    return lines;
+}
+/**
+ * Parse a single CSV line into field values.
+ * Handles quoted fields and escaped double-quotes.
+ */
+function parseCsvLine(line) {
+    const fields = [];
+    let i = 0;
+    while (i <= line.length) {
+        if (i === line.length) {
+            // Trailing comma → empty field
+            break;
+        }
+        if (line[i] === '"') {
+            // Quoted field
+            let value = '';
+            i++; // skip opening quote
+            while (i < line.length) {
+                if (line[i] === '"') {
+                    if (i + 1 < line.length && line[i + 1] === '"') {
+                        value += '"';
+                        i += 2;
+                    }
+                    else {
+                        i++; // skip closing quote
+                        break;
+                    }
+                }
+                else {
+                    value += line[i];
+                    i++;
+                }
+            }
+            fields.push(value);
+            // Skip comma after quoted field
+            if (i < line.length && line[i] === ',') {
+                i++;
+            }
+        }
+        else {
+            // Unquoted field
+            const nextComma = line.indexOf(',', i);
+            if (nextComma === -1) {
+                fields.push(line.slice(i));
+                i = line.length;
+            }
+            else {
+                fields.push(line.slice(i, nextComma));
+                i = nextComma + 1;
+            }
+        }
+    }
+    return fields;
+}
+//# sourceMappingURL=csv-parser.js.map

package/dist/adapters/freee-csv-adapter.d.ts

@@ -0,0 +1,14 @@
+import { Transaction } from '../classifier/types.js';
+import { CsvAdapter, CsvSource } from './types.js';
+export declare class FreeeCsvAdapter implements CsvAdapter {
+    readonly source: CsvSource;
+    readonly label = "freee CSV \u30A8\u30AF\u30B9\u30DD\u30FC\u30C8";
+    detectFormat(headers: string[]): boolean;
+    parseRow(row: Record<string, string>, rowNumber: number): {
+        transaction: Transaction | null;
+        skip_reason: string | null;
+    };
+    private parseDate;
+    private parseAmount;
+}
+//# sourceMappingURL=freee-csv-adapter.d.ts.map

package/dist/adapters/freee-csv-adapter.js

@@ -0,0 +1,67 @@
+// freee CSV export adapter.
+//
+// Parses the freee 取引データ CSV export format.
+//
+// freee export path: 取引 → 一覧 → CSVエクスポート
+// Encoding: UTF-8 (freee defaults to UTF-8)
+//
+// Typical columns:
+//   収支区分, 取引日, 決済日, 取引先, 勘定科目, 税区分, 金額, 税額,
+//   備考, 品目, 部門, メモタグ, セグメント1, セグメント2, セグメント3
+const FREEE_REQUIRED = ['取引日', '金額'];
+const FREEE_DETECTION = ['勘定科目', '税区分', '収支区分'];
+export class FreeeCsvAdapter {
+    source = 'freee_export';
+    label = 'freee CSV エクスポート';
+    detectFormat(headers) {
+        const h = headers.map(s => s.trim());
+        // Must have 取引日 + 金額, and at least one freee-specific column
+        return (FREEE_REQUIRED.every(req => h.includes(req)) &&
+            FREEE_DETECTION.some(det => h.includes(det)));
+    }
+    parseRow(row, rowNumber) {
+        // Parse date
+        const dateStr = row['取引日']?.trim();
+        if (!dateStr) {
+            return { transaction: null, skip_reason: '取引日が空' };
+        }
+        const date = this.parseDate(dateStr);
+        if (!date) {
+            return { transaction: null, skip_reason: `日付パース失敗: "${dateStr}"` };
+        }
+        // Parse amount
+        const amountStr = row['金額']?.trim();
+        const amount = this.parseAmount(amountStr);
+        if (!amount || amount === 0) {
+            return { transaction: null, skip_reason: '金額が0またはパース失敗' };
+        }
+        // Build memo: 備考 or 勘定科目 or fallback
+        const bikou = row['備考']?.trim() || '';
+        const kanjou = row['勘定科目']?.trim() || '';
+        const memoTag = row['メモタグ']?.trim() || '';
+        const memo = bikou || kanjou || memoTag || `取引 ${date}`;
+        // Partner name
+        const partner_name = row['取引先']?.trim() || undefined;
+        return {
+            transaction: { amount, memo, date, partner_name },
+            skip_reason: null,
+        };
+    }
+    parseDate(dateStr) {
+        // freee uses "YYYY-MM-DD" or "YYYY/MM/DD"
+        const match = dateStr.match(/^(\d{4})[\/\-](\d{1,2})[\/\-](\d{1,2})$/);
+        if (match) {
+            const [, y, m, d] = match;
+            return `${y}-${m.padStart(2, '0')}-${d.padStart(2, '0')}`;
+        }
+        return null;
+    }
+    parseAmount(amountStr) {
+        if (!amountStr)
+            return 0;
+        const cleaned = amountStr.replace(/[,、￥¥\s]/g, '');
+        const num = parseInt(cleaned, 10);
+        return isNaN(num) ? 0 : Math.abs(num);
+    }
+}
+//# sourceMappingURL=freee-csv-adapter.js.map

package/dist/adapters/generic-adapter.d.ts

@@ -0,0 +1,20 @@
+import { Transaction } from '../classifier/types.js';
+import { CsvAdapter, CsvSource, ColumnMapping } from './types.js';
+export declare class GenericCsvAdapter implements CsvAdapter {
+    readonly source: CsvSource;
+    readonly label = "\u6C4E\u7528 CSV";
+    private mapping;
+    constructor(mapping: ColumnMapping);
+    /**
+     * Generic adapter always returns true for detectFormat
+     * (it's the fallback when no other adapter matches).
+     */
+    detectFormat(_headers: string[]): boolean;
+    parseRow(row: Record<string, string>, rowNumber: number): {
+        transaction: Transaction | null;
+        skip_reason: string | null;
+    };
+    private parseDate;
+    private parseAmount;
+}
+//# sourceMappingURL=generic-adapter.d.ts.map

package/dist/adapters/generic-adapter.js

@@ -0,0 +1,73 @@
+// Generic CSV adapter with user-specified column mapping.
+//
+// Used when the CSV doesn't match any known format (弥生/freee/MF).
+// User provides a ColumnMapping to specify which columns contain date, amount, memo.
+export class GenericCsvAdapter {
+    source = 'generic';
+    label = '汎用 CSV';
+    mapping;
+    constructor(mapping) {
+        this.mapping = mapping;
+    }
+    /**
+     * Generic adapter always returns true for detectFormat
+     * (it's the fallback when no other adapter matches).
+     */
+    detectFormat(_headers) {
+        return true;
+    }
+    parseRow(row, rowNumber) {
+        // Date
+        const dateStr = row[this.mapping.date]?.trim();
+        if (!dateStr) {
+            return { transaction: null, skip_reason: `日付列 "${this.mapping.date}" が空` };
+        }
+        const date = this.parseDate(dateStr);
+        if (!date) {
+            return { transaction: null, skip_reason: `日付パース失敗: "${dateStr}"` };
+        }
+        // Amount
+        const amountStr = row[this.mapping.amount]?.trim();
+        const amount = this.parseAmount(amountStr);
+        if (!amount || amount === 0) {
+            return { transaction: null, skip_reason: '金額が0またはパース失敗' };
+        }
+        // Memo
+        const memo = row[this.mapping.memo]?.trim() || '';
+        if (!memo) {
+            return { transaction: null, skip_reason: `摘要列 "${this.mapping.memo}" が空` };
+        }
+        // Partner (optional)
+        const partner_name = this.mapping.partner_name
+            ? row[this.mapping.partner_name]?.trim() || undefined
+            : undefined;
+        return {
+            transaction: { amount, memo, date, partner_name },
+            skip_reason: null,
+        };
+    }
+    parseDate(dateStr) {
+        // Try common date formats
+        // "YYYY/MM/DD", "YYYY-MM-DD", "YYYY.MM.DD"
+        const match = dateStr.match(/^(\d{4})[\/\-\.](\d{1,2})[\/\-\.](\d{1,2})$/);
+        if (match) {
+            const [, y, m, d] = match;
+            return `${y}-${m.padStart(2, '0')}-${d.padStart(2, '0')}`;
+        }
+        // "MM/DD/YYYY" (US format)
+        const usMatch = dateStr.match(/^(\d{1,2})[\/\-](\d{1,2})[\/\-](\d{4})$/);
+        if (usMatch) {
+            const [, m, d, y] = usMatch;
+            return `${y}-${m.padStart(2, '0')}-${d.padStart(2, '0')}`;
+        }
+        return null;
+    }
+    parseAmount(amountStr) {
+        if (!amountStr)
+            return 0;
+        const cleaned = amountStr.replace(/[,、￥¥$\s]/g, '');
+        const num = parseFloat(cleaned);
+        return isNaN(num) ? 0 : Math.abs(Math.round(num));
+    }
+}
+//# sourceMappingURL=generic-adapter.js.map

package/dist/adapters/index.d.ts

@@ -0,0 +1,23 @@
+import { CsvSource, ColumnMapping, ImportResult } from './types.js';
+import { TwoStageClassifier } from '../classifier/two-stage-classifier.js';
+import { ExclusionChecker } from '../exclusion/exclusion-checker.js';
+import { ConfidenceRouter } from '../pipeline/confidence-router.js';
+import { CockpitMemory } from '../memory/cockpit-memory.js';
+import { TaxRuleEngine } from '../tax-rules/tax-rule-engine.js';
+/**
+ * Import CSV and run through the full classification pipeline.
+ *
+ * @param csvText      - Raw CSV text (UTF-8).
+ * @param classifier   - TwoStageClassifier instance.
+ * @param exclusion    - ExclusionChecker instance.
+ * @param router       - ConfidenceRouter instance.
+ * @param opts.source  - Force a specific source format (skip auto-detection).
+ * @param opts.mapping - Column mapping for generic CSV.
+ */
+export declare function importCsv(csvText: string, classifier: TwoStageClassifier, exclusion: ExclusionChecker, router: ConfidenceRouter, opts?: {
+    source?: CsvSource;
+    mapping?: ColumnMapping;
+    memory?: CockpitMemory;
+    taxRuleEngine?: TaxRuleEngine;
+}): Promise<ImportResult>;
+//# sourceMappingURL=index.d.ts.map