0xkobold 0.1.0 → 0.3.0

package/dist/src/config/manager.js

@@ -0,0 +1,222 @@
+/**
+ * Configuration Manager - v0.3.0
+ *
+ * Centralized config management with validation.
+ */
+import * as fs from "node:fs/promises";
+import * as path from "node:path";
+const DEFAULT_CONFIG = {
+    version: "0.3.0",
+    persona: {
+        name: "0xKobold",
+        emoji: "🐉",
+        description: "Your personal AI assistant",
+    },
+    heartbeat: {
+        enabled: true,
+        morning: "09:00",
+        evening: "18:00",
+        idleMinutes: 30,
+    },
+    gateway: {
+        enabled: true,
+        port: 7777,
+        host: "localhost",
+        cors: ["http://localhost:3000"],
+        remote: {
+            enabled: false,
+            url: undefined,
+            token: undefined,
+            password: undefined,
+            autoReconnect: true,
+            reconnectDelay: 1000,
+        },
+    },
+    channels: {
+        discord: { enabled: false },
+        whatsapp: { enabled: false, sessionPath: "~/.0xkobold/whatsapp-session" },
+        telegram: { enabled: false, mode: "polling" },
+        slack: { enabled: false },
+    },
+    security: {
+        deviceAuth: true,
+        dockerSandbox: true,
+        tokenExpiryHours: 168,
+    },
+    media: {
+        vision: { enabled: false, provider: "claude" },
+        audio: { enabled: false, provider: "openai" },
+        pdf: { enabled: true, maxPages: 100 },
+    },
+    llm: {
+        provider: "claude",
+        model: "claude-3-sonnet-20240229",
+        maxTokens: 4000,
+        temperature: 0.7,
+    },
+};
+class ConfigManager {
+    config;
+    configPath;
+    constructor(configPath = path.join(process.env.HOME || "~", ".0xkobold", "config.json")) {
+        this.configPath = configPath;
+        this.config = { ...DEFAULT_CONFIG };
+    }
+    /**
+     * Load configuration from file
+     */
+    async load() {
+        try {
+            const content = await fs.readFile(this.configPath, "utf-8");
+            const loaded = JSON.parse(content);
+            this.config = this.mergeConfig(DEFAULT_CONFIG, loaded);
+            return this.config;
+        }
+        catch {
+            // No config file yet, use defaults
+            return this.config;
+        }
+    }
+    /**
+     * Save configuration to file
+     */
+    async save() {
+        await fs.mkdir(path.dirname(this.configPath), { recursive: true });
+        await fs.writeFile(this.configPath, JSON.stringify(this.config, null, 2), "utf-8");
+    }
+    /**
+     * Get config value
+     */
+    get(key) {
+        return this.config[key];
+    }
+    /**
+     * Set config value
+     */
+    set(key, value) {
+        this.config[key] = value;
+    }
+    /**
+     * Get nested config
+     */
+    getNested(path) {
+        return path.split(".").reduce((obj, key) => obj?.[key], this.config);
+    }
+    /**
+     * Set nested config
+     */
+    setNested(path, value) {
+        const keys = path.split(".");
+        const lastKey = keys.pop();
+        const target = keys.reduce((obj, key) => {
+            if (!obj[key])
+                obj[key] = {};
+            return obj[key];
+        }, this.config);
+        target[lastKey] = value;
+    }
+    /**
+     * Validate configuration
+     */
+    validate() {
+        const errors = [];
+        // Check required fields
+        if (!this.config.version)
+            errors.push("version is required");
+        if (!this.config.persona.name)
+            errors.push("persona.name is required");
+        // Check LLM config
+        if (!this.config.llm.model)
+            errors.push("llm.model is required");
+        if (this.config.llm.temperature < 0 || this.config.llm.temperature > 2) {
+            errors.push("llm.temperature must be between 0 and 2");
+        }
+        // Check port range
+        if (this.config.gateway.port < 1 || this.config.gateway.port > 65535) {
+            errors.push("gateway.port must be between 1 and 65535");
+        }
+        return { valid: errors.length === 0, errors };
+    }
+    /**
+     * Get full config
+     */
+    getConfig() {
+        return { ...this.config };
+    }
+    /**
+     * Reset to defaults
+     */
+    reset() {
+        this.config = { ...DEFAULT_CONFIG };
+    }
+    /**
+     * Merge configs (deep merge)
+     */
+    mergeConfig(defaults, loaded) {
+        const result = { ...defaults };
+        for (const key in loaded) {
+            if (loaded[key] !== null && typeof loaded[key] === "object" && !Array.isArray(loaded[key])) {
+                result[key] = this.mergeConfig(defaults[key] || {}, loaded[key]);
+            }
+            else {
+                result[key] = loaded[key];
+            }
+        }
+        return result;
+    }
+    /**
+     * Export as environment variables
+     */
+    async exportToEnv() {
+        const lines = [];
+        if (this.config.channels.telegram?.token) {
+            lines.push(`TELEGRAM_BOT_TOKEN=${this.config.channels.telegram.token}`);
+        }
+        if (this.config.channels.discord?.token) {
+            lines.push(`DISCORD_BOT_TOKEN=${this.config.channels.discord.token}`);
+        }
+        if (this.config.channels.slack?.webhookUrl) {
+            lines.push(`SLACK_WEBHOOK_URL=${this.config.channels.slack.webhookUrl}`);
+        }
+        if (this.config.llm.apiKey) {
+            lines.push(`LLM_API_KEY=${this.config.llm.apiKey}`);
+        }
+        return lines.join("\n");
+    }
+    /**
+     * Import from environment
+     */
+    importFromEnv() {
+        if (process.env.TELEGRAM_BOT_TOKEN) {
+            this.config.channels.telegram ??= { enabled: true, mode: "polling" };
+            this.config.channels.telegram.token = process.env.TELEGRAM_BOT_TOKEN;
+            this.config.channels.telegram.enabled = true;
+        }
+        if (process.env.DISCORD_BOT_TOKEN) {
+            this.config.channels.discord ??= { enabled: true };
+            this.config.channels.discord.token = process.env.DISCORD_BOT_TOKEN;
+            this.config.channels.discord.enabled = true;
+        }
+        if (process.env.ANTHROPIC_API_KEY) {
+            this.config.llm.apiKey = process.env.ANTHROPIC_API_KEY;
+        }
+        if (process.env.OPENAI_API_KEY) {
+            this.config.media.audio ??= { enabled: false, provider: "openai" };
+            this.config.media.audio.apiKey = process.env.OPENAI_API_KEY;
+        }
+    }
+}
+// Singleton
+let instance = null;
+export function getConfigManager(configPath) {
+    if (!instance) {
+        instance = new ConfigManager(configPath);
+    }
+    return instance;
+}
+export function resetConfigManager() {
+    instance = null;
+}
+export { ConfigManager, DEFAULT_CONFIG };
+export default ConfigManager;
+//# sourceMappingURL=manager.js.map

package/dist/src/config/manager.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"manager.js","sourceRoot":"","sources":["../../../src/config/manager.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,KAAK,EAAE,MAAM,kBAAkB,CAAC;AACvC,OAAO,KAAK,IAAI,MAAM,WAAW,CAAC;AAuElC,MAAM,cAAc,GAAiB;IACnC,OAAO,EAAE,OAAO;IAEhB,OAAO,EAAE;QACP,IAAI,EAAE,UAAU;QAChB,KAAK,EAAE,IAAI;QACX,WAAW,EAAE,4BAA4B;KAC1C;IAED,SAAS,EAAE;QACT,OAAO,EAAE,IAAI;QACb,OAAO,EAAE,OAAO;QAChB,OAAO,EAAE,OAAO;QAChB,WAAW,EAAE,EAAE;KAChB;IAED,OAAO,EAAE;QACP,OAAO,EAAE,IAAI;QACb,IAAI,EAAE,IAAI;QACV,IAAI,EAAE,WAAW;QACjB,IAAI,EAAE,CAAC,uBAAuB,CAAC;QAC/B,MAAM,EAAE;YACN,OAAO,EAAE,KAAK;YACd,GAAG,EAAE,SAAS;YACd,KAAK,EAAE,SAAS;YAChB,QAAQ,EAAE,SAAS;YACnB,aAAa,EAAE,IAAI;YACnB,cAAc,EAAE,IAAI;SACrB;KACF;IAED,QAAQ,EAAE;QACR,OAAO,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE;QAC3B,QAAQ,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE,WAAW,EAAE,8BAA8B,EAAE;QACzE,QAAQ,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE,IAAI,EAAE,SAAS,EAAE;QAC7C,KAAK,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE;KAC1B;IAED,QAAQ,EAAE;QACR,UAAU,EAAE,IAAI;QAChB,aAAa,EAAE,IAAI;QACnB,gBAAgB,EAAE,GAAG;KACtB;IAED,KAAK,EAAE;QACL,MAAM,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,QAAQ,EAAE;QAC9C,KAAK,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,QAAQ,EAAE;QAC7C,GAAG,EAAE,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,GAAG,EAAE;KACtC;IAED,GAAG,EAAE;QACH,QAAQ,EAAE,QAAQ;QAClB,KAAK,EAAE,0BAA0B;QACjC,SAAS,EAAE,IAAI;QACf,WAAW,EAAE,GAAG;KACjB;CACF,CAAC;AAEF,MAAM,aAAa;IACT,MAAM,CAAe;IACrB,UAAU,CAAS;IAE3B,YAAY,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,IAAI,GAAG,EAAE,WAAW,EAAE,aAAa,CAAC;QACrF,IAAI,CAAC,UAAU,GAAG,UAAU,CAAC;QAC7B,IAAI,CAAC,MAAM,GAAG,EAAE,GAAG,cAAc,EAAE,CAAC;IACtC,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,IAAI;QACR,IAAI,CAAC;YACH,MAAM,OAAO,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,EAAE,OAAO,CAAC,CAAC;YAC5D,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;YACnC,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,WAAW,CAAC,cAAc,EAAE,MAAM,CAAC,CAAC;YACvD,OAAO,IAAI,CAAC,MAAM,CAAC;QACrB,CAAC;QAAC,MAAM,CAAC;YACP,mCAAmC;YACnC,OAAO,IAAI,CAAC,MAAM,CAAC;QACrB,CAAC;IACH,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,IAAI;QACR,MAAM,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,UAAU,CAAC,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QACnE,MAAM,EAAE,CAAC,SAAS,CAChB,IAAI,CAAC,UAAU,EACf,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,EACpC,OAAO,CACR,CAAC;IACJ,CAAC;IAED;;OAEG;IACH,GAAG,CAA+B,GAAM;QACtC,OAAO,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC;IAC1B,CAAC;IAED;;OAEG;IACH,GAAG,CAA+B,GAAM,EAAE,KAAsB;QAC9D,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC;IAC3B,CAAC;IAED;;OAEG;IACH,SAAS,CAAC,IAAY;QACpB,OAAO,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,GAAQ,EAAE,GAAG,EAAE,EAAE,CAAC,GAAG,EAAE,CAAC,GAAG,CAAC,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;IAC5E,CAAC;IAED;;OAEG;IACH,SAAS,CAAC,IAAY,EAAE,KAAc;QACpC,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QAC7B,MAAM,OAAO,GAAG,IAAI,CAAC,GAAG,EAAG,CAAC;QAC5B,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,GAAQ,EAAE,GAAG,EAAE,EAAE;YAC3C,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;gBAAE,GAAG,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC;YAC7B,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;QAClB,CAAC,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;QAChB,MAAM,CAAC,OAAO,CAAC,GAAG,KAAK,CAAC;IAC1B,CAAC;IAED;;OAEG;IACH,QAAQ;QACN,MAAM,MAAM,GAAa,EAAE,CAAC;QAE5B,wBAAwB;QACxB,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,OAAO;YAAE,MAAM,CAAC,IAAI,CAAC,qBAAqB,CAAC,CAAC;QAC7D,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC,IAAI;YAAE,MAAM,CAAC,IAAI,CAAC,0BAA0B,CAAC,CAAC;QAEvE,mBAAmB;QACnB,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,KAAK;YAAE,MAAM,CAAC,IAAI,CAAC,uBAAuB,CAAC,CAAC;QACjE,IAAI,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,WAAW,GAAG,CAAC,IAAI,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,WAAW,GAAG,CAAC,EAAE,CAAC;YACvE,MAAM,CAAC,IAAI,CAAC,yCAAyC,CAAC,CAAC;QACzD,CAAC;QAED,mBAAmB;QACnB,IAAI,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC,IAAI,GAAG,CAAC,IAAI,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC,IAAI,GAAG,KAAK,EAAE,CAAC;YACrE,MAAM,CAAC,IAAI,CAAC,0CAA0C,CAAC,CAAC;QAC1D,CAAC;QAED,OAAO,EAAE,KAAK,EAAE,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,MAAM,EAAE,CAAC;IAChD,CAAC;IAED;;OAEG;IACH,SAAS;QACP,OAAO,EAAE,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;IAC5B,CAAC;IAED;;OAEG;IACH,KAAK;QACH,IAAI,CAAC,MAAM,GAAG,EAAE,GAAG,cAAc,EAAE,CAAC;IACtC,CAAC;IAED;;OAEG;IACK,WAAW,CAAC,QAAa,EAAE,MAAW;QAC5C,MAAM,MAAM,GAAG,EAAE,GAAG,QAAQ,EAAE,CAAC;QAE/B,KAAK,MAAM,GAAG,IAAI,MAAM,EAAE,CAAC;YACzB,IAAI,MAAM,CAAC,GAAG,CAAC,KAAK,IAAI,IAAI,OAAO,MAAM,CAAC,GAAG,CAAC,KAAK,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC;gBAC3F,MAAM,CAAC,GAAG,CAAC,GAAG,IAAI,CAAC,WAAW,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,EAAE,EAAE,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC;YACnE,CAAC;iBAAM,CAAC;gBACN,MAAM,CAAC,GAAG,CAAC,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;YAC5B,CAAC;QACH,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,WAAW;QACf,MAAM,KAAK,GAAa,EAAE,CAAC;QAE3B,IAAI,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,EAAE,KAAK,EAAE,CAAC;YACzC,KAAK,CAAC,IAAI,CAAC,sBAAsB,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,CAAC,CAAC;QAC1E,CAAC;QACD,IAAI,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,KAAK,EAAE,CAAC;YACxC,KAAK,CAAC,IAAI,CAAC,qBAAqB,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,KAAK,EAAE,CAAC,CAAC;QACxE,CAAC;QACD,IAAI,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,EAAE,UAAU,EAAE,CAAC;YAC3C,KAAK,CAAC,IAAI,CAAC,qBAAqB,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,UAAU,EAAE,CAAC,CAAC;QAC3E,CAAC;QACD,IAAI,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC;YAC3B,KAAK,CAAC,IAAI,CAAC,eAAe,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,CAAC;QACtD,CAAC;QAED,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC1B,CAAC;IAED;;OAEG;IACH,aAAa;QACX,IAAI,OAAO,CAAC,GAAG,CAAC,kBAAkB,EAAE,CAAC;YACnC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,IAAI,EAAE,SAAS,EAAE,CAAC;YACrE,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,kBAAkB,CAAC;YACrE,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,OAAO,GAAG,IAAI,CAAC;QAC/C,CAAC;QAED,IAAI,OAAO,CAAC,GAAG,CAAC,iBAAiB,EAAE,CAAC;YAClC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;YACnD,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,iBAAiB,CAAC;YACnE,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,OAAO,GAAG,IAAI,CAAC;QAC9C,CAAC;QAED,IAAI,OAAO,CAAC,GAAG,CAAC,iBAAiB,EAAE,CAAC;YAClC,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,iBAAiB,CAAC;QACzD,CAAC;QAED,IAAI,OAAO,CAAC,GAAG,CAAC,cAAc,EAAE,CAAC;YAC/B,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,QAAQ,EAAE,CAAC;YACnE,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC;QAC9D,CAAC;IACH,CAAC;CACF;AAED,YAAY;AACZ,IAAI,QAAQ,GAAyB,IAAI,CAAC;AAE1C,MAAM,UAAU,gBAAgB,CAAC,UAAmB;IAClD,IAAI,CAAC,QAAQ,EAAE,CAAC;QACd,QAAQ,GAAG,IAAI,aAAa,CAAC,UAAU,CAAC,CAAC;IAC3C,CAAC;IACD,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,MAAM,UAAU,kBAAkB;IAChC,QAAQ,GAAG,IAAI,CAAC;AAClB,CAAC;AAED,OAAO,EAAE,aAAa,EAAE,cAAc,EAAE,CAAC;AACzC,eAAe,aAAa,CAAC"}

package/dist/src/documents/index.js

@@ -0,0 +1,3 @@
+// Documents Module - v0.3.0
+export { PDFExtractor, extractPDF, extractPDFText, } from "./pdf.js";
+//# sourceMappingURL=index.js.map

package/dist/src/documents/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../src/documents/index.ts"],"names":[],"mappings":"AAAA,4BAA4B;AAC5B,OAAO,EACL,YAAY,EACZ,UAAU,EACV,cAAc,GAGf,MAAM,UAAU,CAAC"}

package/dist/src/documents/pdf.js

@@ -0,0 +1,168 @@
+/**
+ * PDF Document Support - v0.3.0
+ *
+ * Text extraction from PDF documents.
+ */
+import * as fs from "node:fs/promises";
+class PDFExtractor {
+    config;
+    constructor(config = {}) {
+        this.config = {
+            maxPages: 100,
+            extractText: true,
+            extractMetadata: true,
+            ...config,
+        };
+    }
+    /**
+     * Extract from file
+     */
+    async extractFromFile(filePath) {
+        const buffer = await fs.readFile(filePath);
+        return this.extractFromBuffer(buffer);
+    }
+    /**
+     * Extract from buffer
+     */
+    async extractFromBuffer(buffer) {
+        // PDF magic number check
+        if (buffer.slice(0, 5).toString() !== "%PDF-") {
+            throw new Error("Invalid PDF file");
+        }
+        // Parse basic PDF structure
+        const result = {
+            text: "",
+            metadata: {},
+            pages: [],
+        };
+        // Extract text using regex (basic implementation)
+        // In production, use pdf-parse or pdf-lib
+        const textContent = this.extractTextFromPDF(buffer);
+        result.text = textContent;
+        result.pages = [{
+                pageNumber: 1,
+                text: textContent.slice(0, 5000), // First page preview
+            }];
+        // Extract metadata
+        if (this.config.extractMetadata) {
+            result.metadata = this.extractMetadata(buffer);
+        }
+        return result;
+    }
+    /**
+     * Extract text using basic PDF parsing
+     */
+    extractTextFromPDF(buffer) {
+        const content = buffer.toString("latin1");
+        // Find text streams (BT ... ET)
+        const textRegex = /BT\s*((?:(?!ET).|\s)*)\s*ET/g;
+        const matches = content.match(textRegex) || [];
+        const extractedTexts = [];
+        for (const match of matches) {
+            // Extract Tj and TJ operands
+            const text = match
+                .replace(/BT|ET/g, "")
+                .replace(/\[[^\]]*\]TJ/g, (m) => {
+                // Handle TJ arrays
+                const inner = m.slice(1, -3);
+                return inner
+                    .split(" ")
+                    .filter((_, i) => i % 2 === 0)
+                    .join("");
+            })
+                .replace(/\([^)]*\)Tj/g, (m) => {
+                // Handle Tj strings
+                return m.slice(1, -3);
+            })
+                .replace(/\d+\.?\d*\s+\d+\.?\d*\s+Td/g, " ") // Position operators
+                .replace(/\d+\.?\d*\s+TL/g, " ") // Line operators
+                .replace(/\s+/g, " ")
+                .trim();
+            if (text)
+                extractedTexts.push(text);
+        }
+        // Also try obj streams
+        const streamRegex = /stream\r?\n([\s\S]*?)endstream/g;
+        let streamMatch;
+        while ((streamMatch = streamRegex.exec(content)) !== null) {
+            const stream = streamMatch[1];
+            // Check if it looks like text
+            if (stream.includes("(") && stream.includes("Tj")) {
+                const text = stream
+                    .replace(/\([^)]*\)Tj/g, (m) => m.slice(1, -3))
+                    .replace(/\s+/g, " ")
+                    .trim();
+                if (text.length > 3)
+                    extractedTexts.push(text);
+            }
+        }
+        return extractedTexts.join("\n").slice(0, 50000); // Limit output
+    }
+    /**
+     * Extract metadata
+     */
+    extractMetadata(buffer) {
+        const content = buffer.toString("latin1");
+        const metadata = {};
+        // Title
+        const titleMatch = content.match(/\/Title\s*\(([^)]+)\)/);
+        if (titleMatch)
+            metadata.title = this.unescapePDFString(titleMatch[1]);
+        // Author
+        const authorMatch = content.match(/\/Author\s*\(([^)]+)\)/);
+        if (authorMatch)
+            metadata.author = this.unescapePDFString(authorMatch[1]);
+        // Subject
+        const subjectMatch = content.match(/\/Subject\s*\(([^)]+)\)/);
+        if (subjectMatch)
+            metadata.subject = this.unescapePDFString(subjectMatch[1]);
+        // Creator
+        const creatorMatch = content.match(/\/Creator\s*\(([^)]+)\)/);
+        if (creatorMatch)
+            metadata.creator = this.unescapePDFString(creatorMatch[1]);
+        // Producer
+        const producerMatch = content.match(/\/Producer\s*\(([^)]+)\)/);
+        if (producerMatch)
+            metadata.producer = this.unescapePDFString(producerMatch[1]);
+        // Page count
+        const pagesMatch = content.match(/\/Count\s+(\d+)/);
+        if (pagesMatch)
+            metadata.pageCount = parseInt(pagesMatch[1], 10);
+        // Keywords  
+        const keywordsMatch = content.match(/\/Keywords\s*\(([^)]+)\)/);
+        if (keywordsMatch)
+            metadata.keywords = this.unescapePDFString(keywordsMatch[1]);
+        return metadata;
+    }
+    /**
+     * Unescape PDF string
+     */
+    unescapePDFString(str) {
+        return str
+            .replace(/\\n/g, "\n")
+            .replace(/\\r/g, "\r")
+            .replace(/\\t/g, "\t")
+            .replace(/\\\\/g, "\\")
+            .replace(/\\\(/g, "(")
+            .replace(/\\\)/g, ")");
+    }
+    /**
+     * Get PDF info
+     */
+    async getInfo(filePath) {
+        const buffer = await fs.readFile(filePath);
+        return this.extractMetadata(buffer);
+    }
+}
+// Helper functions
+export async function extractPDF(filePath, config) {
+    const extractor = new PDFExtractor(config);
+    return extractor.extractFromFile(filePath);
+}
+export async function extractPDFText(filePath) {
+    const result = await extractPDF(filePath, { extractMetadata: false });
+    return result.text;
+}
+export { PDFExtractor };
+export default PDFExtractor;
+//# sourceMappingURL=pdf.js.map

package/dist/src/documents/pdf.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdf.js","sourceRoot":"","sources":["../../../src/documents/pdf.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,KAAK,EAAE,MAAM,kBAAkB,CAAC;AA4BvC,MAAM,YAAY;IACR,MAAM,CAAY;IAE1B,YAAY,SAA6B,EAAE;QACzC,IAAI,CAAC,MAAM,GAAG;YACZ,QAAQ,EAAE,GAAG;YACb,WAAW,EAAE,IAAI;YACjB,eAAe,EAAE,IAAI;YACrB,GAAG,MAAM;SACV,CAAC;IACJ,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,eAAe,CAAC,QAAgB;QACpC,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAC3C,OAAO,IAAI,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;IACxC,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,iBAAiB,CAAC,MAAc;QACpC,yBAAyB;QACzB,IAAI,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,KAAK,OAAO,EAAE,CAAC;YAC9C,MAAM,IAAI,KAAK,CAAC,kBAAkB,CAAC,CAAC;QACtC,CAAC;QAED,4BAA4B;QAC5B,MAAM,MAAM,GAAwB;YAClC,IAAI,EAAE,EAAE;YACR,QAAQ,EAAE,EAAE;YACZ,KAAK,EAAE,EAAE;SACV,CAAC;QAEF,kDAAkD;QAClD,0CAA0C;QAC1C,MAAM,WAAW,GAAG,IAAI,CAAC,kBAAkB,CAAC,MAAM,CAAC,CAAC;QAEpD,MAAM,CAAC,IAAI,GAAG,WAAW,CAAC;QAC1B,MAAM,CAAC,KAAK,GAAG,CAAC;gBACd,UAAU,EAAE,CAAC;gBACb,IAAI,EAAE,WAAW,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,EAAE,qBAAqB;aACxD,CAAC,CAAC;QAEH,mBAAmB;QACnB,IAAI,IAAI,CAAC,MAAM,CAAC,eAAe,EAAE,CAAC;YAChC,MAAM,CAAC,QAAQ,GAAG,IAAI,CAAC,eAAe,CAAC,MAAM,CAAC,CAAC;QACjD,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED;;OAEG;IACK,kBAAkB,CAAC,MAAc;QACvC,MAAM,OAAO,GAAG,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAE1C,gCAAgC;QAChC,MAAM,SAAS,GAAG,8BAA8B,CAAC;QACjD,MAAM,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC;QAE/C,MAAM,cAAc,GAAa,EAAE,CAAC;QAEpC,KAAK,MAAM,KAAK,IAAI,OAAO,EAAE,CAAC;YAC5B,6BAA6B;YAC7B,MAAM,IAAI,GAAG,KAAK;iBACf,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC;iBACrB,OAAO,CAAC,eAAe,EAAE,CAAC,CAAC,EAAE,EAAE;gBAC9B,mBAAmB;gBACnB,MAAM,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;gBAC7B,OAAO,KAAK;qBACT,KAAK,CAAC,GAAG,CAAC;qBACV,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;qBAC7B,IAAI,CAAC,EAAE,CAAC,CAAC;YACd,CAAC,CAAC;iBACD,OAAO,CAAC,cAAc,EAAE,CAAC,CAAC,EAAE,EAAE;gBAC7B,oBAAoB;gBACpB,OAAO,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;YACxB,CAAC,CAAC;iBACD,OAAO,CAAC,6BAA6B,EAAE,GAAG,CAAC,CAAC,qBAAqB;iBACjE,OAAO,CAAC,iBAAiB,EAAE,GAAG,CAAC,CAAC,iBAAiB;iBACjD,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;iBACpB,IAAI,EAAE,CAAC;YAEV,IAAI,IAAI;gBAAE,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACtC,CAAC;QAED,uBAAuB;QACvB,MAAM,WAAW,GAAG,iCAAiC,CAAC;QACtD,IAAI,WAAW,CAAC;QAChB,OAAO,CAAC,WAAW,GAAG,WAAW,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC1D,MAAM,MAAM,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC;YAC9B,8BAA8B;YAC9B,IAAI,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;gBAClD,MAAM,IAAI,GAAG,MAAM;qBAChB,OAAO,CAAC,cAAc,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;qBAC9C,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;qBACpB,IAAI,EAAE,CAAC;gBACV,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;oBAAE,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YACjD,CAAC;QACH,CAAC;QAED,OAAO,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,eAAe;IACnE,CAAC;IAED;;OAEG;IACK,eAAe,CAAC,MAAc;QACpC,MAAM,OAAO,GAAG,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAC1C,MAAM,QAAQ,GAAoC,EAAE,CAAC;QAErD,QAAQ;QACR,MAAM,UAAU,GAAG,OAAO,CAAC,KAAK,CAAC,uBAAuB,CAAC,CAAC;QAC1D,IAAI,UAAU;YAAE,QAAQ,CAAC,KAAK,GAAG,IAAI,CAAC,iBAAiB,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;QAEvE,SAAS;QACT,MAAM,WAAW,GAAG,OAAO,CAAC,KAAK,CAAC,wBAAwB,CAAC,CAAC;QAC5D,IAAI,WAAW;YAAE,QAAQ,CAAC,MAAM,GAAG,IAAI,CAAC,iBAAiB,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC;QAE1E,UAAU;QACV,MAAM,YAAY,GAAG,OAAO,CAAC,KAAK,CAAC,yBAAyB,CAAC,CAAC;QAC9D,IAAI,YAAY;YAAE,QAAQ,CAAC,OAAO,GAAG,IAAI,CAAC,iBAAiB,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;QAE7E,UAAU;QACV,MAAM,YAAY,GAAG,OAAO,CAAC,KAAK,CAAC,yBAAyB,CAAC,CAAC;QAC9D,IAAI,YAAY;YAAE,QAAQ,CAAC,OAAO,GAAG,IAAI,CAAC,iBAAiB,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;QAE7E,WAAW;QACX,MAAM,aAAa,GAAG,OAAO,CAAC,KAAK,CAAC,0BAA0B,CAAC,CAAC;QAChE,IAAI,aAAa;YAAE,QAAQ,CAAC,QAAQ,GAAG,IAAI,CAAC,iBAAiB,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC;QAEhF,aAAa;QACb,MAAM,UAAU,GAAG,OAAO,CAAC,KAAK,CAAC,iBAAiB,CAAC,CAAC;QACpD,IAAI,UAAU;YAAE,QAAQ,CAAC,SAAS,GAAG,QAAQ,CAAC,UAAU,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QAEjE,aAAa;QACb,MAAM,aAAa,GAAG,OAAO,CAAC,KAAK,CAAC,0BAA0B,CAAC,CAAC;QAChE,IAAI,aAAa;YAAE,QAAQ,CAAC,QAAQ,GAAG,IAAI,CAAC,iBAAiB,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC;QAEhF,OAAO,QAAQ,CAAC;IAClB,CAAC;IAED;;OAEG;IACK,iBAAiB,CAAC,GAAW;QACnC,OAAO,GAAG;aACP,OAAO,CAAC,MAAM,EAAE,IAAI,CAAC;aACrB,OAAO,CAAC,MAAM,EAAE,IAAI,CAAC;aACrB,OAAO,CAAC,MAAM,EAAE,IAAI,CAAC;aACrB,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC;aACtB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;aACrB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;IAC3B,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,OAAO,CAAC,QAAgB;QAC5B,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAC3C,OAAO,IAAI,CAAC,eAAe,CAAC,MAAM,CAAC,CAAC;IACtC,CAAC;CACF;AAED,mBAAmB;AACnB,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,QAAgB,EAChB,MAA2B;IAE3B,MAAM,SAAS,GAAG,IAAI,YAAY,CAAC,MAAM,CAAC,CAAC;IAC3C,OAAO,SAAS,CAAC,eAAe,CAAC,QAAQ,CAAC,CAAC;AAC7C,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,cAAc,CAAC,QAAgB;IACnD,MAAM,MAAM,GAAG,MAAM,UAAU,CAAC,QAAQ,EAAE,EAAE,eAAe,EAAE,KAAK,EAAE,CAAC,CAAC;IACtE,OAAO,MAAM,CAAC,IAAI,CAAC;AACrB,CAAC;AAED,OAAO,EAAE,YAAY,EAAE,CAAC;AACxB,eAAe,YAAY,CAAC"}