@promptbook/browser 0.72.0-8 → 0.72.0-9

package/esm/typings/src/llm-providers/remote/interfaces/RemoteLlmExecutionToolsOptions.d.ts

@@ -1,14 +1,14 @@
-import type { CommonExecutionToolsOptions } from '../../../execution/CommonExecutionToolsOptions';
+import type { CommonToolsOptions } from '../../../execution/CommonToolsOptions';
 import type { string_base_url } from '../../../types/typeAliases';
 import type { string_uri } from '../../../types/typeAliases';
 import type { string_user_id } from '../../../types/typeAliases';
-import type { LlmToolsConfiguration } from '../../_common/LlmToolsConfiguration';
+import type { LlmToolsConfiguration } from '../../_common/register/LlmToolsConfiguration';
 /**
  * Options for `RemoteLlmExecutionTools`
  *
  * @public exported from `@promptbook/remote-client`
  */
-export type RemoteLlmExecutionToolsOptions = CommonExecutionToolsOptions & {
+export type RemoteLlmExecutionToolsOptions = CommonToolsOptions & {
     /**
      * URL of the remote PROMPTBOOK server
      * On this server will be connected to the socket.io server

package/esm/typings/src/llm-providers/remote/interfaces/RemoteServerOptions.d.ts

@@ -1,5 +1,5 @@
 import type { PipelineCollection } from '../../../collection/PipelineCollection';
-import type { CommonExecutionToolsOptions } from '../../../execution/CommonExecutionToolsOptions';
+import type { CommonToolsOptions } from '../../../execution/CommonToolsOptions';
 import type { LlmExecutionTools } from '../../../execution/LlmExecutionTools';
 import type { string_uri } from '../../../types/typeAliases';
 import type { string_user_id } from '../../../types/typeAliases';
@@ -17,7 +17,7 @@ import type { string_user_id } from '../../../types/typeAliases';
  * @public exported from `@promptbook/remote-client`
  * @public exported from `@promptbook/remote-server`
  */
-export type RemoteServerOptions = CommonExecutionToolsOptions & {
+export type RemoteServerOptions = CommonToolsOptions & {
     /**
      * Port on which the server will listen
      */

package/esm/typings/src/personas/preparePersona.d.ts

@@ -1,3 +1,4 @@
+import type { ExecutionTools } from '../execution/ExecutionTools';
 import type { PrepareAndScrapeOptions } from '../prepare/PrepareAndScrapeOptions';
 import type { PersonaPreparedJson } from '../types/PipelineJson/PersonaJson';
 import type { string_persona_description } from '../types/typeAliases';
@@ -7,7 +8,7 @@ import type { string_persona_description } from '../types/typeAliases';
  * @see https://github.com/webgptorg/promptbook/discussions/22
  * @public exported from `@promptbook/core`
  */
-export declare function preparePersona(personaDescription: string_persona_description, options: PrepareAndScrapeOptions): Promise<PersonaPreparedJson['modelRequirements']>;
+export declare function preparePersona(personaDescription: string_persona_description, tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions): Promise<PersonaPreparedJson['modelRequirements']>;
 /**
  * TODO: [🔃][main] !!!!! If the persona was prepared with different version or different set of models, prepare it once again
  * TODO: [🏢] !! Check validity of `modelName` in pipeline

package/esm/typings/src/prepare/PrepareAndScrapeOptions.d.ts

@@ -1,20 +1,17 @@
-import type { LlmExecutionTools } from '../execution/LlmExecutionTools';
 import type { string_dirname } from '../types/typeAliases';
 /**
  * Options for preparation of the pipeline
  */
 export type PrepareAndScrapeOptions = {
-    /**
-     * LLM tools
-     */
-    readonly llmTools?: LlmExecutionTools;
     /**
      * Path to the root folder of the pipeline
      *
      * Note: When the pipeline is not created from files, it is `null`
-     * Note: This folder must exist
+     * Note: This folder must exist (=it is not created recursively)
+     *
+     * @default process.cwd()
      */
-    readonly rootDirname: string_dirname | null;
+    readonly rootDirname?: string_dirname | null;
     /**
      * Path to the cache folder
      *
@@ -55,6 +52,10 @@ export type PrepareAndScrapeOptions = {
          */
         readonly libreOfficePath?: string;
     };
+    /**
+     * If true, the missing software is automatically installed
+     */
+    readonly isAutoInstalled?: boolean;
     /**
      * If true, the preparation logs additional information
      *

package/esm/typings/src/prepare/preparePipeline.d.ts

@@ -1,3 +1,4 @@
+import type { ExecutionTools } from '../execution/ExecutionTools';
 import type { PipelineJson } from '../types/PipelineJson/PipelineJson';
 import type { PrepareAndScrapeOptions } from './PrepareAndScrapeOptions';
 /**
@@ -8,7 +9,7 @@ import type { PrepareAndScrapeOptions } from './PrepareAndScrapeOptions';
  * Note: When the pipeline is already prepared, it returns the same pipeline
  * @public exported from `@promptbook/core`
  */
-export declare function preparePipeline(pipeline: PipelineJson, options: PrepareAndScrapeOptions): Promise<PipelineJson>;
+export declare function preparePipeline(pipeline: PipelineJson, tools: Pick<ExecutionTools, 'llm' | 'fs' | 'scrapers'>, options: PrepareAndScrapeOptions): Promise<PipelineJson>;
 /**
  * TODO: Write tests for `preparePipeline`
  * TODO: [🏏] Leverage the batch API and build queues @see https://platform.openai.com/docs/guides/batch

package/esm/typings/src/prepare/prepareTemplates.d.ts

@@ -1,3 +1,4 @@
+import type { ExecutionTools } from '../execution/ExecutionTools';
 import type { PipelineJson } from '../types/PipelineJson/PipelineJson';
 import type { TemplateJson } from '../types/PipelineJson/TemplateJson';
 import type { PrepareAndScrapeOptions } from './PrepareAndScrapeOptions';
@@ -18,7 +19,7 @@ type PreparedTemplates = {
  *
  * @public exported from `@promptbook/core`
  */
-export declare function prepareTemplates(pipeline: PrepareTemplateInput, options: PrepareAndScrapeOptions): Promise<PreparedTemplates>;
+export declare function prepareTemplates(pipeline: PrepareTemplateInput, tools: Pick<ExecutionTools, 'llm' | 'fs' | 'scrapers'>, options: PrepareAndScrapeOptions): Promise<PreparedTemplates>;
 export {};
 /**
  * TODO: [🧠] Add context to each template (if missing)

package/esm/typings/src/scrapers/_common/Converter.d.ts

@@ -1,6 +1,4 @@
-import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
-import type { string_mime_type } from '../../types/typeAliases';
-import type { string_promptbook_documentation_url } from '../../types/typeAliases';
+import type { ScraperAndConverterMetadata } from './register/ScraperAndConverterMetadata';
 import type { ScraperSourceHandler } from './Scraper';
 import type { ScraperIntermediateSource } from './ScraperIntermediateSource';
 /**
@@ -9,13 +7,9 @@ import type { ScraperIntermediateSource } from './ScraperIntermediateSource';
  */
 export type Converter = {
     /**
-     * Mime types that this scraper can handle
+     * Metadata of the converter which includes title, mime types, etc.
      */
-    readonly mimeTypes: Array<string_mime_type>;
-    /**
-     * Link to documentation
-     */
-    readonly documentationUrl: string_promptbook_documentation_url;
+    readonly metadata: ScraperAndConverterMetadata;
     /**
      * Convert the the file and returns intermediate source or `null` if it can't convert it
      *
@@ -24,5 +18,5 @@ export type Converter = {
      *
      * Note: `$` is used to indicate that this function is not a pure function - it leaves files on the disk and you are responsible for cleaning them by calling `destroy` method of returned object
      */
-    $convert(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promise<ScraperIntermediateSource>;
+    $convert(source: ScraperSourceHandler): Promise<ScraperIntermediateSource>;
 };

package/esm/typings/src/scrapers/_common/Scraper.d.ts

@@ -1,28 +1,23 @@
 import type { Promisable } from 'type-fest';
-import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
 import type { string_filename } from '../../types/typeAliases';
 import type { string_knowledge_source_link } from '../../types/typeAliases';
 import type { string_mime_type } from '../../types/typeAliases';
-import type { string_promptbook_documentation_url } from '../../types/typeAliases';
 import type { string_url } from '../../types/typeAliases';
+import type { ScraperAndConverterMetadata } from './register/ScraperAndConverterMetadata';
 /**
  * @@@
  *
  */
 export type Scraper = {
     /**
-     * Mime types that this scraper can handle
+     * Metadata of the scraper which includes title, mime types, etc.
      */
-    readonly mimeTypes: Array<string_mime_type>;
-    /**
-     * Link to documentation
-     */
-    readonly documentationUrl: string_promptbook_documentation_url;
+    readonly metadata: ScraperAndConverterMetadata;
     /**
      * Scrapes the markdown file and returns the knowledge pieces or `null` if it can't scrape it
      */
-    scrape(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promisable<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
+    scrape(source: ScraperSourceHandler): Promisable<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
 };
 /**
  * @@@

package/esm/typings/src/scrapers/_common/prepareKnowledgePieces.d.ts

@@ -1,13 +1,14 @@
 import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
 import type { KnowledgeSourceJson } from '../../types/PipelineJson/KnowledgeSourceJson';
+import type { ExecutionTools } from '../../execution/ExecutionTools';
 /**
  * Prepares the knowle
  *
  * @see https://github.com/webgptorg/promptbook/discussions/41
  * @public exported from `@promptbook/core`
  */
-export declare function prepareKnowledgePieces(knowledgeSources: Array<KnowledgeSourceJson>, options: PrepareAndScrapeOptions): Promise<Array<Omit<KnowledgePiecePreparedJson, 'preparationIds'>>>;
+export declare function prepareKnowledgePieces(knowledgeSources: Array<KnowledgeSourceJson>, tools: Pick<ExecutionTools, 'llm' | 'fs' | 'scrapers'>, options: PrepareAndScrapeOptions): Promise<Array<Omit<KnowledgePiecePreparedJson, 'preparationIds'>>>;
 /**
  * TODO: [🧊] In future one preparation can take data from previous preparation and save tokens and time
  *       Put `knowledgePieces` into `PrepareKnowledgeOptions`

package/esm/typings/src/scrapers/_common/register/$provideFilesystemForNode.d.ts

@@ -0,0 +1,11 @@
+import type { FilesystemTools } from '../../../execution/FilesystemTools';
+import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
+/**
+ * @@@
+ *
+ * @public exported from `@promptbook/node`
+ */
+export declare function $provideFilesystemForNode(options?: Pick<PrepareAndScrapeOptions, 'isVerbose'>): FilesystemTools;
+/**
+ * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
+ */

package/esm/typings/src/scrapers/_common/register/$provideScrapersForBrowser.d.ts

@@ -0,0 +1,12 @@
+import type { ExecutionTools } from '../../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
+import type { Scraper } from '../Scraper';
+/**
+ * @@@!!!!!!
+ *
+ * 1) @@@
+ * 2) @@@
+ *
+ * @public exported from `@promptbook/browser`
+ */
+export declare function $provideScrapersForBrowser(tools: Pick<ExecutionTools, 'llm'>, options?: PrepareAndScrapeOptions): Promise<Array<Scraper>>;

package/esm/typings/src/scrapers/_common/register/$provideScrapersForNode.d.ts

@@ -0,0 +1,15 @@
+import type { ExecutionTools } from '../../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
+import type { Scraper } from '../Scraper';
+/**
+ * !!!!!!
+ *
+ * 1) @@@
+ * 2) @@@
+ *
+ * @public exported from `@promptbook/node`
+ */
+export declare function $provideScrapersForNode(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options?: PrepareAndScrapeOptions): Promise<Array<Scraper>>;
+/**
+ * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
+ */

package/esm/typings/src/scrapers/_common/register/$registeredScrapersMessage.d.ts

@@ -0,0 +1,12 @@
+import type { string_markdown } from '../../../types/typeAliases';
+/**
+ * Creates a message with all registered scrapers
+ *
+ * Note: This function is used to create a (error) message when there is no scraper for particular mime type
+ *
+ * @private internal function of `createScrapersFromConfiguration` and `createScrapersFromEnv`
+ */
+export declare function $registeredScrapersMessage(): string_markdown;
+/**
+ * TODO: [®] DRY Register logic
+ */

package/esm/typings/src/scrapers/_common/register/$scrapersMetadataRegister.d.ts

@@ -0,0 +1,13 @@
+import { $Register } from '../../../utils/$Register';
+import type { ScraperAndConverterMetadata } from './ScraperAndConverterMetadata';
+/**
+ * @@@
+ *
+ * Note: `$` is used to indicate that this interacts with the global scope
+ * @singleton Only one instance of each register is created per build, but thare can be more @@@
+ * @public exported from `@promptbook/core`
+ */
+export declare const $scrapersMetadataRegister: $Register<ScraperAndConverterMetadata>;
+/**
+ * TODO: [®] DRY Register logic
+ */

package/esm/typings/src/scrapers/_common/register/$scrapersRegister.d.ts

@@ -0,0 +1,13 @@
+import { $Register } from '../../../utils/$Register';
+import type { ScraperConstructor } from './ScraperConstructor';
+/**
+ * @@@
+ *
+ * Note: `$` is used to indicate that this interacts with the global scope
+ * @singleton Only one instance of each register is created per build, but thare can be more @@@
+ * @public exported from `@promptbook/core`
+ */
+export declare const $scrapersRegister: $Register<ScraperConstructor>;
+/**
+ * TODO: [®] DRY Register logic
+ */

package/esm/typings/src/scrapers/_common/register/ScraperAndConverterMetadata.d.ts

@@ -0,0 +1,41 @@
+import type { string_mime_type } from '../../../types/typeAliases';
+import type { string_promptbook_documentation_url } from '../../../types/typeAliases';
+import type { string_title } from '../../../types/typeAliases';
+import type { Registered } from '../../../utils/$Register';
+import type { TODO_any } from '../../../utils/organization/TODO_any';
+/**
+ * @@@
+ *
+ * @@@
+ * x) `Scraper`
+ * x) `Converter`
+ * x) `ScraperConstructor`
+ * x) `Registered`
+ * x) `ExecutionTools`
+ * x) `ScraperAndConverterMetadata`
+ * x) `PrepareAndScrapeOptions`
+ * x) `ScraperConfiguration`
+ * x) `ScraperOptions`
+ */
+export type ScraperAndConverterMetadata = Registered & {
+    /**
+     * @@@
+     */
+    readonly title: string_title;
+    /**
+     * Mime types that this scraper can handle
+     */
+    readonly mimeTypes: ReadonlyArray<string_mime_type>;
+    /**
+     * @@@
+     */
+    readonly isAvilableInBrowser: boolean;
+    /**
+     * @@@
+     */
+    readonly requiredExecutables: TODO_any;
+    /**
+     * Link to documentation
+     */
+    readonly documentationUrl: string_promptbook_documentation_url;
+};

package/esm/typings/src/scrapers/_common/register/ScraperConstructor.d.ts

@@ -0,0 +1,12 @@
+import type { ExecutionTools } from '../../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
+import type { Registered } from '../../../utils/$Register';
+import type { Scraper } from '../Scraper';
+import type { ScraperAndConverterMetadata } from './ScraperAndConverterMetadata';
+/**
+ * @@@
+ */
+export type ScraperConstructor = Registered & ScraperAndConverterMetadata & ((tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions) => Scraper);
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/_common/utils/getScraperIntermediateSource.d.ts

@@ -30,4 +30,5 @@ export {};
  * 1) Need to store more than serialized JSONs
  * 2) Need to switch between a `rootDirname` and `cacheDirname` <- TODO: !!!!
  * TODO: [🐱‍🐉][🧠] Make some smart crop
+ * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
  */

package/esm/typings/src/scrapers/_common/utils/makeKnowledgeSourceHandler.d.ts

@@ -1,4 +1,5 @@
 import type { SetOptional } from 'type-fest';
+import type { ExecutionTools } from '../../../execution/ExecutionTools';
 import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgeSourceJson } from '../../../types/PipelineJson/KnowledgeSourceJson';
 import type { ScraperSourceHandler } from '../Scraper';
@@ -7,4 +8,4 @@ import type { ScraperSourceHandler } from '../Scraper';
  *
  * @private for scraper utilities
  */
-export declare function makeKnowledgeSourceHandler(knowledgeSource: SetOptional<KnowledgeSourceJson, 'name'>, options?: Pick<PrepareAndScrapeOptions, 'rootDirname' | 'isVerbose'>): Promise<ScraperSourceHandler>;
+export declare function makeKnowledgeSourceHandler(knowledgeSource: SetOptional<KnowledgeSourceJson, 'name'>, tools: Pick<ExecutionTools, 'fs'>, options?: Pick<PrepareAndScrapeOptions, 'rootDirname' | 'isVerbose'>): Promise<ScraperSourceHandler>;

package/esm/typings/src/scrapers/document/{documentScraper.d.ts → DocumentScraper.d.ts}

@@ -1,37 +1,43 @@
-import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import type { Converter } from '../_common/Converter';
+import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
+import type { Scraper } from '../_common/Scraper';
 import type { ScraperSourceHandler } from '../_common/Scraper';
 import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
 /**
  * Scraper of .docx and .odt files
  *
  * @see `documentationUrl` for more details
- * @public exported from `@promptbook/core`
+ * @public exported from `@promptbook/documents`
  */
-export declare const documentScraper: {
+export declare class DocumentScraper implements Converter, Scraper {
+    private readonly tools;
+    private readonly options;
     /**
-     * Mime types that this scraper can handle
+     * Metadata of the scraper which includes title, mime types, etc.
      */
-    mimeTypes: string[];
+    get metadata(): ScraperAndConverterMetadata;
     /**
-     * Link to documentation
+     * Markdown scraper is used internally
      */
-    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    private readonly markdownScraper;
+    constructor(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options: PrepareAndScrapeOptions);
     /**
      * Convert the `.docx` or `.odt`  to `.md` file and returns intermediate source
      *
      * Note: `$` is used to indicate that this function is not a pure function - it leaves files on the disk and you are responsible for cleaning them by calling `destroy` method of returned object
      */
-    $convert(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promise<ScraperIntermediateSource>;
+    $convert(source: ScraperSourceHandler): Promise<ScraperIntermediateSource>;
     /**
      * Scrapes the docx file and returns the knowledge pieces or `null` if it can't scrape it
      */
-    scrape(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
-};
+    scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
+}
 /**
  * TODO: [👣] Converted documents can act as cached items - there is no need to run conversion each time
- * TODO: [🦖] Make some system for putting scrapers to separete packages
  * TODO: [🪂] Do it in parallel 11:11
- * TODO: [🦷] Ideally use `as const satisfies Converter & Scraper` BUT this combination throws errors
  * Note: No need to aggregate usage here, it is done by intercepting the llmTools
+ * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
  */

package/esm/typings/src/scrapers/document/createDocumentScraper.d.ts

@@ -0,0 +1,20 @@
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import { DocumentScraper } from './DocumentScraper';
+/**
+ * @@@
+ *
+ * @public exported from `@promptbook/documents`
+ */
+export declare const createDocumentScraper: ((tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions) => DocumentScraper) & import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: false;
+    requiredExecutables: string[];
+}>;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/document/register-constructor.d.ts

@@ -0,0 +1,13 @@
+import type { Registration } from '../../utils/$Register';
+/**
+ * Registration of known scraper
+ *
+ * Warning: This is not useful for the end user, it is just a side effect of the mechanism that handles all available known scrapers
+ *
+ * @public exported from `@promptbook/documents`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _DocumentScraperRegistration: Registration;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/document/register-metadata.d.ts

@@ -0,0 +1,24 @@
+import type { Registration } from '../../utils/$Register';
+/**
+ * Metadata of the scraper
+ *
+ * @private within the scraper directory
+ */
+export declare const documentScraperMetadata: import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: false;
+    requiredExecutables: string[];
+}>;
+/**
+ * Registration of known scraper metadata
+ *
+ * Warning: This is not useful for the end user, it is just a side effect of the mechanism that handles all available known scrapers
+ *
+ * @public exported from `@promptbook/core`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _DocumentScraperMetadataRegistration: Registration;

package/esm/typings/src/scrapers/document-legacy/{legacyDocumentScraper.d.ts → LegacyDocumentScraper.d.ts}

@@ -1,37 +1,43 @@
-import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import type { Converter } from '../_common/Converter';
+import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
+import type { Scraper } from '../_common/Scraper';
 import type { ScraperSourceHandler } from '../_common/Scraper';
 import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
 /**
  * Scraper for .docx files
  *
  * @see `documentationUrl` for more details
- * @public exported from `@promptbook/core`
+ * @public exported from `@promptbook/legacy-documents`
  */
-export declare const legacyDocumentScraper: {
+export declare class LegacyDocumentScraper implements Converter, Scraper {
+    private readonly tools;
+    private readonly options;
     /**
-     * Mime types that this scraper can handle
+     * Metadata of the scraper which includes title, mime types, etc.
      */
-    mimeTypes: string[];
+    get metadata(): ScraperAndConverterMetadata;
     /**
-     * Link to documentation
+     * Document scraper is used internally
      */
-    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    private readonly documentScraper;
+    constructor(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options: PrepareAndScrapeOptions);
     /**
      * Convert the `.doc` or `.rtf`  to `.doc` file and returns intermediate source
      *
      * Note: `$` is used to indicate that this function is not a pure function - it leaves files on the disk and you are responsible for cleaning them by calling `destroy` method of returned object
      */
-    $convert(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promise<ScraperIntermediateSource>;
+    $convert(source: ScraperSourceHandler): Promise<ScraperIntermediateSource>;
     /**
      * Scrapes the `.doc` or `.rtf` file and returns the knowledge pieces or `null` if it can't scrape it
      */
-    scrape(source: ScraperSourceHandler, options: PrepareAndScrapeOptions): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
-};
+    scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
+}
 /**
  * TODO: [👣] Converted documents can act as cached items - there is no need to run conversion each time
- * TODO: [🦖] Make some system for putting scrapers to separete packages
  * TODO: [🪂] Do it in parallel 11:11
- * TODO: [🦷] Ideally use `as const satisfies Converter & Scraper` BUT this combination throws errors
  * Note: No need to aggregate usage here, it is done by intercepting the llmTools
+ * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
  */

package/esm/typings/src/scrapers/document-legacy/createLegacyDocumentScraper.d.ts

@@ -0,0 +1,20 @@
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import { LegacyDocumentScraper } from './LegacyDocumentScraper';
+/**
+ * @@@
+ *
+ * @public exported from `@promptbook/legacy-documents`
+ */
+export declare const createLegacyDocumentScraper: ((tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions) => LegacyDocumentScraper) & import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: false;
+    requiredExecutables: string[];
+}>;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/document-legacy/register-constructor.d.ts

@@ -0,0 +1,13 @@
+import type { Registration } from '../../utils/$Register';
+/**
+ * Registration of known scraper
+ *
+ * Warning: This is not useful for the end user, it is just a side effect of the mechanism that handles all available known scrapers
+ *
+ * @public exported from `@promptbook/legacy-documents`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _LegacyDocumentScraperRegistration: Registration;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/document-legacy/register-metadata.d.ts

@@ -0,0 +1,24 @@
+import type { Registration } from '../../utils/$Register';
+/**
+ * Metadata of the scraper
+ *
+ * @private within the scraper directory
+ */
+export declare const legacyDocumentScraperMetadata: import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: false;
+    requiredExecutables: string[];
+}>;
+/**
+ * Registration of known scraper metadata
+ *
+ * Warning: This is not useful for the end user, it is just a side effect of the mechanism that handles all available known scrapers
+ *
+ * @public exported from `@promptbook/core`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _LegacyDocumentScraperMetadataRegistration: Registration;

package/esm/typings/src/scrapers/markdown/MarkdownScraper.d.ts

@@ -0,0 +1,29 @@
+import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
+import type { Scraper } from '../_common/Scraper';
+import type { ScraperSourceHandler } from '../_common/Scraper';
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
+/**
+ * Scraper for markdown files
+ *
+ * @see `documentationUrl` for more details
+ * @public exported from `@promptbook/markdown-utils`
+ */
+export declare class MarkdownScraper implements Scraper {
+    private readonly tools;
+    private readonly options;
+    /**
+     * Metadata of the scraper which includes title, mime types, etc.
+     */
+    get metadata(): ScraperAndConverterMetadata;
+    constructor(tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions);
+    /**
+     * Scrapes the markdown file and returns the knowledge pieces or `null` if it can't scrape it
+     */
+    scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
+}
+/**
+ * TODO: [🪂] Do it in parallel 11:11
+ * Note: No need to aggregate usage here, it is done by intercepting the llmTools
+ */