@promptbook/openai 0.72.0-7 → 0.72.0-9

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
+ */

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

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

package/esm/typings/src/scrapers/markdown/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/markdown-utils`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _MarkdownScraperRegistration: Registration;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/markdown/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 markdownScraperMetadata: import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: true;
+    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 _MarkdownScraperMetadataRegistration: Registration;

package/esm/typings/src/scrapers/pdf/PdfScraper.d.ts

@@ -0,0 +1,40 @@
+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 { Converter } from '../_common/Converter';
+import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
+import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
+/**
+ * Scraper for .docx files
+ *
+ * @see `documentationUrl` for more details
+ * @public exported from `@promptbook/pdf`
+ */
+export declare class PdfScraper implements Converter, Scraper {
+    private readonly tools;
+    private readonly options;
+    /**
+     * Metadata of the scraper which includes title, mime types, etc.
+     */
+    get metadata(): ScraperAndConverterMetadata;
+    /**
+     * Markdown scraper is used internally
+     */
+    private readonly markdownScraper;
+    constructor(tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions);
+    /**
+     * Converts the `.pdf` file to `.md` file and returns intermediate source
+     */
+    $convert(source: ScraperSourceHandler): Promise<ScraperIntermediateSource>;
+    /**
+     * Scrapes the `.pdf` 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: [👣] Converted pdf documents can act as cached items - there is no need to run conversion each time
+ * TODO: [🪂] Do it in parallel 11:11
+ * Note: No need to aggregate usage here, it is done by intercepting the llmTools
+ */

package/esm/typings/src/scrapers/pdf/createPdfScraper.d.ts

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

package/esm/typings/src/scrapers/pdf/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/pdf`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _PdfScraperRegistration: Registration;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */

package/esm/typings/src/scrapers/pdf/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 pdfScraperMetadata: import("type-fest/source/readonly-deep").ReadonlyObjectDeep<{
+    title: string;
+    packageName: string;
+    className: string;
+    mimeTypes: string[];
+    documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
+    isAvilableInBrowser: true;
+    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 _PdfScraperMetadataRegistration: Registration;

package/esm/typings/src/scrapers/website/{websiteScraper.d.ts → WebsiteScraper.d.ts}

@@ -1,43 +1,47 @@
-import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
 import type { KnowledgePiecePreparedJson } from '../../types/PipelineJson/KnowledgePieceJson';
 import type { string_markdown } from '../../types/typeAliases';
+import type { Converter } from '../_common/Converter';
+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';
 import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
 /**
  * Scraper for .docx files
  *
  * @see `documentationUrl` for more details
- * @public exported from `@promptbook/core`
+ * @public exported from `@promptbook/website-crawler`
  */
-export declare const websiteScraper: {
+export declare class WebsiteScraper 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 website  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<import("destroyable").IDestroyable & {
-        readonly filename: string;
-    } & {
+    $convert(source: ScraperSourceHandler): Promise<ScraperIntermediateSource & {
         markdown: string_markdown;
     }>;
     /**
      * Scrapes the website 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: !!!!!! Put into separate package
  * TODO: [👣] Scraped website in .md can act as cache item - 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/website/createWebsiteScraper.d.ts

@@ -0,0 +1,20 @@
+import type { ExecutionTools } from '../../execution/ExecutionTools';
+import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOptions';
+import { WebsiteScraper } from './WebsiteScraper';
+/**
+ * @@@
+ *
+ * @public exported from `@promptbook/website-crawler`
+ */
+export declare const createWebsiteScraper: ((tools: Pick<ExecutionTools, 'llm'>, options: PrepareAndScrapeOptions) => WebsiteScraper) & 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/website/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/website-crawler`
+ * @public exported from `@promptbook/cli`
+ */
+export declare const _WebsiteScraperRegistration: Registration;
+/**
+ * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
+ */