@promptbook/browser 0.72.0-9 → 0.72.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (133) hide show
  1. package/README.md +4 -8
  2. package/esm/index.es.js +224 -27
  3. package/esm/index.es.js.map +1 -1
  4. package/esm/typings/src/_packages/core.index.d.ts +24 -18
  5. package/esm/typings/src/_packages/node.index.d.ts +2 -0
  6. package/esm/typings/src/_packages/types.index.d.ts +18 -12
  7. package/esm/typings/src/_packages/utils.index.d.ts +2 -2
  8. package/esm/typings/src/collection/PipelineCollection.d.ts +1 -1
  9. package/esm/typings/src/collection/SimplePipelineCollection.d.ts +2 -2
  10. package/esm/typings/src/collection/collectionToJson.d.ts +1 -1
  11. package/esm/typings/src/collection/constructors/createCollectionFromJson.d.ts +1 -1
  12. package/esm/typings/src/collection/constructors/createCollectionFromPromise.d.ts +1 -1
  13. package/esm/typings/src/commands/_common/types/CommandParser.d.ts +5 -5
  14. package/esm/typings/src/config.d.ts +21 -14
  15. package/esm/typings/src/executables/$provideExecutablesForNode.d.ts +12 -0
  16. package/esm/typings/src/executables/apps/locateLibreoffice.d.ts +11 -0
  17. package/esm/typings/src/executables/apps/locatePandoc.d.ts +11 -0
  18. package/esm/typings/src/executables/locateApp.d.ts +33 -0
  19. package/esm/typings/src/executables/locateApp.test.d.ts +1 -0
  20. package/esm/typings/src/executables/platforms/locateAppOnLinux.d.ts +12 -0
  21. package/esm/typings/src/executables/platforms/locateAppOnMacOs.d.ts +12 -0
  22. package/esm/typings/src/executables/platforms/locateAppOnWindows.d.ts +12 -0
  23. package/esm/typings/src/execution/EmbeddingVector.d.ts +1 -1
  24. package/esm/typings/src/execution/Executables.d.ts +18 -0
  25. package/esm/typings/src/execution/ExecutionTools.d.ts +9 -3
  26. package/esm/typings/src/execution/FilesystemTools.d.ts +1 -1
  27. package/esm/typings/src/execution/LlmExecutionTools.d.ts +2 -2
  28. package/esm/typings/src/execution/PipelineExecutorResult.d.ts +2 -2
  29. package/esm/typings/src/execution/assertsExecutionSuccessful.d.ts +3 -2
  30. package/esm/typings/src/execution/createPipelineExecutor/00-CreatePipelineExecutorOptions.d.ts +29 -6
  31. package/esm/typings/src/execution/createPipelineExecutor/10-executePipeline.d.ts +2 -11
  32. package/esm/typings/src/execution/createPipelineExecutor/20-executeTemplate.d.ts +4 -13
  33. package/esm/typings/src/execution/createPipelineExecutor/40-executeAttempts.d.ts +9 -14
  34. package/esm/typings/src/execution/translation/automatic-translate/automatic-translators/LindatAutomaticTranslator.d.ts +11 -3
  35. package/esm/typings/src/execution/utils/addUsage.d.ts +1 -1
  36. package/esm/typings/src/execution/utils/forEachAsync.d.ts +1 -1
  37. package/esm/typings/src/formats/_common/FormatDefinition.d.ts +2 -2
  38. package/esm/typings/src/formats/_common/FormatSubvalueDefinition.d.ts +2 -2
  39. package/esm/typings/src/llm-providers/_common/register/$provideLlmToolsForCli.d.ts +2 -2
  40. package/esm/typings/src/llm-providers/_common/register/LlmToolsConfiguration.d.ts +1 -1
  41. package/esm/typings/src/llm-providers/_common/register/createLlmToolsFromConfiguration.d.ts +7 -0
  42. package/esm/typings/src/llm-providers/_common/utils/cache/CacheLlmToolsOptions.d.ts +4 -1
  43. package/esm/typings/src/llm-providers/anthropic-claude/AnthropicClaudeExecutionTools.d.ts +1 -1
  44. package/esm/typings/src/llm-providers/anthropic-claude/AnthropicClaudeExecutionToolsOptions.d.ts +2 -2
  45. package/esm/typings/src/llm-providers/anthropic-claude/anthropic-claude-models.d.ts +1 -1
  46. package/esm/typings/src/llm-providers/azure-openai/AzureOpenAiExecutionTools.d.ts +1 -1
  47. package/esm/typings/src/llm-providers/azure-openai/AzureOpenAiExecutionToolsOptions.d.ts +2 -1
  48. package/esm/typings/src/llm-providers/langtail/LangtailExecutionTools.d.ts +1 -1
  49. package/esm/typings/src/llm-providers/mocked/$fakeTextToExpectations.d.ts +1 -1
  50. package/esm/typings/src/llm-providers/mocked/MockedEchoLlmExecutionTools.d.ts +1 -1
  51. package/esm/typings/src/llm-providers/mocked/MockedFackedLlmExecutionTools.d.ts +1 -1
  52. package/esm/typings/src/llm-providers/multiple/MultipleLlmExecutionTools.d.ts +4 -5
  53. package/esm/typings/src/llm-providers/multiple/joinLlmExecutionTools.d.ts +1 -1
  54. package/esm/typings/src/llm-providers/openai/OpenAiAssistantExecutionTools.d.ts +0 -1
  55. package/esm/typings/src/llm-providers/openai/OpenAiExecutionTools.d.ts +2 -2
  56. package/esm/typings/src/llm-providers/openai/OpenAiExecutionToolsOptions.d.ts +2 -2
  57. package/esm/typings/src/llm-providers/openai/openai-models.d.ts +1 -1
  58. package/esm/typings/src/llm-providers/remote/RemoteLlmExecutionTools.d.ts +4 -4
  59. package/esm/typings/src/llm-providers/remote/interfaces/PromptbookServer_ListModels_Request.d.ts +15 -6
  60. package/esm/typings/src/llm-providers/remote/interfaces/PromptbookServer_ListModels_Response.d.ts +4 -4
  61. package/esm/typings/src/llm-providers/remote/interfaces/PromptbookServer_Prompt_Request.d.ts +6 -12
  62. package/esm/typings/src/llm-providers/remote/interfaces/PromptbookServer_Prompt_Response.d.ts +3 -3
  63. package/esm/typings/src/llm-providers/remote/interfaces/RemoteLlmExecutionToolsOptions.d.ts +9 -14
  64. package/esm/typings/src/llm-providers/remote/interfaces/RemoteServerOptions.d.ts +23 -8
  65. package/esm/typings/src/llm-providers/remote/startRemoteServer.d.ts +1 -1
  66. package/esm/typings/src/prepare/PrepareAndScrapeOptions.d.ts +4 -23
  67. package/esm/typings/src/prepare/prepareTemplates.d.ts +1 -1
  68. package/esm/typings/src/scrapers/_common/Scraper.d.ts +2 -9
  69. package/esm/typings/src/scrapers/_common/prepareKnowledgePieces.d.ts +1 -1
  70. package/esm/typings/src/scrapers/_common/register/$provideScrapersForBrowser.d.ts +2 -2
  71. package/esm/typings/src/scrapers/_common/register/$provideScrapersForNode.d.ts +2 -2
  72. package/esm/typings/src/scrapers/_common/register/ScraperAndConverterMetadata.d.ts +2 -2
  73. package/esm/typings/src/scrapers/_common/utils/getScraperIntermediateSource.d.ts +2 -2
  74. package/esm/typings/src/scrapers/_common/utils/makeKnowledgeSourceHandler.d.ts +1 -1
  75. package/esm/typings/src/scrapers/document/DocumentScraper.d.ts +2 -2
  76. package/esm/typings/src/scrapers/document/createDocumentScraper.d.ts +1 -1
  77. package/esm/typings/src/scrapers/document/register-metadata.d.ts +1 -1
  78. package/esm/typings/src/scrapers/document-legacy/LegacyDocumentScraper.d.ts +3 -3
  79. package/esm/typings/src/scrapers/document-legacy/createLegacyDocumentScraper.d.ts +1 -1
  80. package/esm/typings/src/scrapers/document-legacy/register-metadata.d.ts +1 -1
  81. package/esm/typings/src/scrapers/markdown/MarkdownScraper.d.ts +1 -1
  82. package/esm/typings/src/scrapers/markdown/createMarkdownScraper.d.ts +1 -1
  83. package/esm/typings/src/scrapers/markdown/register-metadata.d.ts +1 -1
  84. package/esm/typings/src/scrapers/pdf/PdfScraper.d.ts +2 -2
  85. package/esm/typings/src/scrapers/pdf/createPdfScraper.d.ts +1 -1
  86. package/esm/typings/src/scrapers/pdf/register-metadata.d.ts +1 -1
  87. package/esm/typings/src/scrapers/website/WebsiteScraper.d.ts +6 -3
  88. package/esm/typings/src/scrapers/website/createWebsiteScraper.d.ts +1 -1
  89. package/esm/typings/src/scrapers/website/register-metadata.d.ts +1 -1
  90. package/esm/typings/src/scrapers/website/utils/createShowdownConverter.d.ts +7 -0
  91. package/esm/typings/src/scrapers/website/utils/createShowdownConverter.test.d.ts +1 -0
  92. package/esm/typings/src/scripting/javascript/utils/preserve.d.ts +1 -1
  93. package/esm/typings/src/storage/file-cache-storage/utils/nameToSubfolderPath.d.ts +1 -1
  94. package/esm/typings/src/types/Arrayable.d.ts +1 -1
  95. package/esm/typings/src/types/IntermediateFilesStrategy.d.ts +7 -0
  96. package/esm/typings/src/types/PipelineJson/KnowledgePieceJson.d.ts +4 -4
  97. package/esm/typings/src/types/PipelineJson/KnowledgeSourceJson.d.ts +1 -1
  98. package/esm/typings/src/types/PipelineJson/PersonaJson.d.ts +1 -1
  99. package/esm/typings/src/types/PipelineJson/TemplateJsonCommon.d.ts +2 -2
  100. package/esm/typings/src/types/Prompt.d.ts +2 -1
  101. package/esm/typings/src/types/execution-report/ExecutionReportJson.d.ts +1 -1
  102. package/esm/typings/src/types/typeAliases.d.ts +11 -8
  103. package/esm/typings/src/utils/$Register.d.ts +1 -1
  104. package/esm/typings/src/utils/FromtoItems.d.ts +1 -1
  105. package/esm/typings/src/utils/arrayableToArray.d.ts +1 -1
  106. package/esm/typings/src/utils/emojis.d.ts +1 -1
  107. package/esm/typings/src/utils/execCommand/$execCommand.d.ts +4 -2
  108. package/esm/typings/src/utils/execCommand/$execCommandNormalizeOptions.d.ts +12 -0
  109. package/esm/typings/src/utils/execCommand/$execCommands.d.ts +1 -0
  110. package/esm/typings/src/utils/execCommand/ExecCommandOptions.d.ts +45 -0
  111. package/esm/typings/src/utils/expectation-counters/countSentences.d.ts +1 -1
  112. package/esm/typings/src/utils/files/isExecutable.d.ts +11 -0
  113. package/esm/typings/src/utils/markdown/extractAllBlocksFromMarkdown.d.ts +1 -1
  114. package/esm/typings/src/utils/markdown/splitMarkdownIntoSections.d.ts +1 -1
  115. package/esm/typings/src/utils/normalization/IKeywords.d.ts +2 -2
  116. package/esm/typings/src/utils/normalization/parseKeywords.d.ts +2 -2
  117. package/esm/typings/src/utils/normalization/parseKeywordsFromString.d.ts +2 -2
  118. package/esm/typings/src/utils/normalization/searchKeywords.d.ts +2 -2
  119. package/esm/typings/src/utils/organization/TODO_USE.d.ts +1 -1
  120. package/esm/typings/src/utils/organization/keepUnused.d.ts +1 -1
  121. package/esm/typings/src/utils/random/$randomSeed.d.ts +1 -1
  122. package/esm/typings/src/utils/sets/intersection.d.ts +1 -1
  123. package/esm/typings/src/utils/sets/union.d.ts +1 -1
  124. package/esm/typings/src/utils/unwrapResult.d.ts +4 -4
  125. package/package.json +4 -3
  126. package/umd/index.umd.js +224 -27
  127. package/umd/index.umd.js.map +1 -1
  128. package/esm/typings/src/execution/createPipelineExecutor/00-CreatePipelineExecutorSettings.d.ts +0 -29
  129. package/esm/typings/src/scrapers/website/utils/markdownConverter.d.ts +0 -12
  130. package/esm/typings/src/utils/execCommand/IExecCommandOptions.d.ts +0 -23
  131. package/esm/typings/src/utils/execCommand/execCommandNormalizeOptions.d.ts +0 -10
  132. /package/esm/typings/src/{scrapers/website/WebsiteScraper.test.d.ts → executables/apps/locateLibreoffice.test.d.ts} +0 -0
  133. /package/esm/typings/src/{scrapers/website/utils/markdownConverter.test.d.ts → executables/apps/locatePandoc.test.d.ts} +0 -0
@@ -4,9 +4,9 @@ import type { PromptResult } from '../../../execution/PromptResult';
4
4
  *
5
5
  * This is sent from server to client when the generated text is completed
6
6
  */
7
- export interface PromptbookServer_Prompt_Response {
7
+ export type PromptbookServer_Prompt_Response = {
8
8
  /**
9
9
  * The result of the prompt
10
10
  */
11
- promptResult: PromptResult;
12
- }
11
+ readonly promptResult: PromptResult;
12
+ };
@@ -3,12 +3,13 @@ import type { string_base_url } from '../../../types/typeAliases';
3
3
  import type { string_uri } from '../../../types/typeAliases';
4
4
  import type { string_user_id } from '../../../types/typeAliases';
5
5
  import type { LlmToolsConfiguration } from '../../_common/register/LlmToolsConfiguration';
6
+ import type { CollectionRemoteServerClientOptions } from './RemoteServerOptions';
6
7
  /**
7
8
  * Options for `RemoteLlmExecutionTools`
8
9
  *
9
10
  * @public exported from `@promptbook/remote-client`
10
11
  */
11
- export type RemoteLlmExecutionToolsOptions = CommonToolsOptions & {
12
+ export type RemoteLlmExecutionToolsOptions<TCustomOptions> = CommonToolsOptions & {
12
13
  /**
13
14
  * URL of the remote PROMPTBOOK server
14
15
  * On this server will be connected to the socket.io server
@@ -29,7 +30,7 @@ export type RemoteLlmExecutionToolsOptions = CommonToolsOptions & {
29
30
  /**
30
31
  * Use anonymous server with anonymous mode
31
32
  */
32
- isAnonymous: true;
33
+ readonly isAnonymous: true;
33
34
  /**
34
35
  * Configuration for the LLM tools
35
36
  */
@@ -37,22 +38,16 @@ export type RemoteLlmExecutionToolsOptions = CommonToolsOptions & {
37
38
  /**
38
39
  * Identifier of the end user
39
40
  *
40
- * Note: this is passed to the certain model providers to identify misuse
41
- * Note: In anonymous mode it is not required to identify
41
+ * Note: This is passed to the certain model providers to identify misuse
42
+ * Note: In anonymous mode, there is no need to identify yourself, nor does it change the actual configuration of LLM Tools (unlike in application mode).
42
43
  */
43
- readonly userId?: string_user_id;
44
- } | {
44
+ readonly userId: string_user_id | null;
45
+ } | ({
45
46
  /**
46
47
  * Use anonymous server with client identification and fixed collection
47
48
  */
48
- isAnonymous: false;
49
- /**
50
- * Identifier of the end user
51
- *
52
- * Note: this is passed to the certain model providers to identify misuse
53
- */
54
- readonly userId: string_user_id;
55
- });
49
+ readonly isAnonymous: false;
50
+ } & CollectionRemoteServerClientOptions<TCustomOptions>));
56
51
  /**
57
52
  * TODO: [🧠][🛍] Maybe not `isAnonymous: boolean` BUT `mode: 'ANONYMOUS'|'COLLECTION'`
58
53
  * TODO: [🧠][🧜‍♂️] Maybe join remoteUrl and path into single value
@@ -1,6 +1,8 @@
1
+ import type { Promisable } from 'type-fest';
1
2
  import type { PipelineCollection } from '../../../collection/PipelineCollection';
2
3
  import type { CommonToolsOptions } from '../../../execution/CommonToolsOptions';
3
4
  import type { LlmExecutionTools } from '../../../execution/LlmExecutionTools';
5
+ import type { string_app_id } from '../../../types/typeAliases';
4
6
  import type { string_uri } from '../../../types/typeAliases';
5
7
  import type { string_user_id } from '../../../types/typeAliases';
6
8
  /**
@@ -8,7 +10,7 @@ import type { string_user_id } from '../../../types/typeAliases';
8
10
  *
9
11
  * There are two modes of remote server:
10
12
  *
11
- * 1) **Collection mode** Server will recieve `collection` and execute prompts only from this collection
13
+ * 1) **Application mode** Server will recieve `collection` and execute prompts only from this collection
12
14
  * 2) **Anonymous mode** Server will recieve full `LlmToolsConfiguration` (with api keys) and just acts as a proxy
13
15
  * In anonymous mode, `collection` will be ignored and any prompt will be executed
14
16
  *
@@ -17,7 +19,7 @@ import type { string_user_id } from '../../../types/typeAliases';
17
19
  * @public exported from `@promptbook/remote-client`
18
20
  * @public exported from `@promptbook/remote-server`
19
21
  */
20
- export type RemoteServerOptions = CommonToolsOptions & {
22
+ export type RemoteServerOptions<TCustomOptions> = CommonToolsOptions & {
21
23
  /**
22
24
  * Port on which the server will listen
23
25
  */
@@ -29,18 +31,18 @@ export type RemoteServerOptions = CommonToolsOptions & {
29
31
  * @example '/promptbook/socket.io'
30
32
  */
31
33
  readonly path: string_uri;
32
- } & (AnonymousRemoteServerOptions | CollectionRemoteServerOptions | (AnonymousRemoteServerOptions & CollectionRemoteServerOptions));
34
+ } & (AnonymousRemoteServerOptions | CollectionRemoteServerOptions<TCustomOptions> | (AnonymousRemoteServerOptions & CollectionRemoteServerOptions<TCustomOptions>));
33
35
  export type AnonymousRemoteServerOptions = {
34
36
  /**
35
37
  * Enable anonymous mode
36
38
  */
37
39
  readonly isAnonymousModeAllowed: true;
38
40
  };
39
- export type CollectionRemoteServerOptions = {
41
+ export type CollectionRemoteServerOptions<TCustomOptions> = {
40
42
  /**
41
- * Enable collection mode
43
+ * Enable application mode
42
44
  */
43
- readonly isCollectionModeAllowed: true;
45
+ readonly isApplicationModeAllowed: true;
44
46
  /**
45
47
  * Promptbook collection to use
46
48
  *
@@ -50,9 +52,22 @@ export type CollectionRemoteServerOptions = {
50
52
  /**
51
53
  * Creates llm execution tools for each client
52
54
  */
53
- createLlmExecutionTools(userId: string_user_id | undefined): LlmExecutionTools;
55
+ createLlmExecutionTools(options: CollectionRemoteServerClientOptions<TCustomOptions>): Promisable<LlmExecutionTools>;
56
+ };
57
+ export type CollectionRemoteServerClientOptions<TCustomOptions> = {
58
+ /**
59
+ * @@@
60
+ */
61
+ readonly appId: string_app_id | null;
62
+ /**
63
+ * @@@
64
+ */
65
+ readonly userId: string_user_id | null;
66
+ /**
67
+ * @@@
68
+ */
69
+ readonly customOptions?: TCustomOptions;
54
70
  };
55
71
  /**
56
72
  * TODO: Constrain anonymous mode for specific models / providers
57
- * TODO: [🧠][🤺] Remove `createLlmExecutionTools`, pass just `llmExecutionTools`
58
73
  */
@@ -9,7 +9,7 @@ import type { RemoteServerOptions } from './interfaces/RemoteServerOptions';
9
9
  * @see https://github.com/webgptorg/promptbook#remote-server
10
10
  * @public exported from `@promptbook/remote-server`
11
11
  */
12
- export declare function startRemoteServer(options: RemoteServerOptions): IDestroyable;
12
+ export declare function startRemoteServer<TCustomOptions = undefined>(options: RemoteServerOptions<TCustomOptions>): IDestroyable;
13
13
  /**
14
14
  * TODO: Maybe use `$asDeeplyFrozenSerializableJson`
15
15
  * TODO: [🧠][🛍] Maybe not `isAnonymous: boolean` BUT `mode: 'ANONYMOUS'|'COLLECTION'`
@@ -1,3 +1,4 @@
1
+ import type { IntermediateFilesStrategy } from '../types/IntermediateFilesStrategy';
1
2
  import type { string_dirname } from '../types/typeAliases';
2
3
  /**
3
4
  * Options for preparation of the pipeline
@@ -21,37 +22,17 @@ export type PrepareAndScrapeOptions = {
21
22
  */
22
23
  readonly cacheDirname?: string_dirname;
23
24
  /**
24
- * If true, the cache is cleaned after the scraping
25
+ * Strategy for caching the intermediate results for knowledge source
25
26
  *
26
- *
27
- * @default false // <- TODO: !!!!!! Change to `cacheStrategy`/`intermediateFiles`, Put to global config, change to `true` and explicitly set to `false` in all playgrounds
27
+ * @default DEFAULT_INTERMEDIATE_FILES_STRATEGY
28
28
  */
29
- readonly isCacheCleaned?: boolean;
29
+ readonly intermediateFilesStrategy?: IntermediateFilesStrategy;
30
30
  /**
31
31
  * Maximum number of tasks running in parallel
32
32
  *
33
33
  * @default MAX_PARALLEL_COUNT
34
34
  */
35
35
  readonly maxParallelCount?: number;
36
- /**
37
- * Path to the external programs executables
38
- *
39
- * TODO: !!!!!! Transform to scrapers and make them Classy
40
- */
41
- readonly externalProgramsPaths?: {
42
- /**
43
- * Path to the `pandoc` executable
44
- *
45
- * @example 'C:/Users/me/AppData/Local/Pandoc/pandoc.exe'
46
- */
47
- readonly pandocPath?: string;
48
- /**
49
- * Path to the LibreOffice executable
50
- *
51
- * @example 'C:/Program Files/LibreOffice/program/swriter.exe'
52
- */
53
- readonly libreOfficePath?: string;
54
- };
55
36
  /**
56
37
  * If true, the missing software is automatically installed
57
38
  */
@@ -12,7 +12,7 @@ type PreparedTemplates = {
12
12
  /**
13
13
  * @@@ Sequence of templates that are chained together to form a pipeline
14
14
  */
15
- readonly templatesPrepared: Array<TemplateJson>;
15
+ readonly templatesPrepared: ReadonlyArray<TemplateJson>;
16
16
  };
17
17
  /**
18
18
  * @@@
@@ -17,7 +17,7 @@ export type Scraper = {
17
17
  /**
18
18
  * Scrapes the markdown file and returns the knowledge pieces or `null` if it can't scrape it
19
19
  */
20
- scrape(source: ScraperSourceHandler): Promisable<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
20
+ scrape(source: ScraperSourceHandler): Promisable<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
21
21
  };
22
22
  /**
23
23
  * @@@
@@ -51,16 +51,9 @@ export type ScraperSourceHandler = {
51
51
  * Get the content as a utf-8 string
52
52
  */
53
53
  asText(): Promisable<string>;
54
- /**
55
- * Get the content as a blob
56
- */
57
- asBlob(): Promisable<Blob>;
58
54
  };
59
55
  /**
56
+ * TODO: [🥽] Add ` asBlob(): Promisable<Blob>;` or asFile
60
57
  * TODO: [🐝] @@@ Annotate all
61
58
  * TODO: [🔼] Export via types
62
59
  */
63
- /**
64
- * TODO: !!!!!! Test that this is catched
65
- * Note: [⚫] Code in this file should never be published in any package
66
- */
@@ -8,7 +8,7 @@ import type { ExecutionTools } from '../../execution/ExecutionTools';
8
8
  * @see https://github.com/webgptorg/promptbook/discussions/41
9
9
  * @public exported from `@promptbook/core`
10
10
  */
11
- export declare function prepareKnowledgePieces(knowledgeSources: Array<KnowledgeSourceJson>, tools: Pick<ExecutionTools, 'llm' | 'fs' | 'scrapers'>, options: PrepareAndScrapeOptions): Promise<Array<Omit<KnowledgePiecePreparedJson, 'preparationIds'>>>;
11
+ export declare function prepareKnowledgePieces(knowledgeSources: ReadonlyArray<KnowledgeSourceJson>, tools: Pick<ExecutionTools, 'llm' | 'fs' | 'scrapers'>, options: PrepareAndScrapeOptions): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'preparationIds'>>>;
12
12
  /**
13
13
  * TODO: [🧊] In future one preparation can take data from previous preparation and save tokens and time
14
14
  * Put `knowledgePieces` into `PrepareKnowledgeOptions`
@@ -2,11 +2,11 @@ import type { ExecutionTools } from '../../../execution/ExecutionTools';
2
2
  import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
3
3
  import type { Scraper } from '../Scraper';
4
4
  /**
5
- * @@@!!!!!!
5
+ * @@@
6
6
  *
7
7
  * 1) @@@
8
8
  * 2) @@@
9
9
  *
10
10
  * @public exported from `@promptbook/browser`
11
11
  */
12
- export declare function $provideScrapersForBrowser(tools: Pick<ExecutionTools, 'llm'>, options?: PrepareAndScrapeOptions): Promise<Array<Scraper>>;
12
+ export declare function $provideScrapersForBrowser(tools: Pick<ExecutionTools, 'llm'>, options?: PrepareAndScrapeOptions): Promise<ReadonlyArray<Scraper>>;
@@ -2,14 +2,14 @@ import type { ExecutionTools } from '../../../execution/ExecutionTools';
2
2
  import type { PrepareAndScrapeOptions } from '../../../prepare/PrepareAndScrapeOptions';
3
3
  import type { Scraper } from '../Scraper';
4
4
  /**
5
- * !!!!!!
5
+ * @@@
6
6
  *
7
7
  * 1) @@@
8
8
  * 2) @@@
9
9
  *
10
10
  * @public exported from `@promptbook/node`
11
11
  */
12
- export declare function $provideScrapersForNode(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options?: PrepareAndScrapeOptions): Promise<Array<Scraper>>;
12
+ export declare function $provideScrapersForNode(tools: Pick<ExecutionTools, 'fs' | 'llm' | 'executables'>, options?: PrepareAndScrapeOptions): Promise<ReadonlyArray<Scraper>>;
13
13
  /**
14
14
  * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
15
15
  */
@@ -1,8 +1,8 @@
1
+ import type { Executables } from '../../../execution/Executables';
1
2
  import type { string_mime_type } from '../../../types/typeAliases';
2
3
  import type { string_promptbook_documentation_url } from '../../../types/typeAliases';
3
4
  import type { string_title } from '../../../types/typeAliases';
4
5
  import type { Registered } from '../../../utils/$Register';
5
- import type { TODO_any } from '../../../utils/organization/TODO_any';
6
6
  /**
7
7
  * @@@
8
8
  *
@@ -33,7 +33,7 @@ export type ScraperAndConverterMetadata = Registered & {
33
33
  /**
34
34
  * @@@
35
35
  */
36
- readonly requiredExecutables: TODO_any;
36
+ readonly requiredExecutables: ReadonlyArray<Capitalize<keyof Executables extends `${infer N}Path` ? N : never>>;
37
37
  /**
38
38
  * Link to documentation
39
39
  */
@@ -13,7 +13,7 @@ type GetScraperIntermediateSourceSource = Pick<ScraperSourceHandler, 'filename'
13
13
  *
14
14
  * @private internal utility of `getScraperIntermediateSource`
15
15
  */
16
- type GetScraperIntermediateSourceHandler = Required<Pick<PrepareAndScrapeOptions, 'rootDirname' | 'cacheDirname' | 'isCacheCleaned' | 'isVerbose'>> & {
16
+ type GetScraperIntermediateSourceHandler = Required<Pick<PrepareAndScrapeOptions, 'rootDirname' | 'cacheDirname' | 'intermediateFilesStrategy' | 'isVerbose'>> & {
17
17
  readonly extension: string_file_extension;
18
18
  };
19
19
  /**
@@ -28,7 +28,7 @@ export {};
28
28
  /**
29
29
  * Note: Not using `FileCacheStorage` for two reasons:
30
30
  * 1) Need to store more than serialized JSONs
31
- * 2) Need to switch between a `rootDirname` and `cacheDirname` <- TODO: !!!!
31
+ * 2) Need to switch between a `rootDirname` and `cacheDirname` <- TODO: [😡]
32
32
  * TODO: [🐱‍🐉][🧠] Make some smart crop
33
33
  * Note: [🟢] Code in this file should never be never released in packages that could be imported into browser environment
34
34
  */
@@ -6,6 +6,6 @@ import type { ScraperSourceHandler } from '../Scraper';
6
6
  /**
7
7
  * @@@
8
8
  *
9
- * @private for scraper utilities
9
+ * @public exported from `@promptbook/core`
10
10
  */
11
11
  export declare function makeKnowledgeSourceHandler(knowledgeSource: SetOptional<KnowledgeSourceJson, 'name'>, tools: Pick<ExecutionTools, 'fs'>, options?: Pick<PrepareAndScrapeOptions, 'rootDirname' | 'isVerbose'>): Promise<ScraperSourceHandler>;
@@ -23,7 +23,7 @@ export declare class DocumentScraper implements Converter, Scraper {
23
23
  * Markdown scraper is used internally
24
24
  */
25
25
  private readonly markdownScraper;
26
- constructor(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options: PrepareAndScrapeOptions);
26
+ constructor(tools: Pick<ExecutionTools, 'fs' | 'llm' | 'executables'>, options: PrepareAndScrapeOptions);
27
27
  /**
28
28
  * Convert the `.docx` or `.odt` to `.md` file and returns intermediate source
29
29
  *
@@ -33,7 +33,7 @@ export declare class DocumentScraper implements Converter, Scraper {
33
33
  /**
34
34
  * Scrapes the docx file and returns the knowledge pieces or `null` if it can't scrape it
35
35
  */
36
- scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
36
+ scrape(source: ScraperSourceHandler): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
37
37
  }
38
38
  /**
39
39
  * TODO: [👣] Converted documents can act as cached items - there is no need to run conversion each time
@@ -13,7 +13,7 @@ export declare const createDocumentScraper: ((tools: Pick<ExecutionTools, 'llm'>
13
13
  mimeTypes: string[];
14
14
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
15
15
  isAvilableInBrowser: false;
16
- requiredExecutables: string[];
16
+ requiredExecutables: "Pandoc"[];
17
17
  }>;
18
18
  /**
19
19
  * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
@@ -11,7 +11,7 @@ export declare const documentScraperMetadata: import("type-fest/source/readonly-
11
11
  mimeTypes: string[];
12
12
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
13
13
  isAvilableInBrowser: false;
14
- requiredExecutables: string[];
14
+ requiredExecutables: "Pandoc"[];
15
15
  }>;
16
16
  /**
17
17
  * Registration of known scraper metadata
@@ -7,7 +7,7 @@ import type { Scraper } from '../_common/Scraper';
7
7
  import type { ScraperSourceHandler } from '../_common/Scraper';
8
8
  import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
9
9
  /**
10
- * Scraper for .docx files
10
+ * Scraper for old document files (like .doc and .rtf)
11
11
  *
12
12
  * @see `documentationUrl` for more details
13
13
  * @public exported from `@promptbook/legacy-documents`
@@ -23,7 +23,7 @@ export declare class LegacyDocumentScraper implements Converter, Scraper {
23
23
  * Document scraper is used internally
24
24
  */
25
25
  private readonly documentScraper;
26
- constructor(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options: PrepareAndScrapeOptions);
26
+ constructor(tools: Pick<ExecutionTools, 'fs' | 'llm' | 'executables'>, options: PrepareAndScrapeOptions);
27
27
  /**
28
28
  * Convert the `.doc` or `.rtf` to `.doc` file and returns intermediate source
29
29
  *
@@ -33,7 +33,7 @@ export declare class LegacyDocumentScraper implements Converter, Scraper {
33
33
  /**
34
34
  * Scrapes the `.doc` or `.rtf` file and returns the knowledge pieces or `null` if it can't scrape it
35
35
  */
36
- scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
36
+ scrape(source: ScraperSourceHandler): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
37
37
  }
38
38
  /**
39
39
  * TODO: [👣] Converted documents can act as cached items - there is no need to run conversion each time
@@ -13,7 +13,7 @@ export declare const createLegacyDocumentScraper: ((tools: Pick<ExecutionTools,
13
13
  mimeTypes: string[];
14
14
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
15
15
  isAvilableInBrowser: false;
16
- requiredExecutables: string[];
16
+ requiredExecutables: ("Pandoc" | "LibreOffice")[];
17
17
  }>;
18
18
  /**
19
19
  * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
@@ -11,7 +11,7 @@ export declare const legacyDocumentScraperMetadata: import("type-fest/source/rea
11
11
  mimeTypes: string[];
12
12
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
13
13
  isAvilableInBrowser: false;
14
- requiredExecutables: string[];
14
+ requiredExecutables: ("Pandoc" | "LibreOffice")[];
15
15
  }>;
16
16
  /**
17
17
  * Registration of known scraper metadata
@@ -21,7 +21,7 @@ export declare class MarkdownScraper implements Scraper {
21
21
  /**
22
22
  * Scrapes the markdown file and returns the knowledge pieces or `null` if it can't scrape it
23
23
  */
24
- scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
24
+ scrape(source: ScraperSourceHandler): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
25
25
  }
26
26
  /**
27
27
  * TODO: [🪂] Do it in parallel 11:11
@@ -13,7 +13,7 @@ export declare const createMarkdownScraper: ((tools: Pick<ExecutionTools, 'llm'>
13
13
  mimeTypes: string[];
14
14
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
15
15
  isAvilableInBrowser: true;
16
- requiredExecutables: string[];
16
+ requiredExecutables: never[];
17
17
  }>;
18
18
  /**
19
19
  * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
@@ -11,7 +11,7 @@ export declare const markdownScraperMetadata: import("type-fest/source/readonly-
11
11
  mimeTypes: string[];
12
12
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
13
13
  isAvilableInBrowser: true;
14
- requiredExecutables: string[];
14
+ requiredExecutables: never[];
15
15
  }>;
16
16
  /**
17
17
  * Registration of known scraper metadata
@@ -7,7 +7,7 @@ import type { Converter } from '../_common/Converter';
7
7
  import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
8
8
  import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
9
9
  /**
10
- * Scraper for .docx files
10
+ * Scraper for .pdf files
11
11
  *
12
12
  * @see `documentationUrl` for more details
13
13
  * @public exported from `@promptbook/pdf`
@@ -31,7 +31,7 @@ export declare class PdfScraper implements Converter, Scraper {
31
31
  /**
32
32
  * Scrapes the `.pdf` file and returns the knowledge pieces or `null` if it can't scrape it
33
33
  */
34
- scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
34
+ scrape(source: ScraperSourceHandler): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
35
35
  }
36
36
  /**
37
37
  * TODO: [👣] Converted pdf documents can act as cached items - there is no need to run conversion each time
@@ -13,7 +13,7 @@ export declare const createPdfScraper: ((tools: Pick<ExecutionTools, 'llm'>, opt
13
13
  mimeTypes: string[];
14
14
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
15
15
  isAvilableInBrowser: true;
16
- requiredExecutables: string[];
16
+ requiredExecutables: never[];
17
17
  }>;
18
18
  /**
19
19
  * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
@@ -11,7 +11,7 @@ export declare const pdfScraperMetadata: import("type-fest/source/readonly-deep"
11
11
  mimeTypes: string[];
12
12
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
13
13
  isAvilableInBrowser: true;
14
- requiredExecutables: string[];
14
+ requiredExecutables: never[];
15
15
  }>;
16
16
  /**
17
17
  * Registration of known scraper metadata
@@ -8,7 +8,7 @@ import type { PrepareAndScrapeOptions } from '../../prepare/PrepareAndScrapeOpti
8
8
  import type { ScraperAndConverterMetadata } from '../_common/register/ScraperAndConverterMetadata';
9
9
  import type { ScraperIntermediateSource } from '../_common/ScraperIntermediateSource';
10
10
  /**
11
- * Scraper for .docx files
11
+ * Scraper for websites
12
12
  *
13
13
  * @see `documentationUrl` for more details
14
14
  * @public exported from `@promptbook/website-crawler`
@@ -24,6 +24,10 @@ export declare class WebsiteScraper implements Converter, Scraper {
24
24
  * Markdown scraper is used internally
25
25
  */
26
26
  private readonly markdownScraper;
27
+ /**
28
+ * Showdown converter is used internally
29
+ */
30
+ private readonly showdownConverter;
27
31
  constructor(tools: Pick<ExecutionTools, 'fs' | 'llm'>, options: PrepareAndScrapeOptions);
28
32
  /**
29
33
  * Convert the website to `.md` file and returns intermediate source
@@ -36,10 +40,9 @@ export declare class WebsiteScraper implements Converter, Scraper {
36
40
  /**
37
41
  * Scrapes the website and returns the knowledge pieces or `null` if it can't scrape it
38
42
  */
39
- scrape(source: ScraperSourceHandler): Promise<Array<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
43
+ scrape(source: ScraperSourceHandler): Promise<ReadonlyArray<Omit<KnowledgePiecePreparedJson, 'sources' | 'preparationIds'>> | null>;
40
44
  }
41
45
  /**
42
- * TODO: !!!!!! Put into separate package
43
46
  * TODO: [👣] Scraped website in .md can act as cache item - there is no need to run conversion each time
44
47
  * TODO: [🪂] Do it in parallel 11:11
45
48
  * Note: No need to aggregate usage here, it is done by intercepting the llmTools
@@ -13,7 +13,7 @@ export declare const createWebsiteScraper: ((tools: Pick<ExecutionTools, 'llm'>,
13
13
  mimeTypes: string[];
14
14
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
15
15
  isAvilableInBrowser: false;
16
- requiredExecutables: string[];
16
+ requiredExecutables: never[];
17
17
  }>;
18
18
  /**
19
19
  * TODO: [🎶] Naming "constructor" vs "creator" vs "factory"
@@ -11,7 +11,7 @@ export declare const websiteScraperMetadata: import("type-fest/source/readonly-d
11
11
  mimeTypes: string[];
12
12
  documentationUrl: "https://github.com/webgptorg/promptbook/discussions/@@";
13
13
  isAvilableInBrowser: false;
14
- requiredExecutables: string[];
14
+ requiredExecutables: never[];
15
15
  }>;
16
16
  /**
17
17
  * Registration of known scraper metadata
@@ -0,0 +1,7 @@
1
+ import { Converter as ShowdownConverter } from 'showdown';
2
+ /**
3
+ * Create a new showdown converter instance
4
+ *
5
+ * @private utility of `WebsiteScraper`
6
+ */
7
+ export declare function createShowdownConverter(): ShowdownConverter;
@@ -7,7 +7,7 @@ import type { really_any } from '../../../utils/organization/really_any';
7
7
  * @returns nothing
8
8
  * @private internal function of `JavascriptExecutionTools` and `JavascriptEvalExecutionTools`
9
9
  */
10
- export declare function preserve(func: (...params: Array<really_any>) => unknown): void;
10
+ export declare function preserve(func: (...params: ReadonlyArray<really_any>) => unknown): void;
11
11
  /**
12
12
  * TODO: !! [1] This maybe does memory leak
13
13
  */
@@ -4,4 +4,4 @@ import type { string_name } from '../../../types/typeAliases';
4
4
  *
5
5
  * @private for `FileCacheStorage`
6
6
  */
7
- export declare function nameToSubfolderPath(name: string_name): Array<string>;
7
+ export declare function nameToSubfolderPath(name: string_name): ReadonlyArray<string>;
@@ -7,4 +7,4 @@
7
7
  *
8
8
  * @private internal type
9
9
  */
10
- export type Arrayable<TItem> = TItem | Array<TItem> | undefined;
10
+ export type Arrayable<TItem> = TItem | ReadonlyArray<TItem> | undefined;
@@ -0,0 +1,7 @@
1
+ /**
2
+ * @@@
3
+ */
4
+ export type IntermediateFilesStrategy = 'HIDE_AND_CLEAN' | 'HIDE_AND_KEEP';
5
+ /**
6
+ * TODO: [🎅] Maube add options for all kinds of cache, unite with `isCacheReloaded` and make `cacheStrategy` (which is not specific only for intermediate files)
7
+ */
@@ -31,7 +31,7 @@ export type KnowledgePiecePreparedJson = {
31
31
  /**
32
32
  * List of sources where the information comes from
33
33
  */
34
- readonly sources: Array<{
34
+ readonly sources: ReadonlyArray<{
35
35
  /**
36
36
  * Identifier of the source
37
37
  */
@@ -48,11 +48,11 @@ export type KnowledgePiecePreparedJson = {
48
48
  /**
49
49
  * List of keywords that are associated with the knowledge piece
50
50
  */
51
- readonly keywords: Array<string_keyword>;
51
+ readonly keywords: ReadonlyArray<string_keyword>;
52
52
  /**
53
53
  * List of models embeddings that are associated with the knowledge piece
54
54
  */
55
- readonly index: Array<{
55
+ readonly index: ReadonlyArray<{
56
56
  /**
57
57
  * Model name which generated the embedding
58
58
  */
@@ -65,7 +65,7 @@ export type KnowledgePiecePreparedJson = {
65
65
  /**
66
66
  * List of preparation ids that were used to prepare this knowledge piece
67
67
  */
68
- readonly preparationIds: Array<number_id>;
68
+ readonly preparationIds: ReadonlyArray<number_id>;
69
69
  };
70
70
  /**
71
71
  * TODO: [🍙] Make some standard order of json properties
@@ -37,7 +37,7 @@ export type KnowledgeSourcePreparedJson = KnowledgeSourceJson & {
37
37
  /**
38
38
  * List of preparation ids that were used to prepare this knowledge source to knowledge pieces
39
39
  */
40
- readonly preparationIds: Array<number_id>;
40
+ readonly preparationIds: ReadonlyArray<number_id>;
41
41
  };
42
42
  /**
43
43
  * TODO: [🍙] Make some standard order of json properties
@@ -38,7 +38,7 @@ export type PersonaPreparedJson = PersonaJson & {
38
38
  /**
39
39
  * List of preparation ids that were used to prepare this persona
40
40
  */
41
- readonly preparationIds: Array<number_id>;
41
+ readonly preparationIds: ReadonlyArray<number_id>;
42
42
  };
43
43
  /**
44
44
  * TODO: [🍙] Make some standard order of json properties