chore: rename files in core module (#277)

---------

Co-authored-by: zhouxiao.shaw <[email protected]>

Author: yuyutaotao

apps/site/docs/zh/index.mdx

@@ -1,6 +1,6 @@
 # Midscene.js - AI 加持，带来愉悦的 UI 自动化体验
 
-UI 自动化太难维护了。UI 自动化脚本里往往到处都是选择器，比如 `#ids`、`data-test`、`.selectors`。在需要重构的时候，这可能会让人感到非常头疼，尽管在这种情况下，UI 自动化应该能够发挥作用。
+传统 UI 自动化太难维护了。自动化脚本里往往到处都是选择器，比如 `#ids`、`data-test`、`.selectors`。在需要重构的时候，这可能会让人感到非常头疼，尽管在这种情况下，UI 自动化应该能够发挥作用。
 
 我们在这里推出 Midscene.js，助你重拾编码的乐趣。
 

packages/midscene/src/action/executor.ts renamed to packages/midscene/src/ai-model/action-executor.ts

@@ -1,13 +1,11 @@
 import assert from 'node:assert';
-import { MIDSCENE_MODEL_TEXT_ONLY, getAIConfig } from '@/env';
 import type { AIUsageInfo } from '@/types';
 
 import type {
-  ChatCompletionContentPart,
   ChatCompletionSystemMessageParam,
   ChatCompletionUserMessageParam,
 } from 'openai/resources';
-import { callToGetJSONObject, checkAIConfig } from './openai';
+import { callToGetJSONObject, checkAIConfig } from './service-caller';
 
 export type AIArgs = [
   ChatCompletionSystemMessageParam,

packages/midscene/src/ai-model/common.ts

@@ -1,5 +1,5 @@
-export { callToGetJSONObject } from './openai';
-export { systemPromptToFindElement } from './prompt/element-inspector';
+export { callToGetJSONObject } from './service-caller';
+export { systemPromptToLocateElement } from './prompt/llm-locator';
 export { describeUserPage } from './prompt/util';
 
 export type { ChatCompletionMessageParam } from 'openai/resources';
@@ -11,7 +11,6 @@ export {
   transformElementPositionToId,
 } from './inspect';
 
-export { plan } from './automation';
+export { plan } from './llm-planning';
 export { callAiFn } from './common';
-
-export { vlmPlanning } from './vlm-planning';
+export { vlmPlanning } from './ui-tars-planning';

packages/midscene/src/ai-model/index.ts

@@ -1,5 +1,4 @@
 import assert from 'node:assert';
-import path from 'node:path';
 import type {
   AIAssertionResponse,
   AIElementResponse,
@@ -9,29 +8,26 @@ import type {
   AIUsageInfo,
   BaseElement,
   ElementById,
-  Point,
   Size,
   UIContext,
 } from '@/types';
-import { savePositionImg } from '@midscene/shared/img';
 import type {
   ChatCompletionSystemMessageParam,
   ChatCompletionUserMessageParam,
 } from 'openai/resources';
 import { AIActionType, callAiFn } from './common';
-import { call, callToGetJSONObject } from './openai';
 import {
   findElementPrompt,
-  multiDescription,
-  systemPromptToFindElement,
-} from './prompt/element-inspector';
+  systemPromptToLocateElement,
+} from './prompt/llm-locator';
 import {
   describeUserPage,
   elementByPositionWithElementInfo,
   extractDataPrompt,
   systemPromptToAssert,
   systemPromptToExtract,
 } from './prompt/util';
+import { callToGetJSONObject } from './service-caller';
 
 export type AIArgs = [
   ChatCompletionSystemMessageParam,
@@ -69,11 +65,7 @@ export async function transformElementPositionToId(
       },
       size,
     );
-    // await savePositionImg({
-    //   inputImgBase64: screenshotBase64,
-    //   rect: absolutePosition,
-    //   outputPath: path.join(__dirname, 'test-data', `output-${index++}.png`),
-    // });
+
     const element = elementByPositionWithElementInfo(
       elementsInfo,
       absolutePosition,
@@ -182,7 +174,7 @@ export async function AiInspectElement<
     targetElementDescription,
     multi,
   });
-  const systemPrompt = systemPromptToFindElement();
+  const systemPrompt = systemPromptToLocateElement();
   const msgs: AIArgs = [
     { role: 'system', content: systemPrompt },
     {

packages/midscene/src/ai-model/inspect.ts

@@ -6,7 +6,7 @@ import {
   automationUserPrompt,
   generateTaskBackgroundContext,
   systemPromptToTaskPlanning,
-} from './prompt/planning';
+} from './prompt/llm-planning';
 import { describeUserPage } from './prompt/util';
 
 export async function plan(

packages/midscene/src/ai-model/automation.ts renamed to packages/midscene/src/ai-model/llm-planning.ts

@@ -1,11 +1,11 @@
 import { MATCH_BY_POSITION, getAIConfig } from '@/env';
 import { PromptTemplate } from '@langchain/core/prompts';
 import type { ResponseFormatJSONSchema } from 'openai/resources';
-import { systemPromptToFindElementPosition } from './element-point';
+import { systemPromptToLocateElementPosition } from './ui-tars-locator';
 
-export function systemPromptToFindElement() {
+export function systemPromptToLocateElement() {
   if (getAIConfig(MATCH_BY_POSITION)) {
-    return systemPromptToFindElementPosition();
+    return systemPromptToLocateElementPosition();
   }
   return `
 ## Role:
@@ -144,12 +144,6 @@ Output Example:
   `;
 }
 
-export function multiDescription(multi: boolean) {
-  return multi
-    ? 'multiple elements matching the description (two or more)'
-    : 'The element closest to the description (only one)';
-}
-
 export const findElementSchema: ResponseFormatJSONSchema = {
   type: 'json_schema',
   json_schema: {

packages/midscene/src/ai-model/prompt/element-inspector.ts renamed to packages/midscene/src/ai-model/prompt/llm-locator.ts

@@ -1,5 +1,5 @@
 // claude 3.5 sonnet computer The ability to understand the content of the image is better, Does not provide element snapshot effect
-export function systemPromptToFindElementPosition() {
+export function systemPromptToLocateElementPosition() {
   return `
   You are a GUI agent. You are given a task and your action history, with screenshots. You need to perform the next action to complete the task.
   

packages/midscene/src/ai-model/prompt/planning.ts renamed to packages/midscene/src/ai-model/prompt/llm-planning.ts

@@ -1,4 +1,4 @@
-import type { Action } from '../vlm-planning';
+import type { Action } from '../ui-tars-planning';
 
 export const uiTarsPlanningPrompt = `
 You are a GUI agent. You are given a task and your action history, with screenshots. You need to perform the next action to complete the task.