mirror of
https://github.com/labring/FastGPT.git
synced 2025-07-23 21:13:50 +00:00
修改插件：Doc2X插件适配文件上传功能 (#2284)
* Change to download first and check contentType * Up to date * Some bug fix, still some bug with img * Update tool to read from file * improve formatting in PDF * Add tool of img file OCR , but meet some bug * Bug fix for parameter passing error. * Modification Introduction
This commit is contained in:
Menghuan1918
@@ -13,6 +13,8 @@ const staticPluginList = [
|
||||
'Doc2X',
|
||||
'Doc2X/URLPDF2text',
|
||||
'Doc2X/URLImg2text',
|
||||
`Doc2X/FilePDF2text`,
|
||||
`Doc2X/FileImg2text`,
|
||||
'feishu'
|
||||
];
|
||||
// Run in worker thread (Have npm packages)
|
||||
|
||||
172
packages/plugins/src/Doc2X/FileImg2text/index.ts
Normal file
172
packages/plugins/src/Doc2X/FileImg2text/index.ts
Normal file
@@ -0,0 +1,172 @@
|
||||
import { delay } from '@fastgpt/global/common/system/utils';
|
||||
import { addLog } from '@fastgpt/service/common/system/log';
|
||||
|
||||
type Props = {
|
||||
apikey: string;
|
||||
files: Array<string>;
|
||||
img_correction: boolean;
|
||||
formula: boolean;
|
||||
};
|
||||
|
||||
type Response = Promise<{
|
||||
result: string;
|
||||
failreason: string;
|
||||
success: boolean;
|
||||
}>;
|
||||
|
||||
const main = async ({ apikey, files, img_correction, formula }: Props): Response => {
|
||||
// Check the apikey
|
||||
if (!apikey) {
|
||||
return {
|
||||
result: '',
|
||||
failreason: `API key is required`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
let real_api_key = apikey;
|
||||
if (!apikey.startsWith('sk-')) {
|
||||
const response = await fetch('https://api.doc2x.noedgeai.com/api/token/refresh', {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
Authorization: `Bearer ${apikey}`
|
||||
}
|
||||
});
|
||||
if (response.status !== 200) {
|
||||
return {
|
||||
result: '',
|
||||
failreason: `Get token failed: ${await response.text()}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
const data = await response.json();
|
||||
real_api_key = data.data.token;
|
||||
}
|
||||
|
||||
let final_result = '';
|
||||
let fail_reason = '';
|
||||
let flag = false;
|
||||
//Process each file one by one
|
||||
for await (const url of files) {
|
||||
// Fetch the image and check its content type
|
||||
const imageResponse = await fetch(url);
|
||||
if (!imageResponse.ok) {
|
||||
fail_reason += `\n---\nFile:${url} \n<Content>\nFailed to fetch image from URL\n</Content>\n`;
|
||||
flag = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
const contentType = imageResponse.headers.get('content-type');
|
||||
const fileName = url.match(/read\?filename=([^&]+)/)?.[1] || 'unknown.png';
|
||||
if (!contentType || !contentType.startsWith('image/')) {
|
||||
fail_reason += `\n---\nFile:${url} \n<Content>\nThe provided URL does not point to an image: ${contentType}\n</Content>\n`;
|
||||
flag = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
const blob = await imageResponse.blob();
|
||||
const formData = new FormData();
|
||||
formData.append('file', blob, fileName);
|
||||
formData.append('img_correction', img_correction ? '1' : '0');
|
||||
formData.append('equation', formula ? '1' : '0');
|
||||
|
||||
let upload_url = 'https://api.doc2x.noedgeai.com/api/platform/async/img';
|
||||
if (real_api_key.startsWith('sk-')) {
|
||||
upload_url = 'https://api.doc2x.noedgeai.com/api/v1/async/img';
|
||||
}
|
||||
|
||||
let uuid;
|
||||
let upload_flag = true;
|
||||
const uploadAttempts = [1, 2, 3];
|
||||
for await (const attempt of uploadAttempts) {
|
||||
const upload_response = await fetch(upload_url, {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
Authorization: `Bearer ${real_api_key}`
|
||||
},
|
||||
body: formData
|
||||
});
|
||||
|
||||
if (!upload_response.ok) {
|
||||
// Rate limit, wait for 10s and retry at most 3 times
|
||||
if (upload_response.status === 429 && attempt < 3) {
|
||||
await delay(10000);
|
||||
continue;
|
||||
}
|
||||
fail_reason += `\n---\nFile:${fileName}\n<Content>\nFailed to upload file: ${await upload_response.text()}\n</Content>\n`;
|
||||
flag = true;
|
||||
upload_flag = false;
|
||||
break;
|
||||
}
|
||||
if (!upload_flag) {
|
||||
continue;
|
||||
}
|
||||
|
||||
const upload_data = await upload_response.json();
|
||||
uuid = upload_data.data.uuid;
|
||||
break;
|
||||
}
|
||||
|
||||
// Get the result by uuid
|
||||
let result_url = 'https://api.doc2x.noedgeai.com/api/platform/async/status?uuid=' + uuid;
|
||||
if (real_api_key.startsWith('sk-')) {
|
||||
result_url = 'https://api.doc2x.noedgeai.com/api/v1/async/status?uuid=' + uuid;
|
||||
}
|
||||
|
||||
let required_flag = true;
|
||||
const maxAttempts = 100;
|
||||
// Wait for the result, at most 100s
|
||||
for await (const _ of Array(maxAttempts).keys()) {
|
||||
const result_response = await fetch(result_url, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${real_api_key}`
|
||||
}
|
||||
});
|
||||
if (!result_response.ok) {
|
||||
fail_reason += `\n---\nFile:${fileName}\n<Content>\nFailed to get result: ${await result_response.text()}\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
}
|
||||
const result_data = await result_response.json();
|
||||
if (['ready', 'processing'].includes(result_data.data.status)) {
|
||||
await delay(1000);
|
||||
} else if (result_data.data.status === 'pages limit exceeded') {
|
||||
fail_reason += `\n---\nFile:${fileName}\n<Content>\nFailed to get result: pages limit exceeded\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
} else if (result_data.data.status === 'success') {
|
||||
let result;
|
||||
try {
|
||||
result = result_data.data.result.pages[0].md;
|
||||
result = result.replace(/\\[\(\)]/g, '$').replace(/\\[\[\]]/g, '$$');
|
||||
} catch {
|
||||
// no pages
|
||||
final_result += `\n---\nFile:${fileName}\n<Content>\n \n</Content>\n`;
|
||||
required_flag = false;
|
||||
}
|
||||
final_result += `\n---\nFile:${fileName}\n<Content>\n${result}\n</Content>\n`;
|
||||
required_flag = false;
|
||||
break;
|
||||
} else {
|
||||
fail_reason += `\n---\nFile:${fileName}\n<Content>\nFailed to get result: ${result_data.data.status}\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (required_flag) {
|
||||
fail_reason += `\n---\nFile:${fileName}\n<Content>\nTimeout waiting for result\n</Content>\n`;
|
||||
flag = true;
|
||||
}
|
||||
}
|
||||
|
||||
return {
|
||||
result: final_result,
|
||||
failreason: fail_reason,
|
||||
success: !flag
|
||||
};
|
||||
};
|
||||
|
||||
export default main;
|
||||
437
packages/plugins/src/Doc2X/FileImg2text/template.json
Normal file
437
packages/plugins/src/Doc2X/FileImg2text/template.json
Normal file
@@ -0,0 +1,437 @@
|
||||
{
|
||||
"author": "Menghuan1918",
|
||||
"version": "488",
|
||||
"name": "Doc2X 图像(文件)识别",
|
||||
"avatar": "plugins/doc2x",
|
||||
"intro": "将上传的图片文件发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"inputExplanationUrl": "https://fael3z0zfze.feishu.cn/wiki/Rkc5witXWiJoi5kORd2cofh6nDg?fromScene=spaceOverview",
|
||||
"showStatus": true,
|
||||
"weight": 10,
|
||||
|
||||
"isTool": true,
|
||||
"templateType": "tools",
|
||||
|
||||
"workflow": {
|
||||
"nodes": [
|
||||
{
|
||||
"nodeId": "pluginInput",
|
||||
"name": "自定义插件输入",
|
||||
"intro": "可以配置插件需要哪些输入，利用这些输入来运行插件",
|
||||
"avatar": "core/workflow/template/workflowStart",
|
||||
"flowNodeType": "pluginInput",
|
||||
"showStatus": false,
|
||||
"position": {
|
||||
"x": 362.9862638626885,
|
||||
"y": 6.16353826540589
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"renderTypeList": ["input"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"description": "Doc2X的验证密匙，对于个人用户可以从Doc2X官网 - 个人信息 - 身份令牌获得",
|
||||
"required": true,
|
||||
"toolDescription": "",
|
||||
"defaultValue": ""
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "arrayString",
|
||||
"canEdit": true,
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"description": "待处理图片文件",
|
||||
"required": true,
|
||||
"toolDescription": "待处理图片文件"
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["switch"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "img_correction",
|
||||
"label": "img_correction",
|
||||
"description": "是否启用图形矫正功能",
|
||||
"required": true,
|
||||
"toolDescription": "",
|
||||
"defaultValue": false
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["switch"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "formula",
|
||||
"label": "formula",
|
||||
"description": "是否开启纯公式识别(仅适用于图片内容仅有公式时)",
|
||||
"required": true,
|
||||
"toolDescription": "",
|
||||
"defaultValue": false
|
||||
}
|
||||
],
|
||||
"outputs": [
|
||||
{
|
||||
"id": "apikey",
|
||||
"valueType": "string",
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"type": "hidden"
|
||||
},
|
||||
{
|
||||
"id": "url",
|
||||
"valueType": "arrayString",
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"type": "hidden"
|
||||
},
|
||||
{
|
||||
"id": "img_correction",
|
||||
"valueType": "boolean",
|
||||
"key": "img_correction",
|
||||
"label": "img_correction",
|
||||
"type": "hidden"
|
||||
},
|
||||
{
|
||||
"id": "formula",
|
||||
"valueType": "boolean",
|
||||
"key": "formula",
|
||||
"label": "formula",
|
||||
"type": "hidden"
|
||||
}
|
||||
]
|
||||
},
|
||||
{
|
||||
"nodeId": "pluginOutput",
|
||||
"name": "自定义插件输出",
|
||||
"intro": "自定义配置外部输出，使用插件时，仅暴露自定义配置的输出",
|
||||
"avatar": "core/workflow/template/pluginOutput",
|
||||
"flowNodeType": "pluginOutput",
|
||||
"showStatus": false,
|
||||
"position": {
|
||||
"x": 1661.4708279314577,
|
||||
"y": 23.877720915480012
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "result",
|
||||
"label": "result",
|
||||
"description": "处理结果(或者是报错信息)",
|
||||
"value": ["zHG5jJBkXmjB", "xWQuEf50F3mr"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "failreason",
|
||||
"label": "failreason",
|
||||
"description": "文件处理失败原因，由文件名以及报错组成，多个文件之间由横线分隔开",
|
||||
"value": ["zHG5jJBkXmjB", "jbv4nVZvmFXm"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "success",
|
||||
"label": "success",
|
||||
"description": "是否全部文件都处理成功，如有没有处理成功的文件，失败原因将会输出在failreason中",
|
||||
"value": ["zHG5jJBkXmjB", "k46cjNulVk5Y"]
|
||||
}
|
||||
],
|
||||
"outputs": []
|
||||
},
|
||||
{
|
||||
"nodeId": "zHG5jJBkXmjB",
|
||||
"name": "HTTP 请求",
|
||||
"intro": "可以发出一个 HTTP 请求，实现更为复杂的操作（联网搜索、数据库查询等）",
|
||||
"avatar": "core/workflow/template/httpRequest",
|
||||
"flowNodeType": "httpRequest468",
|
||||
"showStatus": true,
|
||||
"position": {
|
||||
"x": 1081.967607938733,
|
||||
"y": -426.08028677656125
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"key": "system_addInputParam",
|
||||
"renderTypeList": ["addInputParam"],
|
||||
"valueType": "dynamic",
|
||||
"label": "",
|
||||
"required": false,
|
||||
"description": "core.module.input.description.HTTP Dynamic Input",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
}
|
||||
},
|
||||
{
|
||||
"key": "system_httpMethod",
|
||||
"renderTypeList": ["custom"],
|
||||
"valueType": "string",
|
||||
"label": "",
|
||||
"value": "POST",
|
||||
"required": true
|
||||
},
|
||||
{
|
||||
"key": "system_httpReqUrl",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "string",
|
||||
"label": "",
|
||||
"description": "core.module.input.description.Http Request Url",
|
||||
"placeholder": "https://api.ai.com/getInventory",
|
||||
"required": false,
|
||||
"value": "Doc2X/FileImg2text"
|
||||
},
|
||||
{
|
||||
"key": "system_httpHeader",
|
||||
"renderTypeList": ["custom"],
|
||||
"valueType": "any",
|
||||
"value": [],
|
||||
"label": "",
|
||||
"description": "core.module.input.description.Http Request Header",
|
||||
"placeholder": "core.module.input.description.Http Request Header",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"key": "system_httpParams",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": [],
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"key": "system_httpJsonBody",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"files\": {{files}},\n \"img_correction\": {{img_correction}},\n \"formula\": {{formula}}\n}",
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "apikey"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "arrayString",
|
||||
"canEdit": true,
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "url"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "img_correction",
|
||||
"label": "img_correction",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "img_correction"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "formula",
|
||||
"label": "formula",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "formula"]
|
||||
}
|
||||
],
|
||||
"outputs": [
|
||||
{
|
||||
"id": "error",
|
||||
"key": "error",
|
||||
"label": "请求错误",
|
||||
"description": "HTTP请求错误信息，成功时返回空",
|
||||
"valueType": "object",
|
||||
"type": "static"
|
||||
},
|
||||
{
|
||||
"id": "httpRawResponse",
|
||||
"key": "httpRawResponse",
|
||||
"label": "原始响应",
|
||||
"required": true,
|
||||
"description": "HTTP请求的原始响应。只能接受字符串或JSON类型响应数据。",
|
||||
"valueType": "any",
|
||||
"type": "static"
|
||||
},
|
||||
{
|
||||
"id": "system_addOutputParam",
|
||||
"key": "system_addOutputParam",
|
||||
"type": "dynamic",
|
||||
"valueType": "dynamic",
|
||||
"label": "",
|
||||
"customFieldConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": false
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "xWQuEf50F3mr",
|
||||
"valueType": "string",
|
||||
"type": "dynamic",
|
||||
"key": "result",
|
||||
"label": "result"
|
||||
},
|
||||
{
|
||||
"id": "jbv4nVZvmFXm",
|
||||
"valueType": "string",
|
||||
"type": "dynamic",
|
||||
"key": "failreason",
|
||||
"label": "failreason"
|
||||
},
|
||||
{
|
||||
"id": "k46cjNulVk5Y",
|
||||
"valueType": "boolean",
|
||||
"type": "dynamic",
|
||||
"key": "success",
|
||||
"label": "success"
|
||||
}
|
||||
]
|
||||
}
|
||||
],
|
||||
"edges": [
|
||||
{
|
||||
"source": "pluginInput",
|
||||
"target": "zHG5jJBkXmjB",
|
||||
"sourceHandle": "pluginInput-source-right",
|
||||
"targetHandle": "zHG5jJBkXmjB-target-left"
|
||||
},
|
||||
{
|
||||
"source": "zHG5jJBkXmjB",
|
||||
"target": "pluginOutput",
|
||||
"sourceHandle": "zHG5jJBkXmjB-source-right",
|
||||
"targetHandle": "pluginOutput-target-left"
|
||||
}
|
||||
]
|
||||
}
|
||||
}
|
||||
165
packages/plugins/src/Doc2X/FilePDF2text/index.ts
Normal file
165
packages/plugins/src/Doc2X/FilePDF2text/index.ts
Normal file
@@ -0,0 +1,165 @@
|
||||
import { delay } from '@fastgpt/global/common/system/utils';
|
||||
import { addLog } from '@fastgpt/service/common/system/log';
|
||||
import { result } from 'lodash';
|
||||
|
||||
type Props = {
|
||||
apikey: string;
|
||||
files: Array<string>;
|
||||
ocr: boolean;
|
||||
};
|
||||
|
||||
// Response type same as HTTP outputs
|
||||
type Response = Promise<{
|
||||
result: string;
|
||||
failreason: string;
|
||||
success: boolean;
|
||||
}>;
|
||||
|
||||
const main = async ({ apikey, files, ocr }: Props): Response => {
|
||||
// Check the apikey
|
||||
if (!apikey) {
|
||||
return {
|
||||
result: '',
|
||||
failreason: `API key is required`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
let real_api_key = apikey;
|
||||
if (!apikey.startsWith('sk-')) {
|
||||
const response = await fetch('https://api.doc2x.noedgeai.com/api/token/refresh', {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
Authorization: `Bearer ${apikey}`
|
||||
}
|
||||
});
|
||||
if (response.status !== 200) {
|
||||
return {
|
||||
result: '',
|
||||
failreason: `Get token failed: ${await response.text()}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
const data = await response.json();
|
||||
real_api_key = data.data.token;
|
||||
}
|
||||
|
||||
let final_result = '';
|
||||
let fail_reason = '';
|
||||
let flag = false;
|
||||
//Process each file one by one
|
||||
for await (const url of files) {
|
||||
//Fetch the pdf and check its contene type
|
||||
const PDFResponse = await fetch(url);
|
||||
if (!PDFResponse.ok) {
|
||||
fail_reason += `\n---\nFile:${url} \n<Content>\nFailed to fetch PDF from URL\n</Content>\n`;
|
||||
flag = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
const contentType = PDFResponse.headers.get('content-type');
|
||||
const file_name = url.match(/read\?filename=([^&]+)/)?.[1] || 'unknown.pdf';
|
||||
if (!contentType || !contentType.startsWith('application/pdf')) {
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nThe provided file does not point to a PDF: ${contentType}\n</Content>\n`;
|
||||
flag = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
const blob = await PDFResponse.blob();
|
||||
const formData = new FormData();
|
||||
formData.append('file', blob, file_name);
|
||||
formData.append('ocr', ocr ? '1' : '0');
|
||||
|
||||
let upload_url = 'https://api.doc2x.noedgeai.com/api/platform/async/pdf';
|
||||
if (real_api_key.startsWith('sk-')) {
|
||||
upload_url = 'https://api.doc2x.noedgeai.com/api/v1/async/pdf';
|
||||
}
|
||||
|
||||
let uuid;
|
||||
let upload_flag = true;
|
||||
const uploadAttempts = [1, 2, 3];
|
||||
for await (const attempt of uploadAttempts) {
|
||||
const upload_response = await fetch(upload_url, {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
Authorization: `Bearer ${real_api_key}`
|
||||
},
|
||||
body: formData
|
||||
});
|
||||
if (!upload_response.ok) {
|
||||
// Rate limit, wait for 10s and retry at most 3 times
|
||||
if (upload_response.status === 429 && attempt < 3) {
|
||||
await delay(10000);
|
||||
continue;
|
||||
}
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nFailed to upload file: ${await upload_response.text()}\n</Content>\n`;
|
||||
flag = true;
|
||||
upload_flag = false;
|
||||
}
|
||||
if (!upload_flag) {
|
||||
continue;
|
||||
}
|
||||
const upload_data = await upload_response.json();
|
||||
uuid = upload_data.data.uuid;
|
||||
break;
|
||||
}
|
||||
|
||||
// Get the result by uuid
|
||||
let result_url = 'https://api.doc2x.noedgeai.com/api/platform/async/status?uuid=' + uuid;
|
||||
if (real_api_key.startsWith('sk-')) {
|
||||
result_url = 'https://api.doc2x.noedgeai.com/api/v1/async/status?uuid=' + uuid;
|
||||
}
|
||||
|
||||
let required_flag = true;
|
||||
let result = '';
|
||||
// Wait for the result, at most 100s
|
||||
const maxAttempts = 100;
|
||||
for await (const _ of Array(maxAttempts).keys()) {
|
||||
const result_response = await fetch(result_url, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${real_api_key}`
|
||||
}
|
||||
});
|
||||
if (!result_response.ok) {
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nFailed to get result: ${await result_response.text()}\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
}
|
||||
const result_data = await result_response.json();
|
||||
if (['ready', 'processing'].includes(result_data.data.status)) {
|
||||
await delay(1000);
|
||||
} else if (result_data.data.status === 'pages limit exceeded') {
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nPages limit exceeded\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
} else if (result_data.data.status === 'success') {
|
||||
result = await Promise.all(
|
||||
result_data.data.result.pages.map((page: { md: any }) => page.md)
|
||||
).then((pages) => pages.join('\n'));
|
||||
result = result.replace(/\\[\(\)]/g, '$').replace(/\\[\[\]]/g, '$$');
|
||||
final_result += `\n---\nFile:${file_name}\n<Content>\n${result}\n</Content>\n`;
|
||||
required_flag = false;
|
||||
break;
|
||||
} else {
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nFailed to get result: ${result_data.data.status}\n</Content>\n`;
|
||||
flag = true;
|
||||
required_flag = false;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (required_flag) {
|
||||
fail_reason += `\n---\nFile:${file_name}\n<Content>\nTimeout after 100s for uuid ${uuid}\n</Content>\n`;
|
||||
flag = true;
|
||||
}
|
||||
}
|
||||
|
||||
return {
|
||||
result: final_result,
|
||||
failreason: fail_reason,
|
||||
success: !flag
|
||||
};
|
||||
};
|
||||
|
||||
export default main;
|
||||
389
packages/plugins/src/Doc2X/FilePDF2text/template.json
Normal file
389
packages/plugins/src/Doc2X/FilePDF2text/template.json
Normal file
@@ -0,0 +1,389 @@
|
||||
{
|
||||
"author": "Menghuan1918",
|
||||
"version": "488",
|
||||
"name": "Doc2X PDF文件(文件)识别",
|
||||
"avatar": "plugins/doc2x",
|
||||
"intro": "将上传的PDF文件发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"inputExplanationUrl": "https://fael3z0zfze.feishu.cn/wiki/Rkc5witXWiJoi5kORd2cofh6nDg?fromScene=spaceOverview",
|
||||
"showStatus": true,
|
||||
"weight": 10,
|
||||
|
||||
"isTool": true,
|
||||
"templateType": "tools",
|
||||
|
||||
"workflow": {
|
||||
"nodes": [
|
||||
{
|
||||
"nodeId": "pluginInput",
|
||||
"name": "自定义插件输入",
|
||||
"intro": "可以配置插件需要哪些输入，利用这些输入来运行插件",
|
||||
"avatar": "core/workflow/template/workflowStart",
|
||||
"flowNodeType": "pluginInput",
|
||||
"showStatus": false,
|
||||
"position": {
|
||||
"x": 388.243055058894,
|
||||
"y": -75.09744210499466
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"renderTypeList": ["input"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"description": "Doc2X的验证密匙，对于个人用户可以从Doc2X官网 - 个人信息 - 身份令牌获得",
|
||||
"required": true,
|
||||
"toolDescription": "",
|
||||
"defaultValue": ""
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "arrayString",
|
||||
"canEdit": true,
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"description": "待处理的PDF文件",
|
||||
"required": true,
|
||||
"toolDescription": "待处理的PDF文件"
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["switch"],
|
||||
"selectedTypeIndex": 0,
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "ocr",
|
||||
"label": "ocr",
|
||||
"description": "是否开启对PDF文件内图片的OCR识别，建议开启",
|
||||
"required": true,
|
||||
"toolDescription": "",
|
||||
"defaultValue": true
|
||||
}
|
||||
],
|
||||
"outputs": [
|
||||
{
|
||||
"id": "apikey",
|
||||
"valueType": "string",
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"type": "hidden"
|
||||
},
|
||||
{
|
||||
"id": "url",
|
||||
"valueType": "arrayString",
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"type": "hidden"
|
||||
},
|
||||
{
|
||||
"id": "formula",
|
||||
"valueType": "boolean",
|
||||
"key": "ocr",
|
||||
"label": "ocr",
|
||||
"type": "hidden"
|
||||
}
|
||||
]
|
||||
},
|
||||
{
|
||||
"nodeId": "pluginOutput",
|
||||
"name": "自定义插件输出",
|
||||
"intro": "自定义配置外部输出，使用插件时，仅暴露自定义配置的输出",
|
||||
"avatar": "core/workflow/template/pluginOutput",
|
||||
"flowNodeType": "pluginOutput",
|
||||
"showStatus": false,
|
||||
"position": {
|
||||
"x": 1649.7796447278438,
|
||||
"y": -96.05331527115042
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "result",
|
||||
"label": "result",
|
||||
"description": "处理结果，由文件名以及文档内容组成，多个文件之间由横线分隔开",
|
||||
"value": ["zHG5jJBkXmjB", "xWQuEf50F3mr"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "failreason",
|
||||
"label": "failreason",
|
||||
"description": "文件处理失败原因，由文件名以及报错组成，多个文件之间由横线分隔开",
|
||||
"value": ["zHG5jJBkXmjB", "yDxzW5CFalGw"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "success",
|
||||
"label": "success",
|
||||
"description": "是否全部文件都处理成功，如有没有处理成功的文件，失败原因将会输出在failreason中",
|
||||
"value": ["zHG5jJBkXmjB", "m6CJJj7GFud5"]
|
||||
}
|
||||
],
|
||||
"outputs": []
|
||||
},
|
||||
{
|
||||
"nodeId": "zHG5jJBkXmjB",
|
||||
"name": "HTTP 请求",
|
||||
"intro": "可以发出一个 HTTP 请求，实现更为复杂的操作（联网搜索、数据库查询等）",
|
||||
"avatar": "core/workflow/template/httpRequest",
|
||||
"flowNodeType": "httpRequest468",
|
||||
"showStatus": true,
|
||||
"position": {
|
||||
"x": 1077.7986740892777,
|
||||
"y": -496.9521622173004
|
||||
},
|
||||
"version": "481",
|
||||
"inputs": [
|
||||
{
|
||||
"key": "system_addInputParam",
|
||||
"renderTypeList": ["addInputParam"],
|
||||
"valueType": "dynamic",
|
||||
"label": "",
|
||||
"required": false,
|
||||
"description": "core.module.input.description.HTTP Dynamic Input",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
}
|
||||
},
|
||||
{
|
||||
"key": "system_httpMethod",
|
||||
"renderTypeList": ["custom"],
|
||||
"valueType": "string",
|
||||
"label": "",
|
||||
"value": "POST",
|
||||
"required": true
|
||||
},
|
||||
{
|
||||
"key": "system_httpReqUrl",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "string",
|
||||
"label": "",
|
||||
"description": "core.module.input.description.Http Request Url",
|
||||
"placeholder": "https://api.ai.com/getInventory",
|
||||
"required": false,
|
||||
"value": "Doc2X/FilePDF2text"
|
||||
},
|
||||
{
|
||||
"key": "system_httpHeader",
|
||||
"renderTypeList": ["custom"],
|
||||
"valueType": "any",
|
||||
"value": [],
|
||||
"label": "",
|
||||
"description": "core.module.input.description.Http Request Header",
|
||||
"placeholder": "core.module.input.description.Http Request Header",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"key": "system_httpParams",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": [],
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"key": "system_httpJsonBody",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"files\": {{files}},\n \"ocr\": {{ocr}}\n}",
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "string",
|
||||
"canEdit": true,
|
||||
"key": "apikey",
|
||||
"label": "apikey",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "apikey"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "arrayString",
|
||||
"canEdit": true,
|
||||
"key": "files",
|
||||
"label": "files",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "url"]
|
||||
},
|
||||
{
|
||||
"renderTypeList": ["reference"],
|
||||
"valueType": "boolean",
|
||||
"canEdit": true,
|
||||
"key": "ocr",
|
||||
"label": "ocr",
|
||||
"customInputConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": true
|
||||
},
|
||||
"required": true,
|
||||
"value": ["pluginInput", "formula"]
|
||||
}
|
||||
],
|
||||
"outputs": [
|
||||
{
|
||||
"id": "error",
|
||||
"key": "error",
|
||||
"label": "请求错误",
|
||||
"description": "HTTP请求错误信息，成功时返回空",
|
||||
"valueType": "object",
|
||||
"type": "static"
|
||||
},
|
||||
{
|
||||
"id": "httpRawResponse",
|
||||
"key": "httpRawResponse",
|
||||
"label": "原始响应",
|
||||
"required": true,
|
||||
"description": "HTTP请求的原始响应。只能接受字符串或JSON类型响应数据。",
|
||||
"valueType": "any",
|
||||
"type": "static"
|
||||
},
|
||||
{
|
||||
"id": "system_addOutputParam",
|
||||
"key": "system_addOutputParam",
|
||||
"type": "dynamic",
|
||||
"valueType": "dynamic",
|
||||
"label": "",
|
||||
"customFieldConfig": {
|
||||
"selectValueTypeList": [
|
||||
"string",
|
||||
"number",
|
||||
"boolean",
|
||||
"object",
|
||||
"arrayString",
|
||||
"arrayNumber",
|
||||
"arrayBoolean",
|
||||
"arrayObject",
|
||||
"any",
|
||||
"chatHistory",
|
||||
"datasetQuote",
|
||||
"dynamic",
|
||||
"selectApp",
|
||||
"selectDataset"
|
||||
],
|
||||
"showDescription": false,
|
||||
"showDefaultValue": false
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "xWQuEf50F3mr",
|
||||
"valueType": "string",
|
||||
"type": "dynamic",
|
||||
"key": "result",
|
||||
"label": "result"
|
||||
},
|
||||
{
|
||||
"id": "m6CJJj7GFud5",
|
||||
"valueType": "boolean",
|
||||
"type": "dynamic",
|
||||
"key": "success",
|
||||
"label": "success"
|
||||
},
|
||||
{
|
||||
"id": "yDxzW5CFalGw",
|
||||
"valueType": "string",
|
||||
"type": "dynamic",
|
||||
"key": "failreason",
|
||||
"label": "failreason"
|
||||
}
|
||||
]
|
||||
}
|
||||
],
|
||||
"edges": [
|
||||
{
|
||||
"source": "pluginInput",
|
||||
"target": "zHG5jJBkXmjB",
|
||||
"sourceHandle": "pluginInput-source-right",
|
||||
"targetHandle": "zHG5jJBkXmjB-target-left"
|
||||
},
|
||||
{
|
||||
"source": "zHG5jJBkXmjB",
|
||||
"target": "pluginOutput",
|
||||
"sourceHandle": "zHG5jJBkXmjB-source-right",
|
||||
"targetHandle": "pluginOutput-target-left"
|
||||
}
|
||||
]
|
||||
}
|
||||
}
|
||||
@@ -8,7 +8,6 @@ type Props = {
|
||||
formula: boolean;
|
||||
};
|
||||
|
||||
// Response type same as HTTP outputs
|
||||
type Response = Promise<{
|
||||
result: string;
|
||||
success: boolean;
|
||||
@@ -41,36 +40,36 @@ const main = async ({ apikey, url, img_correction, formula }: Props): Response =
|
||||
real_api_key = data.data.token;
|
||||
}
|
||||
|
||||
//Get the image binary from the URL
|
||||
const extension = url.split('.').pop()?.toLowerCase();
|
||||
const name = url.split('/').pop()?.split('.').shift();
|
||||
let mini = '';
|
||||
switch (extension) {
|
||||
case 'jpg':
|
||||
case 'jpeg':
|
||||
mini = 'image/jpeg';
|
||||
break;
|
||||
case 'png':
|
||||
mini = 'image/png';
|
||||
break;
|
||||
default:
|
||||
return {
|
||||
result: `Not supported image format, only support jpg/jpeg/png`,
|
||||
success: false
|
||||
};
|
||||
let imageResponse;
|
||||
// Fetch the image and check its content type
|
||||
try {
|
||||
imageResponse = await fetch(url);
|
||||
} catch (e) {
|
||||
return {
|
||||
result: `Failed to fetch image from URL: ${url} with error: ${e}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
const response = await fetch(url);
|
||||
if (!response.ok) {
|
||||
if (!imageResponse.ok) {
|
||||
return {
|
||||
result: `Failed to fetch image from URL: ${url}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
const blob = await response.blob();
|
||||
const contentType = imageResponse.headers.get('content-type');
|
||||
if (!contentType || !contentType.startsWith('image/')) {
|
||||
return {
|
||||
result: `The provided URL does not point to an image: ${contentType}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
const blob = await imageResponse.blob();
|
||||
const formData = new FormData();
|
||||
formData.append('file', new Blob([blob], { type: mini }), name + '.' + extension);
|
||||
const fileName = url.split('/').pop()?.split('?')[0] || 'image';
|
||||
formData.append('file', blob, fileName);
|
||||
formData.append('img_correction', img_correction ? '1' : '0');
|
||||
formData.append('equation', formula ? '1' : '0');
|
||||
|
||||
@@ -135,8 +134,17 @@ const main = async ({ apikey, url, img_correction, formula }: Props): Response =
|
||||
success: false
|
||||
};
|
||||
} else if (result_data.data.status === 'success') {
|
||||
let result = result_data.data.result.pages[0].md;
|
||||
result = result.replace(/\\[\(\)]/g, '$').replace(/\\[\[\]]/g, '$$');
|
||||
let result;
|
||||
try {
|
||||
result = result_data.data.result.pages[0].md;
|
||||
result = result.replace(/\\[\(\)]/g, '$').replace(/\\[\[\]]/g, '$$');
|
||||
} catch {
|
||||
// no pages
|
||||
return {
|
||||
result: '',
|
||||
success: true
|
||||
};
|
||||
}
|
||||
return {
|
||||
result: result,
|
||||
success: true
|
||||
|
||||
@@ -3,7 +3,7 @@
|
||||
"version": "488",
|
||||
"name": "Doc2X 图像(URL)识别",
|
||||
"avatar": "plugins/doc2x",
|
||||
"intro": "将传入的图片(URL)发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"intro": "从URL下载图片并发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"inputExplanationUrl": "https://fael3z0zfze.feishu.cn/wiki/Rkc5witXWiJoi5kORd2cofh6nDg?fromScene=spaceOverview",
|
||||
"showStatus": true,
|
||||
"weight": 10,
|
||||
@@ -220,7 +220,7 @@
|
||||
"key": "system_httpJsonBody",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"url\": \"{{url}}\",\n \"img_correction\": \"{{img_correction}}\",\n \"formula\": \"{{img_correction}}\"\n}",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"url\": \"{{url}}\",\n \"img_correction\": {{img_correction}},\n \"formula\": {{formula}}\n}",
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
|
||||
@@ -40,9 +40,35 @@ const main = async ({ apikey, url, ocr }: Props): Response => {
|
||||
real_api_key = data.data.token;
|
||||
}
|
||||
|
||||
//Get the image binary from the URL
|
||||
//Fetch the pdf and check its contene type
|
||||
let PDFResponse;
|
||||
try {
|
||||
PDFResponse = await fetch(url);
|
||||
} catch (e) {
|
||||
return {
|
||||
result: `Failed to fetch PDF from URL: ${url} with error: ${e}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
if (!PDFResponse.ok) {
|
||||
return {
|
||||
result: `Failed to fetch PDF from URL: ${url}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
const contentType = PDFResponse.headers.get('content-type');
|
||||
if (!contentType || !contentType.startsWith('application/pdf')) {
|
||||
return {
|
||||
result: `The provided URL does not point to a PDF: ${contentType}`,
|
||||
success: false
|
||||
};
|
||||
}
|
||||
|
||||
const blob = await PDFResponse.blob();
|
||||
const formData = new FormData();
|
||||
formData.append('pdf_url', url);
|
||||
const fileName = url.split('/').pop()?.split('?')[0] || 'pdf';
|
||||
formData.append('file', blob, fileName);
|
||||
formData.append('ocr', ocr ? '1' : '0');
|
||||
|
||||
let upload_url = 'https://api.doc2x.noedgeai.com/api/platform/async/pdf';
|
||||
|
||||
@@ -3,7 +3,7 @@
|
||||
"version": "488",
|
||||
"name": "Doc2X PDF文件(URL)识别",
|
||||
"avatar": "plugins/doc2x",
|
||||
"intro": "将传入的PDF文件(URL)发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"intro": "从URL下载PDF文件，并发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本",
|
||||
"inputExplanationUrl": "https://fael3z0zfze.feishu.cn/wiki/Rkc5witXWiJoi5kORd2cofh6nDg?fromScene=spaceOverview",
|
||||
"showStatus": true,
|
||||
"weight": 10,
|
||||
@@ -201,7 +201,7 @@
|
||||
"key": "system_httpJsonBody",
|
||||
"renderTypeList": ["hidden"],
|
||||
"valueType": "any",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"url\": \"{{url}}\",\n \"ocr\": \"{{ocr}}\"\n}",
|
||||
"value": "{\n \"apikey\": \"{{apikey}}\",\n \"url\": \"{{url}}\",\n \"ocr\": {{ocr}}\n}",
|
||||
"label": "",
|
||||
"required": false
|
||||
},
|
||||
|
||||
@@ -3,7 +3,7 @@
|
||||
"version": "488",
|
||||
"name": "Doc2X服务",
|
||||
"avatar": "plugins/doc2x",
|
||||
"intro": "传入的URL形式的图片或PDF文件发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本。",
|
||||
"intro": "将传入的图片或PDF文件发送至Doc2X进行解析，返回带LaTeX公式的markdown格式的文本。",
|
||||
"showStatus": true,
|
||||
"weight": 10,
|
||||
|
||||
|
||||
Reference in New Issue
Block a user