News Extraction

工作流概述

这是一个包含36个节点的复杂工作流,主要用于自动化处理各种任务。

工作流源代码

下载
{
  "id": "xM8Z5vZVNTNjCySL",
  "meta": {
    "instanceId": "b8ef33547995f2a520f12118ac1f7819ea58faa7a1096148cac519fa08be8e99"
  },
  "name": "News Extraction",
  "tags": [],
  "nodes": [
    {
      "id": "97711d12-20de-40aa-b994-d2b10f20a5e5",
      "name": "Extract the HTML with the right css class",
      "type": "n8n-nodes-base.html",
      "position": [
        -500,
        0
      ],
      "parameters": {
        "options": {
          "trimValues": true
        },
        "operation": "extractHtmlContent",
        "extractionValues": {
          "values": [
            {
              "key": "data",
              "attribute": "href",
              "cssSelector": "=div:nth-child(9) > div:nth-child(3) > a:nth-child(2)",
              "returnArray": true,
              "returnValue": "attribute"
            }
          ]
        }
      },
      "typeVersion": 1
    },
    {
      "id": "b874b570-daae-4878-b525-07ac30756eb1",
      "name": "Summary",
      "type": "n8n-nodes-base.openAi",
      "position": [
        -880,
        440
      ],
      "parameters": {
        "model": "gpt-4-1106-preview",
        "prompt": {
          "messages": [
            {
              "content": "=Create a summary in less than 70 words {{ $json[\"content\"] }}"
            }
          ]
        },
        "options": {},
        "resource": "chat"
      },
      "credentials": {
        "openAiApi": {
          "id": "0Vdk5RlVe7AoUdAM",
          "name": "OpenAi account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "72696278-2d44-4073-936a-6fe9df1bc7d8",
      "name": "Keywords",
      "type": "n8n-nodes-base.openAi",
      "position": [
        -880,
        620
      ],
      "parameters": {
        "model": "gpt-4-1106-preview",
        "prompt": {
          "messages": [
            {
              "content": "=name the 3 most important technical keywords in {{ $json[\"content\"] }} ? just name them without any explanations or other sentences"
            }
          ]
        },
        "options": {},
        "resource": "chat"
      },
      "credentials": {
        "openAiApi": {
          "id": "0Vdk5RlVe7AoUdAM",
          "name": "OpenAi account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "0bfdb3be-76ef-4bb3-902f-f0869342b83c",
      "name": "Rename keywords",
      "type": "n8n-nodes-base.set",
      "position": [
        -700,
        620
      ],
      "parameters": {
        "fields": {
          "values": [
            {
              "name": "keywords",
              "stringValue": "={{ $json[\"message\"][\"content\"] }}"
            }
          ]
        },
        "include": "none",
        "options": {}
      },
      "typeVersion": 3.1
    },
    {
      "id": "0387cf34-41c9-4729-8570-1db7b17c42f4",
      "name": "Rename Summary",
      "type": "n8n-nodes-base.set",
      "position": [
        -700,
        440
      ],
      "parameters": {
        "fields": {
          "values": [
            {
              "name": "=summary",
              "stringValue": "={{ $json[\"message\"][\"content\"] }}"
            }
          ]
        },
        "include": "none",
        "options": {}
      },
      "typeVersion": 3.1
    },
    {
      "id": "5fa1702c-f0bf-4524-bc8f-6f550dd83f1e",
      "name": "Merge",
      "type": "n8n-nodes-base.merge",
      "position": [
        -480,
        560
      ],
      "parameters": {
        "mode": "combine",
        "options": {},
        "combinationMode": "mergeByPosition"
      },
      "typeVersion": 2.1
    },
    {
      "id": "25128a71-b0d5-49a4-adb8-c3fbe03c0a85",
      "name": "Extract date",
      "type": "n8n-nodes-base.html",
      "position": [
        -500,
        -160
      ],
      "parameters": {
        "options": {},
        "operation": "extractHtmlContent",
        "extractionValues": {
          "values": [
            {
              "key": "data",
              "cssSelector": "div:nth-child(9) > div:nth-child(2) > span:nth-child(1)",
              "returnArray": true
            }
          ]
        }
      },
      "typeVersion": 1
    },
    {
      "id": "138b3bd6-494a-49b9-b5b8-c9febcfef9fb",
      "name": "Select posts of last 7 days",
      "type": "n8n-nodes-base.code",
      "position": [
        120,
        0
      ],
      "parameters": {
        "jsCode": "const currentDate = new Date();
const sevenDaysAgo = new Date(currentDate.setDate(currentDate.getDate() - 70)); // Change the number of days going back to your liking (e.g. from -7 to -1) -> BUT sync with the cron job (first node)

const filteredItems = items.filter(item => {
    const postDate = new Date(item.json[\"Date\"]); // Assuming \"Date\" is the field name in the extracted html
    return postDate >= sevenDaysAgo;
});

return filteredItems;
"
      },
      "typeVersion": 2
    },
    {
      "id": "1ace953b-e298-4fc2-8970-327f736889ec",
      "name": "Merge date & links",
      "type": "n8n-nodes-base.merge",
      "position": [
        -100,
        0
      ],
      "parameters": {
        "mode": "combine",
        "options": {},
        "combinationMode": "mergeByPosition"
      },
      "typeVersion": 2.1
    },
    {
      "id": "bba692fc-c225-41be-a969-179d8b99c071",
      "name": "HTTP Request1",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        320,
        0
      ],
      "parameters": {
        "url": "={{ $json[\"Link\"] }}",
        "options": {}
      },
      "typeVersion": 4.1
    },
    {
      "id": "26671065-631f-4684-9ee1-15f26b4cf1e4",
      "name": "Merge Content with Date & Link",
      "type": "n8n-nodes-base.merge",
      "position": [
        500,
        260
      ],
      "parameters": {
        "mode": "combine",
        "options": {},
        "combinationMode": "mergeByPosition"
      },
      "typeVersion": 2.1
    },
    {
      "id": "79beb744-97b8-4072-824a-6736b0a080ef",
      "name": "Extract individual posts",
      "type": "n8n-nodes-base.html",
      "position": [
        500,
        0
      ],
      "parameters": {
        "options": {},
        "operation": "extractHtmlContent",
        "extractionValues": {
          "values": [
            {
              "key": "title",
              "cssSelector": "h1.fl-heading > span:nth-child(1)"
            },
            {
              "key": "content",
              "cssSelector": ".fl-node-5c7574ae7d5c6 > div:nth-child(1)"
            }
          ]
        }
      },
      "typeVersion": 1
    },
    {
      "id": "e89d9de5-875b-453e-825a-26f2bebcc8df",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        80,
        -107
      ],
      "parameters": {
        "width": 180.9747474601832,
        "height": 276.31054308676767,
        "content": "Select only the newest news: todays date going back xy days"
      },
      "typeVersion": 1
    },
    {
      "id": "8a603f2f-4208-48c7-b169-e5613f13fa7d",
      "name": "Merge ChatGPT output with Date & Link",
      "type": "n8n-nodes-base.merge",
      "position": [
        -180,
        560
      ],
      "parameters": {
        "mode": "combine",
        "options": {},
        "combinationMode": "mergeByPosition"
      },
      "typeVersion": 2.1
    },
    {
      "id": "e1036421-9ce1-4121-a692-602410ec7c95",
      "name": "Sticky Note1",
      "type": "n8n-nodes-base.stickyNote",
      "disabled": true,
      "position": [
        -539.7802584556148,
        -4.722020203185366
      ],
      "parameters": {
        "width": 182.2748213508401,
        "height": 304.2550759710132,
        "content": "















Extracting the individual links of the press release page in order to retrieve the individual posts on their respective **url**"
      },
      "typeVersion": 1
    },
    {
      "id": "3655ab22-6a17-429a-9d9b-d96bbcc78fee",
      "name": "Sticky Note2",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -538.404803912782,
        -304
      ],
      "parameters": {
        "width": 178.75185894039254,
        "height": 289.463147786618,
        "content": "Extracting the dates of the posts of the press release page.
The right CSS selector has to be chosen.
[More info on datagrab.io](https://datagrab.io/blog/guide-to-css-selectors-for-web-scraping/)"
      },
      "typeVersion": 1
    },
    {
      "id": "2e27fb4c-426a-41e1-b5fb-9b2d78acd2a7",
      "name": "Sticky Note3",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -1300,
        -299.82161760751774
      ],
      "parameters": {
        "width": 334.4404040637068,
        "height": 1127.2017245821128,
        "content": "# Scraping posts of a news site without RSS feed


The [News Site](https://www.colt.net/resources/type/news/) from Colt, a telecom company, does not offer an RSS feed, therefore web scraping is the 
choice to extract and process the news.

The goal is to get only the newest posts, a summary of each post and their respective (technical) keywords.

Note that the news site offers the links to each news post, but not the individual news. We collect first the links and dates of each post before extracting the newest ones.

The result is sent to a SQL database, in this case a NocoDB database.

This process happens each week thru a cron job.

**Requirements**:
- Basic understanding of CSS selectors and how to get them via browser (usually: right click → inspect)
- ChatGPT API account - normal account is not sufficient
- A NocoDB database - of course you may choose any type of output target

**Assumptions**:
- CSS selectors work on the news site
- The post has a date with own CSS selector - meaning date is not part of the news content

**\"Warnings\"**
- Not every site likes to be scraped, especially not in high frequency
- Each website is structured in different ways, the workflow may then need several adaptations.


Happy about any suggestion to improve. You may contact me on **Mastodon**: https://bonn.social/@askans"
      },
      "typeVersion": 1
    },
    {
      "id": "d43bd5b7-2aff-4a07-8aca-ca4747ec6c4d",
      "name": "Sticky Note4",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -927.8447474890202,
        -80
      ],
      "parameters": {
        "width": 153.90180146729315,
        "height": 237.91333335255808,
        "content": "Weekly cron job"
      },
      "typeVersion": 1
    },
    {
      "id": "e732d136-fcf1-4fc3-8bb6-bdcea3c78d9e",
      "name": "Sticky Note5",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -760,
        -80
      ],
      "parameters": {
        "width": 185.41515152389002,
        "height": 241.454848504947,
        "content": "The html of the news site is being retrieved: https://www.colt.net/resources/type/news/"
      },
      "typeVersion": 1
    },
    {
      "id": "d5e29ec3-5ef2-42f3-b316-9350644dbba4",
      "name": "Sticky Note6",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -340,
        -306
      ],
      "parameters": {
        "width": 187.3613302133812,
        "height": 469.2923233086395,
        "content": "As the extraction are returned as arrays, they transformed into individual JSON items to enable looping with other nodes"
      },
      "typeVersion": 1
    },
    {
      "id": "1af15c45-32c0-4abf-a35d-be7206823569",
      "name": "Sticky Note7",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -120,
        -103.54151515238902
      ],
      "parameters": {
        "width": 150,
        "height": 274.50898992724416,
        "content": "The links of the individual posts and the dates of the posts "
      },
      "typeVersion": 1
    },
    {
      "id": "f7c42748-f227-42d0-a9e2-fcb16dbd0f75",
      "name": "Retrieve the web page for further processsing",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        -720,
        0
      ],
      "parameters": {
        "url": "https://www.colt.net/resources/type/news/",
        "options": {
          "response": {
            "response": {
              "responseFormat": "text"
            }
          }
        }
      },
      "typeVersion": 4.1
    },
    {
      "id": "b2c36f26-8221-478f-a4b0-22758b1e5e58",
      "name": "Sticky Note9",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        292,
        -100
      ],
      "parameters": {
        "width": 155.0036363426638,
        "height": 272.1479798256519,
        "content": "Get the html of each individual **newest** post"
      },
      "typeVersion": 1
    },
    {
      "id": "6ae05c31-c09a-4b4e-a013-41571937bc39",
      "name": "Sticky Note10",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        460,
        -100
      ],
      "parameters": {
        "width": 184.07417896879767,
        "height": 269.2504410842093,
        "content": "Extracting the title & content (text) of each individual news post with the right CSS selector"
      },
      "typeVersion": 1
    },
    {
      "id": "e2da76d4-0c8c-4c61-924f-50aa9387e9ab",
      "name": "Sticky Note11",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        460,
        180
      ],
      "parameters": {
        "width": 191.87778190338406,
        "height": 234.13422787857044,
        "content": "Merge link to url, date with content (text) and title of each news psot"
      },
      "typeVersion": 1
    },
    {
      "id": "c124aaac-dce6-4658-9027-bdfe5c0c81e6",
      "name": "Sticky Note12",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -907.2264215202996,
        331.0681740778203
      ],
      "parameters": {
        "width": 150,
        "height": 256.2444361932317,
        "content": "Create a summary of each news post with ChatGPT. You need a ChatGPT API account for this"
      },
      "typeVersion": 1
    },
    {
      "id": "c9037e74-007b-4e44-b7f9-90e78b853eb5",
      "name": "Sticky Note13",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -909.595196087218,
        610.7495589157902
      ],
      "parameters": {
        "width": 152.85976723045226,
        "height": 218.52702200939785,
        "content": "













Get the 3 keywords of each news post"
      },
      "typeVersion": 1
    },
    {
      "id": "756397d9-de80-4114-9dee-b4f4b9593333",
      "name": "Sticky Note14",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -740,
        340
      ],
      "parameters": {
        "width": 182.7735784797001,
        "height": 489.05192374172555,
        "content": "Just a renaming of data fields and eliminating unnecessary ones"
      },
      "typeVersion": 1
    },
    {
      "id": "a0dcb254-f064-45ed-8e22-30a6d079085b",
      "name": "Sticky Note15",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -520,
        480
      ],
      "parameters": {
        "width": 169.7675735887227,
        "height": 254.94383570413422,
        "content": "Merge summary and keywords of each news post"
      },
      "typeVersion": 1
    },
    {
      "id": "82993166-b273-4b82-a954-554c6892f825",
      "name": "Schedule Trigger each week",
      "type": "n8n-nodes-base.scheduleTrigger",
      "position": [
        -900,
        0
      ],
      "parameters": {
        "rule": {
          "interval": [
            {
              "field": "weeks",
              "triggerAtDay": [
                3
              ],
              "triggerAtHour": 4,
              "triggerAtMinute": 32
            }
          ]
        }
      },
      "typeVersion": 1.1
    },
    {
      "id": "3d670eb9-5a36-4cd9-8d2c-40adf848485e",
      "name": "Sticky Note16",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -220,
        477.5081090810816
      ],
      "parameters": {
        "width": 180.1723775015045,
        "height": 260.5279202647822,
        "content": "Add title, link and date to summary and keywords of each news post"
      },
      "typeVersion": 1
    },
    {
      "id": "62021393-e988-4834-9fa2-75a957b42890",
      "name": "NocoDB news database",
      "type": "n8n-nodes-base.nocoDb",
      "position": [
        60,
        560
      ],
      "parameters": {
        "table": "mhbalmu9aaqcun6",
        "fieldsUi": {
          "fieldValues": [
            {
              "fieldName": "=News_Source",
              "fieldValue": "=Colt"
            },
            {
              "fieldName": "Title",
              "fieldValue": "={{ $json[\"title\"] }}"
            },
            {
              "fieldName": "Date",
              "fieldValue": "={{ $json[\"Date\"] }}"
            },
            {
              "fieldName": "Link",
              "fieldValue": "={{ $json[\"Link\"] }}"
            },
            {
              "fieldName": "Summary",
              "fieldValue": "={{ $json[\"summary\"] }}"
            },
            {
              "fieldName": "Keywords",
              "fieldValue": "={{ $json[\"keywords\"] }}"
            }
          ]
        },
        "operation": "create",
        "projectId": "prqu4e8bjj4bv1j",
        "authentication": "nocoDbApiToken"
      },
      "credentials": {
        "nocoDbApiToken": {
          "id": "gjNns0VJMS3P2RQ3",
          "name": "NocoDB Token account"
        }
      },
      "typeVersion": 2
    },
    {
      "id": "e59e9fab-10a7-470b-afa6-e1d4b4e57723",
      "name": "Sticky Note17",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        280,
        480
      ],
      "parameters": {
        "width": 483.95825869942666,
        "height": 268.5678114630957,
        "content": "## News summaries and keywords → database

[NocoDB](https://nocodb.com/) is an SQL database, here we store the news summaries and keywords for further processing. Any other output target can be chosen here, e.g. e-mail, Excel etc.

You need first have that database structured before appending the news summaries and additional fields. The you can shape this node.

Some fields may be edited in the database itself (e.g. relevance of the news to you) and may be filled therefore with a default value or not at all"
      },
      "typeVersion": 1
    },
    {
      "id": "253b414b-9a5b-4a25-892b-9aa011d55d28",
      "name": "Sticky Note18",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        20,
        480
      ],
      "parameters": {
        "width": 262.99083066277313,
        "height": 268.56781146309544,
        "content": ""
      },
      "typeVersion": 1
    },
    {
      "id": "438e8dde-ce0a-4e5e-8d62-d735d19ec189",
      "name": "Create single link items",
      "type": "n8n-nodes-base.itemLists",
      "position": [
        -300,
        0
      ],
      "parameters": {
        "options": {
          "destinationFieldName": "Link"
        },
        "fieldToSplitOut": "data"
      },
      "typeVersion": 3
    },
    {
      "id": "d721776b-fefc-4e72-91ef-6710f10b0393",
      "name": "Create single date items",
      "type": "n8n-nodes-base.itemLists",
      "position": [
        -300,
        -160
      ],
      "parameters": {
        "options": {
          "destinationFieldName": "Date"
        },
        "fieldToSplitOut": "data"
      },
      "typeVersion": 3
    }
  ],
  "active": false,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "ff89d802-3bcf-4b34-9cd9-776b1f3b5eab",
  "connections": {
    "Merge": {
      "main": [
        [
          {
            "node": "Merge ChatGPT output with Date & Link",
            "type": "main",
            "index": 1
          }
        ]
      ]
    },
    "Summary": {
      "main": [
        [
          {
            "node": "Rename Summary",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Keywords": {
      "main": [
        [
          {
            "node": "Rename keywords",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract date": {
      "main": [
        [
          {
            "node": "Create single date items",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "HTTP Request1": {
      "main": [
        [
          {
            "node": "Extract individual posts",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Rename Summary": {
      "main": [
        [
          {
            "node": "Merge",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Rename keywords": {
      "main": [
        [
          {
            "node": "Merge",
            "type": "main",
            "index": 1
          }
        ]
      ]
    },
    "Merge date & links": {
      "main": [
        [
          {
            "node": "Select posts of last 7 days",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Create single date items": {
      "main": [
        [
          {
            "node": "Merge date & links",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Create single link items": {
      "main": [
        [
          {
            "node": "Merge date & links",
            "type": "main",
            "index": 1
          }
        ]
      ]
    },
    "Extract individual posts": {
      "main": [
        [
          {
            "node": "Merge Content with Date & Link",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Schedule Trigger each week": {
      "main": [
        [
          {
            "node": "Retrieve the web page for further processsing",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Select posts of last 7 days": {
      "main": [
        [
          {
            "node": "Merge Content with Date & Link",
            "type": "main",
            "index": 1
          },
          {
            "node": "HTTP Request1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Merge Content with Date & Link": {
      "main": [
        [
          {
            "node": "Summary",
            "type": "main",
            "index": 0
          },
          {
            "node": "Keywords",
            "type": "main",
            "index": 0
          },
          {
            "node": "Merge ChatGPT output with Date & Link",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Merge ChatGPT output with Date & Link": {
      "main": [
        [
          {
            "node": "NocoDB news database",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract the HTML with the right css class": {
      "main": [
        [
          {
            "node": "Create single link items",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Retrieve the web page for further processsing": {
      "main": [
        [
          {
            "node": "Extract the HTML with the right css class",
            "type": "main",
            "index": 0
          },
          {
            "node": "Extract date",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}

功能特点

  • 自动检测新邮件
  • AI智能内容分析
  • 自定义分类规则
  • 批量处理能力
  • 详细的处理日志

技术分析

节点类型及作用

  • Html
  • Openai
  • Set
  • Merge
  • Code

复杂度评估

配置难度:
★★★★☆
维护难度:
★★☆☆☆
扩展性:
★★★★☆

实施指南

前置条件

  • 有效的Gmail账户
  • n8n平台访问权限
  • Google API凭证
  • AI分类服务订阅

配置步骤

  1. 在n8n中导入工作流JSON文件
  2. 配置Gmail节点的认证信息
  3. 设置AI分类器的API密钥
  4. 自定义分类规则和标签映射
  5. 测试工作流执行
  6. 配置定时触发器(可选)

关键参数

参数名称 默认值 说明
maxEmails 50 单次处理的最大邮件数量
confidenceThreshold 0.8 分类置信度阈值
autoLabel true 是否自动添加标签

最佳实践

优化建议

  • 定期更新AI分类模型以提高准确性
  • 根据邮件量调整处理批次大小
  • 设置合理的分类置信度阈值
  • 定期清理过期的分类规则

安全注意事项

  • 妥善保管API密钥和认证信息
  • 限制工作流的访问权限
  • 定期审查处理日志
  • 启用双因素认证保护Gmail账户

性能优化

  • 使用增量处理减少重复工作
  • 缓存频繁访问的数据
  • 并行处理多个邮件分类任务
  • 监控系统资源使用情况

故障排除

常见问题

邮件未被正确分类

检查AI分类器的置信度阈值设置,适当降低阈值或更新训练数据。

Gmail认证失败

确认Google API凭证有效且具有正确的权限范围,重新进行OAuth授权。

调试技巧

  • 启用详细日志记录查看每个步骤的执行情况
  • 使用测试邮件验证分类逻辑
  • 检查网络连接和API服务状态
  • 逐步执行工作流定位问题节点

错误处理

工作流包含以下错误处理机制:

  • 网络超时自动重试(最多3次)
  • API错误记录和告警
  • 处理失败邮件的隔离机制
  • 异常情况下的回滚操作