Extending the web scraper.

[ba11b0y]

There hasn't been much work on the web scraping part.
I am interested to work on this.
Since this is going to be a generic one, what I have thought as of now includes:

1. A generic web scraper which scrapes all images, links and the text.
2. Use scrapy for this maybe.

Still a beginner, any tips or corrections?

[shubhodeep9]

@invinciblycool I like the thought, I would suggest, a detailed list of missing components you find in the current code of scraper, then we will assign you the work.

[ashwini0529]

@invinciblycool XML format could be added.

[ba11b0y]

@ashwini0529 I have added the XML response to web.py. Let me know if any corrections are needed
@shubhodeep9 I will update the detailed list as soon as my exams get over 😄

[ba11b0y]

@ashwini0529 @shubhodeep9 Couldn't resist the excitement 😄
These are some features in my mind which can be added :

• If no JSON response is returned by the URL, only the source of the page is returned. We could have a more better scraper which returns either:

1. A dictionary or a JSON reponse:

{
  "assets":
  {
    "images":
    [
      "link of image1 on the page",
      "link of image2 on the page"
    ],
    "videos":
    [
      "link to embedded video1",
      "link to embedded video2"
    ]
  },
  "content":
  {
    "text": "all raw text from the page",
    "html": "all html from the page"
  }
}

2. Or creates dedicated directories for the above keys of the dictionaries and actually saves the content to the respective directory.(Inspired from httrack)

• Another feature could be adding a specific scrape option.
  For Example:
  web.scrape(url, scrape_content = "images") returns all the links to images in or saves the images locally.

[ashwini0529]

Hey @invinciblycool Sounds good.
Sounds like a great idea to start with. Go ahead. We can add more features. 🎉

[shubhodeep9]

@invinciblycool Add a TO-DO with your PR, and we will keep this issue alive until we feel satisfied. So that whenever someone gets a new idea on web-scraping, they can add to that TO-DO

[ashwini0529]

Also, please add a [WIP] tag in your PR message. 😄

[ba11b0y]

@ashwini0529 To start working if you could make it clear that should the function be returning a response or should create folders and save the content locally. Thanks.
@shubhodeep9 Just confirming a TO-DO with the PR or the issue.

[ashwini0529]

Hey @invinciblycool you can take a look at the QR Code function. I think you can make something like that.
Probable usage like what it was for QRCode:
img = hackr.image.qrcode("https://github.com/pytorn/hackr", dest_path="/tmp/hackr_qrcode.png")

[ba11b0y]

I guess then we agree on saving all the content locally.
Will start working on it ASAP.

[ashwini0529]

Hey @invinciblycool Updates?

[ba11b0y]

Sorry for the delay, I will try opening a PR by this week.
Happy Diwali BTW. ✨

[ashwini0529]

Perfect @invinciblycool
Happy hacking and Happy Diwali! 😄 🎇