Most of my web browsing is reading text, so I wondered about stripping webpages down to plain text. I showed it to my boss at work and he suggested putting it up somewhere as a tool for people to use.
There is no caching so its a bit slow to load.
There are examples in the blog post, but try loading a new yorker article like this:
curl https://earthly-tools.com/text-mode?url=https://www.newyorker.com/news/letter-from-silicon-valley/the-lonely-work-of-moderating-hacker-news | less
Unfortunately, there are no links, so browsing around isn't great experience but reading is.
Next thing I need to add is caching, because some pages are just really slow and text should be very cacheable.
Also, on my phone I need to use widescreen mode to make it viewable the line breaks are in the content and not reactive to browser size.
I hear you; however, changing random broken html found on the web into a readable piece of plain text is not that easy. Think about old school layout done with tables, source code listing, link farms, etc. Dumb down the html is actually easier.
This is fantastic for reading things on my terminal. Thanks for making this. Is there a way to make the url shorter? That will make it much more usable.
But any shorter than that would require a url shortener, or a different domain name. Do any url shorteners support adding url params onto the end of them?
It seems a good idea but should consider also the ideas:
* Character set, if you want to convert to ASCII, or other character encodings that the document to be converted might use.
* Option for including links (possibly in footnotes, and/or section numbers for links to other parts of the same document).
* Option for to find any <link rel="alternate"> in case there is already the plain text version (but you can also disable it if you want to force converting the HTML instead). (This seems unnecessary for the remote service, but if it has it then it might be useful when using locally)
Is good that you have the source codes, though. You might want to run it locally, instead of having to use the remote service to make such a conversion.
Hi, I created this.
Most of my web browsing is reading text, so I wondered about stripping webpages down to plain text. I showed it to my boss at work and he suggested putting it up somewhere as a tool for people to use.
There is no caching so its a bit slow to load.
There are examples in the blog post, but try loading a new yorker article like this:
Or read it right in your terminal:
Unfortunately, there are no links, so browsing around isn't great experience but reading is.
Next thing I need to add is caching, because some pages are just really slow and text should be very cacheable.
Also, on my phone I need to use widescreen mode to make it viewable the line breaks are in the content and not reactive to browser size.
Feature requests?
The same newyorker article from my site: (free registration required) https://roastidio.us/topic/105469?fulltext=t
I strip away all the bells and whistles, keep the main text, fix the img and a links with absolute links, and apply minimum styling.
I also cached all the results and built a full text search index with it.
I was thinking about how if I served in html I could include links and such, but I do like the aesthetics of plain text.
It makes me feel like I'm reading a text file on some BBS or something.
But what you've built is very cool.
I hear you; however, changing random broken html found on the web into a readable piece of plain text is not that easy. Think about old school layout done with tables, source code listing, link farms, etc. Dumb down the html is actually easier.
This is fantastic for reading things on my terminal. Thanks for making this. Is there a way to make the url shorter? That will make it much more usable.
I think I could get this to work:
But any shorter than that would require a url shortener, or a different domain name. Do any url shorteners support adding url params onto the end of them?
It seems a good idea but should consider also the ideas:
* Character set, if you want to convert to ASCII, or other character encodings that the document to be converted might use.
* Option for including links (possibly in footnotes, and/or section numbers for links to other parts of the same document).
* Option for to find any <link rel="alternate"> in case there is already the plain text version (but you can also disable it if you want to force converting the HTML instead). (This seems unnecessary for the remote service, but if it has it then it might be useful when using locally)
Is good that you have the source codes, though. You might want to run it locally, instead of having to use the remote service to make such a conversion.