deathau / markdownload Goto Github PK

Add settings for how HTML is converted into Markdown.

https://github.com/0x6b/copy-selection-as-markdown Has some good options for this, so probably just shamelessly copy them.

Reported by gnatan on the Obsidian Discord

Browser: Opera
System: Mac Os
Extension: 2.4

This is my front-matter template:

---

__Source:__ {baseURI}
__Author:__ {byline} 
__Language:__ English
__Type:__ Web

---

#eng

---

[[Index]] < 

---

This is what appears in downloaded md file:

---


__Source:__ Link
__Author:__  
__Language:__ English
__Type:__ Web


---


#eng



---



[[Index]] < 


---

The line-break is doubled for some reason.

Noticed this in an Obisidian Discord post by jarodise#8379
On the website https://www.perell.com/blog/the-ultimate-guide-to-writing-online, there is a faulty <base> tag. And as my code will now respect an existing tag and add one if not present, the base URL is all wrong if there's a tag in there with nothing.

I'll have to check if the base tag exists and has a valid URL.

As reported by @montemartin in issue #22
Need to get the path we're downloading the md file too (via te title template) and prepend that path to the image prefix so that images download relative to the md file.

When used in conjunction with another extension that affects downloads (e.g. Save In...) then files get downloaded with a Guid filename instead of the title as the filename.

This could be to do with the download being a BLOB and not having a proper URL from which to extract a proper file name.
Have to see if I can figure out a way around this.

I'm experiencing some problems with new lines in lists.
For some reason obsidian doesn't recognise the new lines made by Markdownload.
Example:

- meta:
    - topics: [[to_categorise]]
    - url: https://www.derbund.ch/kanton-bern-hat-sich-bei-neuem-polizeizentrum-verrechnet-453109101204
    - date: [[2020-11-14]]
    
- meta:
    - topics: [[to_categorise]]
    - url: 
    - date: [[2020-11-14]]
------

As you can see the raw text looks identical but the one generated by Markdownload doesn't get processed correctly.
I tried to copy the working code into Markdownload but the output still doesn't work.
Interestingly the line endings seem to be the same but the tabs are not, in VSCode they look different even though they feel both like they are 4 spaces:

Maybe it's a special type of space?
https://www.fileformat.info/info/unicode/char/search.htm?q=%C2%A0&han=Y&preview=entity

I very often want to snip something for my personal notes. Would be amazing to have a keyboard shortcut to activate the dialog, as well as being able to do so via right clicking!

Thanks for all your work!

Os: Mac OS
Browser: Opera

1. Clicking on extension button
2. See the correct markdown in app windows
3. Press DOWNLOAD
4. Nothing is downloading no dialogs pop-up

I tried to enable / disable image download no change in behavior.

No idea if there's even demand for this, as I haven't heard anything.

At first blush, it seems a bit more work to convert for Safari than between chromium/firefox
There is some guidance from Apple here: https://developer.apple.com/library/archive/documentation/UserExperience/Conceptual/SafariExtensionsConversionGuide/Chapters/Chrome.html

Sometimes I just need to collect a webpage link, and I want to get the browser tab as a markdown link with the title. Could that be added as an option to the tab context menu?

Also, I would want to use the tab title in the markdown link from <meta name="title" content="...">. It seems that what is being used as the article title is <meta property="og:title" content="...">. They are not always the same, and I prefer the tab title because it usually includes the website name. Perhaps there could be a new text replacement value for this.

To increase the usability and speed of bookmarking, custom download folder where all MD downloads will be stored is extremely useful. By default it could be something like MarkDownload or MDClips stored in the default browser's Downloads folder.

For example, Obsidian.md uses folders as vaults. This is useful as could be directly integrated in Obsidian.md or transferred without having to search and select for MD files and their resource folders each time.

Firefox 83.0, Win10

I found 2 pages with basically same content (like SiteB is someone blogged to save content on SiteA as a learning note)
The interesting part is the download will silently fail in both pages, no md, no images.
The url, layout... etc are just different so I doubt it's the selected content (very similar) causing the failure.
(I've checked the preview, the preview is working fine.)

Site A
From

To

Site B
From

To

Inspired by good old OneNote, whenever you copy markdown from the clipper, it would be great if it appended a link to the source to the copied text.

This would have to be an option, so you could disable it if you didn't want it.

Hi, thank you for making this wonderful extension! I really love it.

Since I'm using Obsidian markdown, it'll be even easier to extract the contents of web directly to my vault if I could choose where to download the files.

Thanks

For using in Front Matter, I wish, I could have an extract of keywords. So that <meta name="keywords" content="KeePass, Password, Safe, Manager, Database, Secure, Security, Open, Source, Code, Free, Encryption, Key, Master" /> would become KeePass, Password, Safe, Manager, Database, Secure, Security, Open, Source, Code, Free, Encryption, Key, Master". Since personally, I separate Keywords only by space, it would be nice to have an option to remove additionally the comma. I know, this isn't ideal for all, but it would be nice for my workflow.

I use the extension on Linux Mint Ulyana in both Brave (standard browser) and Firefox. It runs well on Firefox but has stopped working on Brave. When I click on the icon, the rectangular window opens and the 'busy' signal keeps on forever. No text appears in the window - nothing. By clicking outside the rectangular window it closes but no download.
It did work and I have tried most options switching on/off but no dice and I'd appreciate some guidance. TIA

It would be nice to have an extra option to strip all images when saving. I'm using the extension to save webpages for my ereader and I don't need any images (or captions).

Additional backslash characters are added to underscores in the markdown output . Each _ is replaced with \_.

Example:
Source:
tp_fan /proc/acpi/ibm/fan
hwmon /sys/class/thermal/thermal_zone0/temp

Markdownload markdown result:
tp_fan /proc/acpi/ibm/fan
hwmon /sys/class/thermal/thermal_zone0/temp

Browser: Firefox Linux 85.0

Feature request

1. Add new button, quick save or something save explanatory
2. Add option to define default folder for saving
3. After clicking on "Quick save button" it will save the file without any prompt into location set in settings

My user case: Im using Obsidian, and all my new articles are going first into the 🌱SEEDBOX folder. Later I'm going into folder and add appropriate tags and link the articles into MOC. With this option it will help to save time to move file from downloads... and in same time it will introduce the very useful concept of ``🌱SEEDBOX``` folder to bigger audience.

Feature request

1. Add option to define if you want to download images offline

2. Add option to define default folder for saving images e.g. (Obsidian/Resources/Images/Clipped/)

3. Images should have prefix of the articles:

   Title: Food is about taste
   Image name: FoodIsAbo_ + image file name

This will help to group images from one article, if article contains a lot of them. It it will also help in general to manage images resources.

4. When it is on, the markdown online image references are replaced with future local references
5. All link to images are downloaded to folder set in settings

You already know the links to relevant images in articles. I see that it is fairly easy to download the links as images, and save them in appropriate folder. It is very beneficial to have the offline version of images and it make sense - as you already have the text offline.

I read epub books on mobile via google play books similar to this workflog.

Play book can store notes and annotation in google drive i.e. (russian)

At this moment extension takes only headings of books and not notes/annotation. (you can try to clip google drive link above).

Would be great to:

• Add h1 # headings to the book headings
• Clip notes/annotations

Thanks

I've extend the list of forbidden characters in file names a bit.
IIRC Obsidian has problems with some of them and some I just consider to be weird to have in a file name :D.
Maybe they should be excluded by default?

[]#^«»,.'`’{}()!$

I want to select text on a page, then hit the button and clip only that (properly markdown formatted, of course).

A request from icebear on the Obsidian Forum: An option to batch download all current tabs in an open window, so you can open a bunch of articles and download the lot at once - without all the clicking involved in doing that now.

Firefox 83.0, Win 10

Also, when using "Always show Save As dialog", images are not saved at path relative to .md

Problem ### solved -> old chromeversion

win 8.1
Chromium Version 83.0.4103.61

same issue in inkognito-mode, Firefox worksa

Pop up testing:

`Uncaught` (in promise) Error: title.replaceAll is not a function
    at n (browser-polyfill.min.js:1)
DevTools failed to load SourceMap: Could not load content for chrome-extension://pcmpcfapbekmbjjkdalcgopdkipoggdi/browser-polyfill.min.js.map: HTTP error: status code 404, net::ERR_UNKNOWN_URL_SCHEME

I wanted to install this, but am not able to from firefox extensions marketplace.

Safari doesn't support the browser.download api. For downloading a markdown file, I've got a workaround of generating a link and clicking it via javascript. Unfortunately, this workaround only works because I'm generating the file and does not work for downloading images.

The only option I see is building a "native" module that downloads the images, which first requires learning how to do that.

It would be helpful to have more distinct visual styling of the current focus inside the dialog window. I like to tab into the note, edit it, and then tab out > enter to download without using the mouse. Right now when the dialog opens there is no indication of what is currently in focus. It might also be helpful to have it start focused on the download button. This would help with the "quick save" part of #21. Once opened the note could be downloaded quickly by pressing enter without any additional tabbing.

Additionally, because the dialog stays open after a download there is no way to tell when the file was sucessfully downloaded. Thankfully on Mac the download icon does bounce once to show a new download, but I still end with several copies sometimes. It would be nice if there were a visual change to indicate the file was downloaded.

Current workflow:

1. Go to a page.
2. Click on the extension.
3. Click on "download".
4. Move from your downloads folder to your preferred location (way too many clicks).

Proposed workflow:

1. Have the configurable download location (example: ~/notes/articles)
2. Find an article.
3. Click on the extension.
4. Click on "download".
5. The file gets automatically put in the right location.

To be honest I have no idea if it's possible to do this with the add-ons, but it would definitely save seconds each time I want to trigger the add-on.

I've found that images won't download from medium.com articles. Can someone try to reproduce? The download functionality is working on other sites.

It would be useful for some people to add a ZK id to the title (or replace it).

Could you please give option to chose what kind of styles for linked images?

Default: ![image](image)
Obsidian format: ![[image]]

• there is no folder path necessary
• %20 this is also not necessary just put there normal space

I often just want to copy text to the clipboard, rather than write to a file. After dragging the mouse over some text to select it in the popup window, it closes.

There are three conditions to reproducing this bug:

1. Copy-selection must be used;
2. A heading (any level: h2/h3/h4, etc.) must be present
3. Text before that heading must be part of the selection.

I tried copy-selection-as-markdown with the following text selected:

<p><strong>Default rule</strong> handles nodes which are not recognised by any other rule. By default, it outputs the node's text content (separated  by blank lines if it is a block-level element). Its behaviour can be customised with the <code>defaultReplacement</code> option.</p>
<h3><a id="user-content-rule-precedence" class="anchor" aria-hidden="true" href="#rule-precedence"></a>Rule Precedence</h3>
<p>Turndown iterates over the set of rules, and picks the first one that matches the <code>filter</code>. The following list describes the order of precedence:</p>

This fails, with the following output:

**Default rule** handles nodes which are not recognised by any other rule. By default, it outputs the node's text content (separated by blank lines if it is a block-level element). Its behaviour can be customised with the `defaultReplacement` option.

### []

[]: #rule-precedence

However, if I omit the line before the heading, and select:

<h3><a id="user-content-rule-precedence" class="anchor" aria-hidden="true" href="#rule-precedence"></a>Rule Precedence</h3>
<p>Turndown iterates over the set of rules, and picks the first one that matches the <code>filter</code>. The following list describes the order of precedence:</p>

I get the correct text, which is:

### []Rule Precedence

Turndown iterates over the set of rules, and picks the first one that matches the `filter`. The following list describes the order of precedence:

[]: #rule-precedence

Please let me know if I can help pin down the bug any further.

Chrome: 87.0.4280.88 (64 bits)
MarkDownload: 2.4.1
OS: Windows 10 (10.0.19042)
Issue Ref: #10

This issue seems to be related to Chrome download API as it was occurred( Issue 1008884: chrome download API downloads file with incorrect filename ) I'm not sure which party should solves this issue first. The same issue appears in different Chrome plugins (such as Save Images Issue ref ). If this problem cannot be solved temporarily, I hope someone can provide some temporary solutions, thank you!

https://www.ershicimi.com/p/044d2f679d3289d3f615b4acf64db379

the image filename in the page above cann't be saved correctly.

If specify a folder to download .md and all relatively images, the .md file can be downloaded at right place, while images will be downloaded in the default DOWNLOAD folder (im using Big Sur with chrome @ latest version), please check, thank you!

For example, clipping https://notes.andymatuschak.org/About_these_notes there are many relative links (for example/z3SjnvsB5aR2ddsycyXofbYR7fCxo7RmKW2be).
When converting them to markdown, they come out with the extension as the domain (e.g. moz-extension://112865e2-6d18-4214-8783-df5c914cce79/z3SjnvsB5aR2ddsycyXofbYR7fCxo7RmKW2be)

As mentioned on the Obsidian Forum

It's already stripping out explicitly disallowed characters in filenames (\/:*?"<>| and \u00A0) but for saving files for use in Obsidian, [] should also be disallowed (due to linking syntax).
They are valid file name characters though 🤔
Perhaps an option to add additional disallowed characters? with [] set by default?

Requested by Mzantsi on the Obsidian discord https://discord.com/channels/686053708261228577/694233507500916796/710384954982400072

Hello
The download works fine. I can open the .md file in Atom an there is nothing wrong with it. When it is opened in Obsidian the Preview-Mode works absoluty fine, but when I switch to the Edit-Mode suddenly there are everywhere small red dots. When I am hovering over them \uad appears. Can anyone help me?

macOS 10.15.7
Firefox 85.0
Markdown 2.4.1
Obsidian v0.10.13
Atom 1.54.0

Inspired by a discussion on the Obsidian discord: https://discord.com/channels/686053708261228577/700466324840775831/710709975219175465

Create, in the settings, a front-matter template whereby you specify which fields you want and the format they should take. For example:

---
URL: {url}
Author: {Author}
---
> {Excerpt}

# {Title}

Provide the possibility to highlight sections of a clipped page, if possible even in multiple colors.

Obsidian for example allows to highlight text in a marker style by using ==example text==. This a simple yellow marker, equivalent to <marker>example test</maker>.

Other bookmarking tools such as diigo even provide multi-colored highlights, which also allows multiple users in to mark the same text or allow to differentiate the marking of multiple passes over the same text. This would be for example equivalent to <marker style="background-color: greenyellow">example test</maker>.

This could be provided as option in the the context menu as Highlight Selection as Markdown with different colors as sub-options.

I noticed that file names with / in them don't get cleaned.
When I for example try to save the following article it tries to make a folder called Woche 10 and a file called 2020.md:
https://www.republik.ch/2020/03/06/joe-biden-ueberholt-sanders-fluechtlinge-auf-lesbos-und-aerger-mit-libra

As requested by BASE on the Obsidian Discord

I primarily download from a site which links to an internal database (I guess at least), so the links do not work in obsidian any longer. Would be very nice, if there was an option to disable that.

It shouldn't be hard to add another option for stripping links from the markdown.

This section will be converted to the following markdown, it seems the links are invalid for markdown form in any means.

かならず‐しも【必ずしも】

の解説

--------------------

［副］《「し」は副助詞、「も」は係助詞》打消しの語を伴って、必ず…というわけではない、…とは限らない、という気持ちを表す。「金持ちが必ずしも幸福ではない」「必ずしもそうとはいえない」

[「かならず【必ず】」の全ての意味を見る](/word/%E5%BF%85%E3%81%9A/#jn-43571)

類語

-   [あながち](/word/%E5%BC%B7%E3%81%A1/#jn-5381)
-   [まんざら](/word/%E6%BA%80%E6%9B%B4/#jn-210389)