[LH3291] [LH4662] Links with special characters won't import ( Beta 2.9.0.11, Win, German)

Scrivener Windows Beta (Archived)
1

Hi everybody,

short intro: I’m a new user and software developer, already found a lot of great uses for Scrivener.

In the latest Beta, adding a website to the Research folder fails when the URL has special characters like Umlauts.

Repro steps (messages and menu items translated from their German texts):

  • Right click “Research”
  • “Add” -> “Website”
  • URL: for instance try freimaurer-wiki.de/index.php/Brüderlichkeit - i.e. the Umlaut is not URL encoded)
  • Enter any title, click OK
  • Error message: “URL import failed”

Keep up the good work!

Phil

2

Thanks for reporting this issue, Phirl. I have filed it in our bug tracking system, Unfortunately I cannot propose a workaround at the moment.

3

Thanks, @tiho_d - my current workaround is copy-pasting into a new document, for my purposes, this works temporarily.

4

You can also try saving the URL into an MHT via the browser and drop it inside the Binder Research folder.

5

The bug still exists. (Version 2.9.0.36 Beta (818046) 64-bit - 28 Jan 2020)

Interesting: If I e.g. enter the link de.wikipedia.org/wiki/J appears. Everything after the “J” is ignored.
But if I look at the link to the imported website in the status bar, the correct link de.wikipedia.org/wiki/Jōmon-Zeit appears there in both cases and if I click on the link, the correct website appears in the browser.
Of course I could convert the corresponding page into a PDF, but then I would not be able to search the text anymore. Therefore I would be very grateful for a fix of this bug.

6

Small addendum: The following link will be imported without problems, although it decodes the url in the same way as the link above:
https://de.wikipedia.org/w/index.php?title=J%C5%8Dmon-Zeit&oldid=195264490

7

We can also import a web page by dragging the left icon from the browser’s URL bar – typically a padlock for a secure site – into the Research section in Scrivener’s binder. The same bug will be effected. Scrivener’s footer bar displays the intended site link, but the page actually imported and displayed reflects a URL truncated at the special character.

Rgds - Jerome

8

Likewise, an addendum. The link with an umlaut, originally reported by PhilR to fail on import, does import successfully in v. 36.

freimaurer-wiki.de/index.php/Br … erlichkeit

Whereas the link reported by Pekoo that truncates on import has an overscored o. A transliteration of the Japanese? Of course we’ll experience the same truncation effect on English-language URLs that include the character. The link below will drag or paste in as Wikimedia Commons Category:J.

commons.wikimedia.org/wiki/Cate … nal_Museum

So an umlaut flies while an overscore flails as of this round.

Cheers - Jerome

9

Websites with special characters (tested with umlauts) are still not imported. Is this bug even processed? It should only be a bug in the used character set. (Beta 2.9.9.8)

10

It’s probably something on your machine. I just tested it and it works perfectly for me. Encoded or not encoded it doesn’t matter, it imports the page without a problem.

Clipboard Image.png
Clipboard Image.png1397×680 300 KB

11

When I visit the website de.wikipedia.org/wiki/K%C3%A4se
with umlauts in the address, I only get an error message (see picture). Within a web page, special characters do not cause any problems for me either - only with URLs.

error scrivener.jpg

12

@Pekoo: We have adjusted the Web Page import function. Give it another test with the next update.

13

As I said, I tested it with umlauts in the address(both encoded and not URL encoded), and it works for me.

14

@Krastev: The Pekoo’s example did not work and I could reproduce the error. Now all example URLs from this thread import successfully.

15

I managed to import the link given, but I noticed the progress bar froze at about 20%, leading me to believe Scriv had frozen. I waited about a minute, and it did complete the import. It was just slow.

16

At least since 2.9.9.9. there has been significant progress: Umlauts are accepted in the address. :slight_smile: Unfortunately the import still aborts as soon as you use an umlaut in the title. :cry:

17

Thanks, Pekoo! Umlauts in the title has been fixed too and will be available in the next update. Avoiding Unicode characters in the Title during import is the workaround at the moment. Once imported, you can change the Binder name with anything you like.

18

Thank you for this information. :smiley: