Suggestion: Include hyperlink action in PDF output for hyperlinks in webpage

[ariya]

[email protected] commented:

PhantomJS v1.2.0

If you rasterie the URL http://www.nu.nl/buitenland/2590433/voedselhulp-wordt-in-mogadishu-gestolen-.html, it contains selectable text in PDF. Which is great!

But this news page contains a lot of link when opened in a normal browser.
The PDF output shows no similar hyperlinks on the available text.

My suggestion: associate a hyperlink action where relevant.

This allows the PDF to more resemble the actual website.

Disclaimer:
This issue was migrated on 2013-03-15 from the project's former issue tracker on Google Code, Issue #196.
🌟   8 people had starred this issue at the time of migration.

[ariya]

[email protected] commented:

Probably need to fix Qt PDF code for this.

 
Metadata Updates

• Label(s) removed:
  • Type-Defect
• Label(s) added:
  • Type-Enhancement

[ariya]

[email protected] commented:

There was a bug in Webkit recently where hyperlinks were not included in the PDF output, that's been fixed. Safari's PDF output had the problem too for some months. It's fixed now. So maybe getting a more recent version of webkit will fix the issue.

[ghost]

[email protected] commented:

Hi, I have upgraded to v1.7 and I still can't get clickable hyperlinks in any PDF output. I have tried it with the Quickstart examples too.

[JamesMGreene]

[email protected] commented:

Related discussion, including links to bugs and patches for wkhtml2pdf: https://groups.google.com/d/topic/phantomjs/DuCbXTEnhRc/discussion

 
Metadata Updates

• Status updated: Accepted

[ariya]

[email protected] commented:

More details :

Here is the original but report in Webkit:

"Bug 65076 - WebKit2: Printing to PDF loses URL links"

https://bugs.webkit.org/show_bug.cgi?id=65076

It got fixed in Safari 5.1.4, which uses ..

http://en.wikipedia.org/wiki/Safari_version_history

... webkit 534.54.16

Using /examples/useragent.js with Phantom 1.8.1 i just upgraded to i get :

The default user agent is Mozilla/5.0 (Unknown; Linux x86_64) AppleWebKit/534.34 (KHTML, like Gecko) PhantomJS/1.8.1 Safari/534.34

And 534.34 is still behind 534.54

Hope it helps.

[thom4parisot]

[email protected] commented:

And since, have you tried rastering an HTML page containing hyperlinks to
PDF and getting them clickable?

Because it's related to a QT lib, not Webkit directly.

[ariya]

[email protected] commented:

I tried, and they are still not clickable.

i don't know if the patch used in the webkit bug report :

https://bug-65076-attachments.webkit.org/attachment.cgi?id=121534

.. is applicable to qtwebkit, it looks like a fix for Safari only, no?
(i'm not familiar at all with webkit's code and how qtwebkit relates to it)
i understand they switched to using PDF Kit, i don't know what is the equivalent od PDF Kit used by qtwebkit.

[On the other end, another issue with page-break-before: always; seems to work for me now.]

[passoir]

It seems the wkhtmlpdf has the clickable links in generated pdf file. Is it possible to apply their webkit patch to phantomjs?

[wachunga]

FWIW, my links are clickable when the pdf is generated on Ubuntu 12.04.1, but not on my mac (10.6.8).

[ddooley]

I have phantomjs 1.9.0 on ubuntu 12.04.2 but no clickable links.

[AndrewEastwood]

I have the same issue (1.9.0 on Ubuntu 11.10)

[duybinh0208]

In summary, Can anyone give me the exactly solution for this problem? I'm using phantomjs 1.9 in windows and The default user agent is Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.34 (KHTML, like Gecko) PhantomJS/1.9.0 Safari/534.34 Your Http User Agent string is: SpecialAgent.

[ddooley]

I feel in the dark too. I tried compiling webkit on my ubuntu 12.10.1 , hoping that success there would make the pdf link problem go away; or that I'd be in a position to recompile phantomjs. But after 5 hours I'm lost - a ./autoconf batch file stops with a libglib-2 version insufficient message. I see dire warnings about trying to remove / reinstall this library (http://unix.stackexchange.com/a/49515), because of all the dependencies (all the latest official versions). So can anyone say which upcoming version of PhantomJS will adopt webkit 534.54 or greater?

Much appreciated!

[JamesMGreene]

Well, as far as I've been able to discern, the WebKit included in Qt 5.0.2 is from revision 136242, which was last changed on 2012-11-30 (just a few days before the Qt 5 release). That seems plenty current.

I'm not sure what the corresponding 5xx.xx version number is, though...?

[ariya]

It's less of a problem with WebKit and more with the PDF generation. Replace WebKit with any document technology and the hyperlink will not exist if both sides do not agree to detect + produce it.

[ddooley]

I see now (from thread #10448) this would go away with Phantom 1.10 or certainly 2.0 because of adoption of Qt 5 . So some time in the fall/winter 2013 ?! Ok.

And certainly I appreciate that there's a lot of work in trying to keep PhantomJs up to date with fundamental changes in its underlying dependencies like qt/webkit !

Cheers,

[wachunga]

One thing I noticed: links are clickable in the PDF if the <a> tag has the same text as the href. So <a href="foo.com">foo.com</a> works, but not <a href="foo.com">Click here</a>.

Not ideal, to be sure. But might help someone.

This is as of phantom 1.9.0.

[wachunga]

Correction to my previous post: anchor tags don't help at all. It's just my PDF viewers that are making the url text clickable.

Acrobat and Chrome PDF viewer pick up anything with http(s):// and/or www. Preview also picks up bare urls like website.com.

So basically the hyperlinks aren't coming through from PhantomJS at all.

[AndrewEastwood]

:(

[apeiniger]

Any news on this?

[tanmaygarg]

Hey, is this issue fixed yet?

[szepczynski]

+1

[apeiniger]

+1

[pofider]

+1

[jobinsjohn]

Any update on the click feature on hyperlinks?

[thom4parisot]

I guess it will be part of PhantomJS 2.

[jobinsjohn]

@oncletom Well i have solved the problem to an extend. Its not a complete solution. But it serves my pupose for the time being. May be the solution will be there in PhantomJS 2.

[jhihn]

Note that the patch was included in Qt 5.6.0, which should be about in about 6 months.
https://bugreports.qt.io/browse/QTBUG-44563

[vitallium]

@jhihn I don't think we will ever migrate (upgrade) to Qt 5.6.0.

[danielbayley]

I don't think we will ever migrate (upgrade) to Qt 5.6.0.

Why not?!

[vitallium]

Because release Qt 5.5.0 was the latest release of QtWebkit.

[astefanutti]

Qt WebKit has indeed been removed from the Qt release packages though it is still maintained and part of the Qt 5.6 release from the source code which is the way PhantomJS is integrating Qt WebKit in its code line. So unless there is an actionable plan to migrate to the new Qt WebEngine or any other underlying Web engine, it is still possible to upgrade to Qt 5.6 and benefit from that enhancement.

[vitallium]

@astefanutti keeping QtBase and QtWebkit out of sync might be dangerous.
We have future development plan, but we don't have people and resources.

[JordanMajd]

+1

Hey guys, I know you have a lot of work on your plate but I'd love to see this patched.

[stgogm]

Currently, if you place your URLs, anchor or plain text, inside a <table>'s <td> element they will be clickable when opened in Chrome or Opera.

[astefanutti]

Given the externalisation of Qt Base and Qt WebKit as submodules, PR #13250 is superseded by the two following PRs:

• Add support for drawing a hyperlink in QPdfEngine vitallium/qtbase#1
• Render anchors as clickable links in PDF documents vitallium/qtwebkit#1

[astefanutti]

As vitallium/qtbase#1 and vitallium/qtwebkit#1 have been merged and are now included in master branch, this can be labelled for next release and closed.

[vitallium]

Landed in master.

[alfchee]

@RomanMinkin I'm having PDF's with links working but I want to avoid what you did about to make:
<a href="http://domain.com">This is a link</a> some text
Looks like: This is a link (http://domain.com) some text

It is making akward some reports that I want to render in PDF.

[alfchee]

Ok, no problem, I found the answer in another issue of this project, alanshaw/markdown-pdf#63

I have solved my problem.

[RomanMinkin]

@alfchee sweet. I used css to display link that way.