[CWB] Set up images in text metadata

[Daniel Wachter]

Ah, thank you! Will try it and report!


Am 14.01.2021 um 16:53 schrieb Uhrig, Peter:
> CQPweb does not need to be able to read it, the web server needs to be able to read it.
> So if you are, say, in Ubuntu, you'd place the images in /var/www/html/images and then have the image links as "image:/images/whatever.jpg"
>
> -----Ursprüngliche Nachricht-----
> Von: cwb-bounces at sslmit.unibo.it <cwb-bounces at sslmit.unibo.it> Im Auftrag von Daniel Wachter
> Gesendet: Donnerstag, 14. Januar 2021 16:51
> An: cwb at sslmit.unibo.it
> Betreff: Re: [CWB] Set up images in text metadata
>
> The admin manual says under "7.5.1 Free text":
>
> "
> Free text metadata fields are allowed to contain links to external resources of one kind or another.
> These are detected by examining the “prefix” of the field value, which is the part of the value before the first colon.
> [...]
> image
> This is a link to an image file (of any kind that the browser can open, including PDF for instance). It will appear as a clickable control which, when clicked, will load the image into an overlay on the page.
> Example: image:../images/orig-text-a1.pdf "
>
> In my experiments the image prefix works: The clickable control is available in the text metadata and CQPweb tries to open the image but can't find the file. Because the manual example uses a relative path, I expected/hoped that I only would have to give CQPweb reading permissions for the directory where the images are stored and then CQPweb would handle the rest... Is this assumption wrong?
>
> Best,
> Daniel
>
>
> Am 14.01.2021 um 16:17 schrieb Stefan Evert:
>>> On 14 Jan 2021, at 15:36, Daniel Wachter <wachter at cl.uni-heidelberg.de> wrote:
>>>
>>> I try to integrate images in my corpus metadata as described in the admin manual p. 72 using the full path of the image files like "image:/var/cqp/upload/covers/7410.jpg". I tried to give cqpweb access to the images with "sudo chown -R root:www-group /var/cqp/upload/covers/" but there's something wrong with the path: When I try to open an image in the installed metadata cqpweb gives me "The requested URL was not found on this server".
>> The image link isn't handled by CQPweb but rather passed through to the client browser.  So you need to provide a public URL (i.e. http: or https:) where the browser can download the image file.
>>
>> "image:" isn't a valid URI scheme as far as I know – where did you find this format?
>>
>>> How can I give cqpweb access to my image files or where should I place them? Or is this rather a question of my server configuration and has nothing to do with cqpweb per se?
>> The usual solution is to make the image files available on your Web server (possibly password protected) independently of CQPweb, and then construct appropriate URLs and insert them in the corpus metadata.
>>
>> Best,
>> Stefan
>> _______________________________________________
>> CWB mailing list
>> CWB at sslmit.unibo.it
>> http://liste.sslmit.unibo.it/mailman/listinfo/cwb
> _______________________________________________
> CWB mailing list
> CWB at sslmit.unibo.it
> http://liste.sslmit.unibo.it/mailman/listinfo/cwb
> _______________________________________________
> CWB mailing list
> CWB at sslmit.unibo.it
> http://liste.sslmit.unibo.it/mailman/listinfo/cwb