[I3ordo]

Hi, i found a situation where this new feature would work wonders.
This would be best fitted to the packagizer...? as it should work on grabbed files after the filtering.)

When link grabber receives a new link, If certain source url is matched and certain file type is matched, set file count limit to 1 (or 2) for that pasted link.
It should sort by alphabetical filename and decide which file is first

My situation for asking this:
A certain site, when it s link pasted, brings loads of img files , most of them are filterable as their download url is different. But the ones that i was not able to filter out to a single one comes from same path.

Normally,in other sites, i would set all the file names to "page title" and have the first one downloaded image , cancel out the preceding ones via file already exists error (intentional)
But i need the filename but not the title and of course i cant make them take the filename of the first arrived file....
Here, **External links are only visible to Support Staff****External links are only visible to Support Staff**
i filtered "/media/images/banner/" and "/media/cache/tuk_model_custom_filter_en
so that only images from url"model_new_big_en/model_images.+\/(.+)" arrives

But that link gives me 2+ images as they are extra images . others links bring more.

this feature would limit the number files to 1 per pasted link.

[pspzockerscene]

Neither do I see us adding this feature nor do I see this working for your case.
Please describe which URLs you want.
I guess thise ones?
b2.3ddd.ru/media/cache/tuk_model_custom_filter_en/model_images/0000/0000/4174/4174740.6274ca53481de.jpeg

Please atach the filter rules you've created so far so we can find a solution on how to filter everything except the image-URLs you want to have.

-psp-

[Jiaz]

@I3ordo: The packagizer has no knowledge/influence of/on other links.

For your case the best would be a LinkCrawler rule, see https://support.jdownloader.org/Know...kcrawler-rules and https://support.jdownloader.org/Know...le-deepdecrypt
so you tell/teach JDownloader which content/links you are interested in.

[I3ordo]

In this case, it s currently impossible....
I was able to filter out a big percentage of unwanted images already , they had different url paths so very easy. but this is hard because they all come from same path

site.com/my_chair/ (source url)
site.com/abc/abc.123.jpg (desired image)
site.com/abc/abd.124.jpg(unwanted)
site.com/abc/abe.126.jpg(unwanted)

the abcs are all random so, i just can NOT define a general rule and tell , "dont download from "abd*" or "abe*", because those paths will become needed, just not for that pasted link....
So the only differentiator is the numerically smallest number.
Currently, i paste link, wait for all links to get grabbed, delete the numerically larger ones. until only one image is left for that pasted link, then move onto the next paste.

ofcourse, i just did it twice then gave up and instead just dragging images to an explorer folder without JD

i gave actual links btw , i just dont think it is doable with current tools.

question:
Is there paste event IDs in JD?
Maybe a script could do this:
For Paste event ID: 123#
get list of files downlaoded for that event
delete all downloaded links until count is 1, reverse numerically sorted.
pff


Or it could be:
For every source url,
delete all jpeg until file count=1
delete all rar until file count=1 ofcourse reverse numerical order...

[Jiaz]

@I3ordo: Instead of filter out unwanted content, you should create/use Linkcrawler rules and tell JDownloader in what content/links you are interested. You create a rule that matches the site url /my_chair... and then specify a deepPattern that matches only content/links you are interested in.
In case you need help with it, just let us know but please try for yourself first.

[raztoki]

I will add in a link crawler rule, you can also specify html components that are adjacent to what you're looking for instead over just a generalised pattern of url itself which is shared/reused in other locations, thus doing what Jiaz recommends and find what you want vs every url pattern.

Count wont work in my opinion as it implies the content you want will be either statically in a given position. My experience this is very rare.