#1
|
|||
|
|||
Linkgrabber Not Parsing Links
Earlier today, for some unknown reason, Linkgrabber has stopped parsing links. As I write I have, on the bottom bar, a huge line of clipboard circles with spinning green clock hands. And it has been this way for about 45 minutes.
I have had to end the JDownloader process in task manager twice today in order to get Linkgrabber working again. This worked the first two times, however Linkgrabber has now ceased to function at all. This affects clipboard links from all websites. I add the links by right clicking and selecting copy link address. However, this issue persists also when right clicking and selecting download with JDownloader. Linkgrabber has worked perfectly until now. Also, I should say that, once there are links that have failed to parse in the clipboard queue, I am also unable to start the download of any links that are held within the Linkgrabber. I can select them, but they do not move to the downloads tab. I have to end the JDownloader process in task manager and restart, then the pre-existing (already on the downloads tab) links can be downloaded. I would really appreciate some help/advice. Log files listed below. 07.02.19 15.29.19 <--> 07.02.19 15.26.44 jdlog://7466576935451/ 07.02.19 19.51.45 <--> 07.02.19 20.31.23 jdlog://9476576935451/ Last edited by gandalf_grey91; 08.02.2019 at 15:03. |
#2
|
|||
|
|||
Another thing I have noticed is that some pre-existing links, once moved to the downloads tab, fail to download with the "invalid download directory" message.
It cannot be an invalid download directory, because there other files currently downloading into that directory in JDownloader. For these troublesome links, I have tried changing the download directory, to another known good directory. However, they continue to display the same "invalid download directory" message. Last edited by gandalf_grey91; 08.02.2019 at 15:01. |
#3
|
||||
|
||||
Will answer by monday or someone else before me
__________________
JD-Dev & Server-Admin |
#4
|
|||
|
|||
This issue means I am unable to use JDownloader, at all.
I would really appreciate some help. Please let me know if you require more information. Last edited by gandalf_grey91; 12.02.2019 at 12:10. |
#5
|
||||
|
||||
It seems the vk plugin is broken and filename parsing is invalid and resulting in very very very very very long filename which fails because it's too long resulting in *invalid download directory* error.
Can you please provide example links that fails because I heavily doubt that it affects *all websites* because in log only those from vk seems to have issues Try to disable Settings-Advanced Settings LinkgrabberSettings.autofilenamecorrectionenabled and LinkgrabberSettings.autopackagematchingcorrectionenabled Looks like you have lots of links within a single package and auto filename correction becomes very slow in such situations. According to log this causes HUGE delay from adding links from linkgrabber to download list as the auto correction is still running
__________________
JD-Dev & Server-Admin |
#6
|
|||
|
|||
Hi Jiaz, many thanks for looking into this issue.
Sorry for the delay in responding, I have been going some experimenting. You are correct, it is the too long VK links that are causing the problem and not all websites links. The other website's links are delayed once the Linkgrabber "pipeline" is delayed with an overly long VK link. But I am not sure that the VK plugin is broken, for some reason, certain VK users like to add every searchable word under the sun, into the filenames, creating gigantic filenames that are a problem for MS Windows. While experimenting, I have found that: Disabling LinkgrabberSettings.autopackagematchingcorrectionenabled and enabling LinkgrabberSettings.autofilenamecorrectionenabled does not resolve the extra long pause to parse Linkgrabber links. But enabling LinkgrabberSettings.autopackagematchingcorrectionenabled and disabling LinkgrabberSettings.autofilenamecorrectionenabled does seem to prevent the extended delays in parsing Linkgrabber links. Equally, disabling both settings also seems to resolve the delayed parsing issue. However, either way. this does not seem to prevent the creation of files with filenames so long that they cannot be handled in Windows Explorer. I had to delete many of these files, using command prompt, because they could not even be renamed in the command prompt. To deal with the problem, I would suggest that JDownloader could use an advanced "LinkgrabberSettings" setting that forces files, with filenames so long that they cannot be handled by Windows Explorer, to be renamed to the standard 8+3 format when downloaded. This setting could be left unticked by default, if it would cause issues to certain users for whatever reason. Just as long as the function is there, for those that need it. This solution could also potentially resolve the "invalid download directory" error. Last edited by gandalf_grey91; 27.02.2019 at 18:18. |
#7
|
||||
|
||||
Nothing to be sorry for
Those two autopackagematchingcorrectionenabled and autopackagematchingcorrectionenabled can be very complex and cpu intensive for longer lists and/or larger packages in list. So disabling them will indeed reduce cpu usage and avoid the delays. Can you provide example links, then maybe I can *easily* fix the filename parser or add workaround. Send them to support@jdownloader.org
__________________
JD-Dev & Server-Admin |
#8
|
|||
|
|||
Hi Jiaz, Many thanks for the help.
I have forwarded 4 examples to support@jdownloader.org, as requested. Please let me know if you require any further examples. |
#9
|
||||
|
||||
Thanks. Will check and respond by mail
__________________
JD-Dev & Server-Admin |
Thread Tools | |
Display Modes | |
|
|