[GUI Plugin] Overdrive Link

[jhowell]

Version 1.42.0 - 13 Jul 2016

Handle domain name change of 3M Cloud Library from ebook.3m.com to ebook.yourcloudlibrary.com.

Allow search for recommendable books at new-style (2016) OverDrive sites.

Fix inability to sign in to the recently modified OCLS Freading site. (Fixes Failed to locate sign in form)

Fix error when checking current availability of Scribd books. (Fixes Exception (Missing Scribd.R.render(Scribd.R.Shared.React.LibraryButto n)))

[Terisa de morgan]

I have a problem with FreeLIbrary FReading search, when I'm searching for a book:

Spoiler:

Search FLP.FR for things by molly harper
Get http://freelibrary.freading.com/sear...r=molly+harper
Search FLP.FR for things by molly harper: : Server response error (missing "div id=all-results")
Traceback (most recent call last):
File "calibre_plugins.overdrive_link.jobs", line 1170, in find_books_at_lib
File "calibre_plugins.overdrive_link.freading", line 183, in find_books
File "calibre_plugins.overdrive_link.freading", line 238, in find_books_by_format
File "calibre_plugins.overdrive_link.parseweb", line 212, in must_find
LibraryError: u'missing "div id=all-results"'

It always happens. I have the last version installed.

[jhowell]

Originally Posted by Terisa de morgan

I have a problem with FreeLIbrary FReading search, when I'm searching for a book...

I have been unable to duplicate this problem. If it is still happening for you then I will need some help to debug it.

It appears that Freading is sending an unexpected response to the search request. You can set a plugin tweak that will cause the plugin to save responses from web sites when errors are detected.

Go to Preferences -> Advanced -> Tweaks -> Plugin Tweaks and add this line:

overdrive_link_save_responses_on_error = True

You will need to restart calibre for this to take effect. The log files will be saved in the current directory of the calibre process so it is best to use a command window to "cd" to your desired directory and then launch calibre from there.

Perform the search that is failing for you. The log will indicate what file name was used to store the response from Freading that leads to the error. You can zip it and attach it to a MobileRead post.

Don't forget to remove the plugin tweak when you are done testing.

[Terisa de morgan]

Thank you, I'll do it this morning (tomorrow is not valid at this moment )

[Terisa de morgan]

The log is attached to this post, I hope it is useful. Thank you for your help.

[jhowell]

Originally Posted by Terisa de morgan

The log is attached to this post, I hope it is useful. Thank you for your help.

Yes, that was very helpful. For some reason your search results from Freading have all of the javascript delivered as part of the main html page, instead of being delivered separately. This lead to incorrect parsing.

I have attached to this post a new version of the plugin that should fix this problem for you. Could you try it out and let me know before I release it generally? Thanks.

[Terisa de morgan]

It works for me now, as it finds some books I knew they were at the library. Thank you very much!

[jhowell]

Originally Posted by Terisa de morgan

Thank you very much!

You're welcome.

Now released:

Version 1.43.0 - 02 Aug 2016

Fix error seen when searching Freading. (Fixes Server response error (missing "div id=all-results"))

3M Cloud Library is now known as Cloud Library. The internal identifier for Cloud Library links will remain '3m' for compatibility with previous plugin versions.

Use the folder "Overdrive Link Logs" in the plugin configuration directory instead of the current directory of the calibre process for storage of temporary debug files.

[seacity1]

I just updated this plugin & now unable to sign in to any library. This is th result for only 1 library but getting a HTTP Error 500: Internal Server Error on 5 public libraries & Open library. Was working fine before upgrade. Have tried restarting both Calibre & my computer.

Starting job: Search for books by 1 author
Overdrive Link 1.43.0, calibre 2.57.1
Search formats: Adobe EPUB eBook, Kindle Book, MP3 Audiobook, Mobipocket eBook, Open EPUB eBook, OverDrive MP3 Audiobook, OverDrive WMA Audiobook, WMA Audiobook
Get http://digitalbooks.brooklynpubliclibrary.org
Signing in to BPL: Retrying digitalbooks.brooklynpubliclibrary.org on exception HTTPError (Internal Server Error)
Signing in to BPL: Retrying digitalbooks.brooklynpubliclibrary.org on exception HTTPError (Internal Server Error)
Signing in to BPL: : HTTPError (Internal Server Error)
Traceback (most recent call last):
File "calibre_plugins.overdrive_link.jobs", line 1200, in sign_in
File "calibre_plugins.overdrive_link.overdrive", line 331, in sign_in
File "calibre_plugins.overdrive_link.net", line 198, in open_url
HTTPError: HTTP Error 500: Internal Server Error

[Terisa de morgan]

The problem is not the plugin but the server. If you go to the library page, you will find that it doesn't work. Same for Orange County.

[seacity1]

Originally Posted by Terisa de morgan

The problem is not the plugin but the server. If you go to the library page, you will find that it doesn't work. Same for Orange County.

Thanks! Seems to be OK now.

[Terisa de morgan]

I am at this moment, too, and the plugin shows no error. A couple of hours ago, when the plugin showed the error, I couldn't login to my account.

[lairdb]

Is there any reason not to pick up availability during baseline search/discovery? (Presuming that the user has enabled ol_wait_weeks?)

[jhowell]

Originally Posted by lairdb

Is there any reason not to pick up availability during baseline search/discovery? (Presuming that the user has enabled ol_wait_weeks?)

The current search algorithm relies on caching information from the past and across multiple libraries to speed up searches. That means that the information needed to determine wait time for books is often not available during search without doing extra work.

I will consider adding an option to automatically do an availability check after a search to save the user a step.

[lairdb]

Originally Posted by jhowell

The current search algorithm relies on caching information from the past and across multiple libraries to speed up searches. That means that the information needed to determine wait time for books is often not available during search without doing extra work.

I will consider adding an option to automatically do an availability check after a search to save the user a step.

Okay -- makes sense.

Such an option would be handy -- at present, I'm prevented from queueing an availability check, since "Search is already in progress...".