Downloading metadata: Blank description overriding?

ownedbycats · 11-26-2022, 12:24 AM

I have a book with mobi-asin:B0B8G1287N.

When I download it with only Amazon for a source (along with title/author), it correctly picks up this book with the description.

If I also enable Goodreads, it combines the entries, but the description is now blank, presumably because the GR entry doesn't have one.

Is this a plugin issue, or should I file a Calibre ticket?

Comfy.n · 11-26-2022, 04:09 AM

Quote:

Originally Posted by ownedbycats

I have a book with mobi-asin:B0B8G1287N.

When I download it with only Amazon for a source (along with title/author), it correctly picks up this book with the description.

If I also enable Goodreads, it combines the entries, but the description is now blank, presumably because the GR entry doesn't have one.

Is this a plugin issue, or should I file a Calibre ticket?

I know that downloading proper metadata can be tricky and I'm always testing new combinations, so I've tried to reproduce this issue by using a dummy file with that same asin. FWIW, the results:

- by using default Amazon PI (the built-in one), I had the same happen which you described: got a blank description when GR is also enabled. Other GR meta fields (grrrating, grvotes, GR identifier etc.) were downloaded as expected.

- using Amazon Multiple Countries (settings: US - Amazon servers): with both GR and AMZ-MC enabled, I got the Amazon description returned -- AN INVASION. A CONSPIRACY. ONE SHOT AT REDEMPTION. (Gee!!

sounds fun enough for me), but then I got no meta from GR returned (not one GR field's meta was downloaded)

(a third result I just managed to get: by setting "Choose server automatically" in AMZ-MC, I had the same output as in the first case (a blank description), so GR meta fields were downloaded)

kovidgoyal · 11-26-2022, 04:16 AM

When merging comments during metadata download calibre picks the longest. The goodreads plugin is probably setting the comments to a string with spaces or similar. It should instead set it to None or the empty string, or just not set it at all when no comments are present.

See the merge() function in sources/identify.py

ownedbycats · 11-26-2022, 12:05 PM

Thanks. I'll ask on the GR plugin thread.

kiwidude · 12-23-2022, 09:02 PM

Quote:

Originally Posted by kovidgoyal

When merging comments during metadata download calibre picks the longest. The goodreads plugin is probably setting the comments to a string with spaces or similar. It should instead set it to None or the empty string, or just not set it at all when no comments are present.

See the merge() function in sources/identify.py

I have checked the Goodreads plugin code. It was returning an empty string in this circumstance. I have tried changing the Goodreads code to ensure it returns None, and I am *still* seeing the same problem reported above from the merge that takes place with the Amazon result.

To replicate, create this empty book:
Title: Lost Galaxy: The Complete Series
Author: Daniel Young
Identifiers: amazon:B0B8G1287N, goodreads:61991150

You can use the attached version of the Goodreads plugin that ensures it always leaves mi.comments as None rather than empty string.

Spoiler:

So if it is a Goodreads plugin problem I'm not quite sure where the problem could be?

kovidgoyal · 12-23-2022, 11:24 PM

For that test case no merging is happening, There is one result from goodreads and one from amazon. They have different titles, the goodreads one has (Books 1-6) append to the title. Thus they are not merged and you are presented with two results, one with comments which comes from amazon and one without which comes from goodreads.

kiwidude · 12-24-2022, 12:08 AM

Quote:

Originally Posted by kovidgoyal

For that test case no merging is happening, There is one result from goodreads and one from amazon. They have different titles, the goodreads one has (Books 1-6) append to the title. Thus they are not merged and you are presented with two results, one with comments which comes from amazon and one without which comes from goodreads.

Hi Kovid, Christmas wishes to you and the family!

Right no merge makes sense in the behaviour being seen. I was confusing myself by seeing both Goodreads and Amazon identifiers on the first row and thinking that was a "merge", which it isn't.

So I'm confused as to how @ownedbycats had an issue in the first place then - as you say these two books do not merge. I've followed their same steps of pulling Amazon metadata first, then enabling Goodreads metadata, trying a download and every single time I get two entries on the left, no merging.

In other words to me it looks like there never was a problem - or at least not one I can replicate with that specific book example.

11-26-2022, 12:24 AM	#1
ownedbycats Custom User Title Posts: 11,347 Karma: 79528341 Join Date: Oct 2018 Location: Canada Device: Kobo Libra H2O, formerly Aura HD	Downloading metadata: Blank description overriding? I have a book with mobi-asin:B0B8G1287N. When I download it with only Amazon for a source (along with title/author), it correctly picks up this book with the description. If I also enable Goodreads, it combines the entries, but the description is now blank, presumably because the GR entry doesn't have one. Is this a plugin issue, or should I file a Calibre ticket? Last edited by ownedbycats; 11-26-2022 at 12:32 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Metadata book description	gmw	Sigil	10	11-26-2012 07:47 AM
Titanium Backup Run has blank description	TonyGamble	Sony Reader Dev Corner	5	09-06-2012 08:55 AM
SONY PRS T1:Regarding metadata overriding by the buggy os	anishd19	Devices	4	01-03-2012 12:36 AM
No HTML tags in description metadata?	RobinD42	Library Management	11	02-22-2011 11:33 AM
Prepending Book with Description from Metadata	doodlydood	Conversion	2	01-29-2011 08:53 PM

11-26-2022, 04:16 AM	#3
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	When merging comments during metadata download calibre picks the longest. The goodreads plugin is probably setting the comments to a string with spaces or similar. It should instead set it to None or the empty string, or just not set it at all when no comments are present. See the merge() function in sources/identify.py

11-26-2022, 12:05 PM	#4
ownedbycats Custom User Title Posts: 11,347 Karma: 79528341 Join Date: Oct 2018 Location: Canada Device: Kobo Libra H2O, formerly Aura HD	Thanks. I'll ask on the GR plugin thread.

12-23-2022, 11:24 PM	#6
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	For that test case no merging is happening, There is one result from goodreads and one from amazon. They have different titles, the goodreads one has (Books 1-6) append to the title. Thus they are not merged and you are presented with two results, one with comments which comes from amazon and one without which comes from goodreads.