Error with news download

jasper 20 · 07-19-2016, 07:50 AM

Hello. I wonder if anyone can help me with this problem?

Since around 4 days ago a news source that I download daily has stopped working. The news source is in the Japanese section and is called "日経新聞電子版(Free, Max)", which is the Nikkei newspaper.

Another news source also from the Japanese section is working as usual.

I have attached a text file of the job details for this morning's attempted download which resulted in an error after 60 mins of trying.

Getting this news each morning is quite important for me and I'd be very grateful for any help in solving the problem. Thanks

jasper 20 · 07-19-2016, 08:10 AM

Just to add...
I'm using Windows 7 and Calibre 2.61

jasper 20 · 07-19-2016, 09:27 PM

To add some more information:

1. I understand now that this is a 'recipe' problem. My apologies, I'm very new to Calibre.

2. I've looked through many recipe problems most of which seem to be about broken recipes or paywalls. I don't think this recipe is broken because it does appear to be downloading some articles as shown in the log:

Fetching http://partsa.nikkei.jp/parts/ds/ima...tlogin_off.gif
Parsed http://www.nikkei.com/article/DGXLAS...6A710C1000000/ in 3.8 seconds
Fetched http://assets.nikkei.jp/release/v1.0...ayout.news.css in 0.324000 seconds

...and many more like this

3. After a while it seems to run into problems with
Recursion limit reached.
and
Initial parse failed, using more forgiving parsers
and
PropertyValue: No match: ('CHAR', u':', 1, 106153)

4. This goes on for some time and finally
Python function terminated unexpectedly
(Error Code: 1)
and
MemoryError

If anyone can help I'd be very grateful.
Thanks

kovidgoyal · 07-19-2016, 11:49 PM

That error just indicates the recipe is downloading too much junk. It needs to be updated to handle whatever change on the website caused it. As I cant read Japanese, it's hard for me to fix, but after a quick glance at the website, this should make it better:

https://github.com/kovidgoyal/calibr...152d8f537393bd

If it still does not work then you will need to wait for somebody that reads Japanese to clean it up further.

jasper 20 · 07-20-2016, 03:44 AM

Oh, thank you very much.

Yes, it works and is much quicker than the earlier version of the recipe.

You are right that the earlier version picked up a lot of junk. I assumed that it was just part of the process and could not be eliminated. The revisions that you kindly made have removed a lot of that junk but, interestingly, along with the articles it now picks up some other stuff which is rather junk-ish

But that doesn't stop me from being able to read the articles and I understand completely that it is almost impossible to understand a Japanese web site without understanding the language. So thanks again for your efforts they are very much appreciated.

I had a look at the recipe via the link you sent and I'm afraid it is quite alien to me so I'm not able to add anything of value in terms of cleaning it up further.

Thanks a lot. You've cheered-up my day!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
News Download Error	wally_k	Calibre	3	02-07-2016 01:34 PM
Download News tags	theant	Library Management	0	03-15-2014 03:03 PM
News Download Conversion Error	JeffGeorge	Recipes	5	10-25-2013 12:07 PM
error on link in ZeitONLINE TOC; Beebook 2010 with calibre news download	eNEWS	Recipes	1	06-27-2011 04:25 AM
Automatic News Download	brewjono	Calibre	3	10-09-2010 08:49 AM

07-19-2016, 08:10 AM	#2
jasper 20 Junior Member Posts: 6 Karma: 10 Join Date: Jul 2016 Device: kindle	Just to add... I'm using Windows 7 and Calibre 2.61

07-19-2016, 09:27 PM	#3
jasper 20 Junior Member Posts: 6 Karma: 10 Join Date: Jul 2016 Device: kindle	To add some more information: 1. I understand now that this is a 'recipe' problem. My apologies, I'm very new to Calibre. 2. I've looked through many recipe problems most of which seem to be about broken recipes or paywalls. I don't think this recipe is broken because it does appear to be downloading some articles as shown in the log: Fetching http://partsa.nikkei.jp/parts/ds/ima...tlogin_off.gif Parsed http://www.nikkei.com/article/DGXLAS...6A710C1000000/ in 3.8 seconds Fetched http://assets.nikkei.jp/release/v1.0...ayout.news.css in 0.324000 seconds ...and many more like this 3. After a while it seems to run into problems with Recursion limit reached. and Initial parse failed, using more forgiving parsers and PropertyValue: No match: ('CHAR', u':', 1, 106153) 4. This goes on for some time and finally Python function terminated unexpectedly (Error Code: 1) and MemoryError If anyone can help I'd be very grateful. Thanks

07-19-2016, 11:49 PM	#4
kovidgoyal creator of calibre Posts: 45,250 Karma: 27110894 Join Date: Oct 2006 Location: Mumbai, India Device: Various	That error just indicates the recipe is downloading too much junk. It needs to be updated to handle whatever change on the website caused it. As I cant read Japanese, it's hard for me to fix, but after a quick glance at the website, this should make it better: https://github.com/kovidgoyal/calibr...152d8f537393bd If it still does not work then you will need to wait for somebody that reads Japanese to clean it up further.

07-20-2016, 03:44 AM	#5
jasper 20 Junior Member Posts: 6 Karma: 10 Join Date: Jul 2016 Device: kindle	Oh, thank you very much. Yes, it works and is much quicker than the earlier version of the recipe. You are right that the earlier version picked up a lot of junk. I assumed that it was just part of the process and could not be eliminated. The revisions that you kindly made have removed a lot of that junk but, interestingly, along with the articles it now picks up some other stuff which is rather junk-ish But that doesn't stop me from being able to read the articles and I understand completely that it is almost impossible to understand a Japanese web site without understanding the language. So thanks again for your efforts they are very much appreciated. I had a look at the recipe via the link you sent and I'm afraid it is quite alien to me so I'm not able to add anything of value in terms of cleaning it up further. Thanks a lot. You've cheered-up my day!

Advert

Advert