(Archived) [Feature Request] import starred articles from Google Reader

[meokey2000 0]

Hello,

 

Google Reader is dying. I believe many users, as myself, have tons of starred articles in Google Reader and they are looking for a method to export them and import into an alternative product for future reference. Evernote is such a prefect product to handle it, because it's born to be a note application with tagging and other rich features that people can make use and easy to search when needed.

 

I hope Evernote will be able to provide such a feature to import starred articles in GR and tag them automatically or with less interaction with users.

 

 

On the other hand, to convince people to use Evernote as an alternative product of GR, competing with many others such as Feedly, GereatNews and so on, and to feel confident to use Evernote as an Internet safer for articles, I suggest Evernote to provide Export function so that people, free users and premium users, will have the capability to export their staff out of Evernote whenever they want to. It doesn't necessarily means Evernote pushing users away, it indeed will attract users by respecting their options and wishes, giving them confidence with Evernote.

 

I can't overemphasize how important the Export function is to Evernote. As people saving money in the banks, Internet users save articles in Evernote. I never ever save my money into a bank I can not withdraw, no matter how good services it provides, no matter it's free of charge or not. Currently Evernote provides simply function to export articles in HTML, I would like to suggest more formats and better flexibilities to export.

 

I sincerely thank Evernote for its fantastic features, and hope it will be better every day.

 

Best Regards,

meokey

[jefito 5,589]

Unless I am misunderstanding what you're suggesting, it doesn't seem to make much sense for Evernote to scramble to add support for a dying (as far as we know) service, particularly when you can already accomplish what you're asking by using IFTTT, or possibly other services (e.g. Zapier?).

[meokey2000 0]

I suggest based on a fact that both GR and Evernote collect/storage valued data (articles, images, etc) from Internet for future reference.

 

Both of them are kinda online knowledge base application. The difference is the technology they employ - GR uses RSS and Evernote uses "clipping", manual selection. However, on the other hand, starring an article in GR is also a process of "manual selection" that collect valued stuff out of huge Internet, and save it as a part of personal knowledge database.

 

I'm not suggesting Evernote to scramble for a dying service, but for valuable personal data.

 

PS. sorry but I cant see how it can be accomplish by using IFTTT, or Zapier?

[jefito 5,589]

I suggest based on a fact that both GR and Evernote collect/storage valued data (articles, images, etc) from Internet for future reference.

Well, yeah, but so does DropBox, so does Pocket, Feedly, etc., etc., etc.

PS. sorry but I cant see how it can be accomplish by using IFTTT, or Zapier?

For IFTTT support, see this recipe: https://ifttt.com/recipes/9134. When use it, if you star an article in Google Reader, then it creates a note in your Evernote. Note content is modifiable in the recipe. I don't use it to fetch note article content, but I believe that you can.

 

I haven't tried to figure this out in Zapier, though I do use it to capture emails from specific sources to Evernote.

[bbattjer 1]

This request, while not very clear, is valid.

 

Likely what Meokey2000 is looking for is the ability to bulk import their JSON export of google reader starred items into Evernote, not support for marking articles as they're being read in Google Reader.

 

That's what I'm looking for.  I have a JSON export that's 23.7MB of data on the items that I marked as starred in Google Reader that's burning a hole in my hard drive to get imported into another service and evernote would be perfect.

[bbattjer 1]

Also, not sure how technical you are or if you have any technical friends, Meokey2000, but I just found this which I'm going to try:

 

https://github.com/kerchen/export_gr2evernote/blob/master/export2enex.py

 

Looks like it'll do exactly what we're both looking for!

[jefito 5,589]

This request, while not very clear, is valid.

 

Likely what Meokey2000 is looking for is the ability to bulk import their JSON export of google reader starred items into Evernote, not support for marking articles as they're being read in Google Reader.

Think that you might be right about that. I use Reader somewhat differently, so that's probably what I'm missing. Thanks for the clarification.

[meokey2000 0]

A possible solution to my request might be that Evernote, once users grant the access permission, run a background process to go though all starred articles in GR, and grab them back to users' Evernote notebooks. Users do not have to run any program on their side. Only 2 things users may have to do: grant Evernote the read access to GR and help Evernote to decide which notebook the articles should go to if Evernote can't determine.

 

@bbattjer, Thanks for the solution, it's awesome. I'm able to run the python program, but I believe many other users can't. So I prefer Evernote will be able to grab the starred items from GR for us.

 

@jefito, I'm sorry if I didn't make myself clear in the first request. I do suggest bulk export the starred items from GR and import them into Evernote automatically once users grant Evernote the access permission.

[whimbrel 0]

Also, not sure how technical you are or if you have any technical friends, Meokey2000, but I just found this which I'm going to try:

 

https://github.com/kerchen/export_gr2evernote/blob/master/export2enex.py

 

Looks like it'll do exactly what we're both looking for!

 

I found that script also, and gave it a run. So far as I can tell, though, it only imports the title and web link, without any of the text of the starred item. You can click on the web link, of course, but there's nothing available off-line and it's hard to know exactly what it was that you'd saved.

 

It's better than nothing, but it's pretty sub-optimal. 

[darrylo 3]

I found that script also, and gave it a run. So far as I can tell, though, it only imports the title and web link, without any of the text of the starred item. You can click on the web link, of course, but there's nothing available off-line and it's hard to know exactly what it was that you'd saved.

 

It's better than nothing, but it's pretty sub-optimal. 

 

In a similar technical vein, I wrote another ugly script to transfer starred notes into evernote:

 

https://github.com/darrylo/stars2evernote

 

This one tries to import the text/html, as well as any user tags.  I've used it to import 99%+ of my starred notes.  However, in addition to the things that Evernote can't handle (e.g,. embedded youtube videos), the one major thing it can't handle is tables -- evernote requires tables to have an explicit height specification, and most tables don't have that.  If you have a note with such a table, it'll likely import, but will never synchronize (evernote will forever banish it to the Unsynced Notes notebook, and evernote will also delete the table, after showing it to you once).  See the README.md for details.

[mattloflin 0]

I ran into this problem about a week ago and came up with a solution that is pretty neat.

 

I have ~2200 starred articles and I quickly hit the google reader email limit (sending it to my evernote email) of about 100 items per day. I wasn't about to use webclipper on all of them *manually*.

 

This method takes advantage of AppleScript iterating through Excel Rows, opening a Chrome Tab, using shortcut keys to use the web clipper, then going onto the next item.

 

Here are the steps I took:

1. Download your .json files via Google.com/takeout
2. Open the starred.json file and copy the xml into Excel into column A
3. Delete everything in Row 1 and enter the following values A1 = "URL" and B1 = "Find"
4. in B2 write this formula: =FIND("""href""",A16) and copy it to the entire column **this finds all the URLs**
5. Filter Column B (Find) to not include #Value
6. Copy and paste the values to a new worksheet **makes it easier to format**
7. Do a Find and Replace on '      "href" : "' (not including the single quotes) and replace it with "" (nothing)
8. Do another Find and Replace on '",' (not including the single quotes) and replace it with "" (nothing)
9. --- you should now be able to copy and paste any value in column A (other than A1) into Chrome and it takes you to the article ---
10. Install Evernote Web Clipper
11. Enable Evernote Keyboard Shortcuts via this Evernote Forum about Half way down by Mike Pacer (I used Command+Shift+E - if you use something else you'll have to update the script)
12. Open Apple Script and copy and paste this code

--set global delayset theDelay to 15--get row count from Excel based on used rangetell application "Microsoft Excel"	set theRowCount to count of rows of used range of active sheetend tell--Get URL From Opened Excel File, Mark it if successfulset theStart to 2set theEnd to 2275 --theRowCountrepeat with thisRow from theStart to theEnd	--display current count	tell me		activate		set theText to thisRow & " out of " & theEnd		display dialog theText as string buttons {"Ok"} default button 1 giving up after 1	end tell	--clear variables	set theRange to ""	set theUrl to ""	set theCell to ""		tell application "Microsoft Excel"		try			set theRange to "A" & thisRow			set theUrl to (get value of range theRange) as list						--Open a url in Google Chrome			tell application "Google Chrome"				activate				tell window 1 to make new tab with properties {URL:theUrl}			end tell						--Delay for the page to load			if theUrl contains "entrepreneur" then				delay 60 -- sometimes entreprenuer.com has ads			else				delay theDelay			end if						--use Evernote Web Clipper to clip the page			tell application "System Events"				keystroke "E" using {command down, shift down}				delay theDelay -- wait on clipper to prepare				keystroke return			end tell						--closes the first tab in Google Chrome (to prevent a lot of open Tab's)			delay theDelay -- Wait on sync			tell application "Google Chrome"				tell window 1					tell tab 1						delete					end tell				end tell			end tell						--Mark the current row as completed w/ "X"			set theCell to "B" & thisRow			set value of cell theCell to current date						--error handling to update Completed w/ Error statement		on error errText number errNum			set theCell to "B" & thisRow			set value of cell theCell to "ERROR " & errText & number & errNum					end try	end tellend repeatdisplay dialog "You're Done!" buttons {"Ok"} default button 1

 

Instructions on how to run:

1. Change theStart and theEnd to your First URL Row (probably 2) and your Last URL Row (You can uncomment theRowCount and it will enter your last used row from excel for you)
2. Change the delay to whatever you want... I have it set to 15 because some pages are slow and sometimes the web clipper is slow. I have a clause that changes it to 60 if it was entrepreneur.com as sometimes their were ads... but i think Evernote Web Clipper ignores those (not verified).
3. Open 2-4 blank chrome tabs ** the script above closes the first tab every iteration, so if you have 2-4 open it allows for the Web Clipper to have a buffer if not finished (e.g. you have 1-4, open a 5th URL, 1st URL closes and 5 becomes 4 4>3, 3>2, 2>1)
4. Run It

Notes about the Script:

1. Everytime it completes one a popup shows X out of Y (e.g. 20 out of 200) that disappears after a second
2. In Excel it logs the timestamp into Column B after each completion
3. When it is complete it prompts you with "You're Done!"
4. You can press stop and it will log it as an error before you start again just update theStart with the row number that has the error.

I ran it on all 2275 rows and it took it about 2 days (1 item takes at least 45 seconds so 2275 = 102375 seconds = 1760 minutes = 28.43 hours) 

 

 

Let me know if you have any questions!

[mattloflin 0]

Note 1: I had my default Web Clipper set to Notebook: Articles and Tags: AppleScript, Articles, Starred and when it imported it got most right but I had to do some bulk changes.

 

Note 2: I found a lot of duplicates and it might be because of the HREF thing... I haven't found a solution but you might spend 30min-1hr deleting duplicate notes.

 

Lastly - i'm using IFTTT.com to automatically create new notes everytime I star an article: https://ifttt.com/recipes/9134 - this only works as a gap until Feedly/Digg/etc. create a Google Reader replacement.

[chirmer 567]

Lastly - i'm using IFTTT.com to automatically create new notes everytime I star an article: https://ifttt.com/recipes/9134 - this only works as a gap until Feedly/Digg/etc. create a Google Reader replacement.

Emphasis mine, but does Feedly not fill this? They even port over your GR subscriptions and add them to your Feedly so that when GR is nuked by Google, you don't have to touch a thing.

[mattloflin 0]

from my understanding feedly currently uses GR backend to save articles and pull new articles - they haven't created their own "service" yet. Eventually Feedly or Digg will fill the gap. I still am going to save them in Evernote so I can have thing in a single repository along with bills, finance, taxes, work docs, etc.

[chirmer 567]

from my understanding feedly currently uses GR backend to save articles and pull new articles - they haven't created their own "service" yet. Eventually Feedly or Digg will fill the gap. I still am going to save them in Evernote so I can have thing in a single repository along with bills, finance, taxes, work docs, etc.

 

"Google announced today that they will be shutting down Google Reader. This is something we have been expecting for some time: We have been working on a project called Normandy which is a feedly clone of the Google Reader API – running on Google App Engine. When Google Reader shuts down, feedly will seamlessly transition to the Normandy back end. So if you are a Google Reader user and using feedly, you are covered: the transition will be seamless."

 

From Feedly's blog post on the subject: http://blog.feedly.com/2013/03/14/google-reader/

 

So, you can feel free to sync your account with Feedly - once that step is completed, you're done. Feedly does the rest.

 

[jefito 5,589]

I am working with Feedly now, though I am not as fond of their Android app as I am of the Web app.

[mattloflin 0]

ahh makes sense. I was grouping the july 1st end of GR with their API being ready for developers. Which makes me wonder why it is that if I star something in GR that it shows up in saved for later on Feedly - kind of makes me think they are just syncing until they finish their engine.

 

Either way - I still had to dump my old starred items to evernote for safekeeping and setup ifttt to automatically create new ones whenever I save them.

[darrylo 3]

I ran it on all 2275 rows and it took it about 2 days (1 item takes at least 45 seconds so 2275 = 102375 seconds = 1760 minutes = 28.43 hours) 

 

Not to lessen what you did, but there are much faster methods.

 

The method I gave converts the json to enex in a few seconds, and using the enex to import 5000+ notes into evernote takes under a minute.

[mattloflin 0]

Darrylo - No worries, I had looked at your github before doing mine and didn't want to just do the snippet of HTML included in the json file. Alot of my articles were the url and then a line or two of information. 

 

I guess the benefit between the two is time. Yours will knock it out in no time and only give you the summary they provide. Mine goes to the URL and does what web clipper does.