Import txt files

[Henrik in Oslo]

Time is approaching for me to import my still active Supermemo 7
learning DB (25 000 items)into Mnemosyne. My ver of SM exports Q/A
text WITH NO LEARNING INFORMATION. I have tested import to Mnem 1.2.2
and it works fine. In SM I am able to extract files depending on their
difficulty ("e-factor").
To establish some learning history for this DB in Mnemosyne I will
1) set the date of the computer back to jan 2005, then import the
easiest items (e-factor 3.0 or higher) into Mnemosye with cat-name e.g
"30" or similar
2) "Auto Learn" imported items
3) Turn computer calendar to july 2005, import e-factor 2.9 cat-name
"29" etc. Learn all.
4) etc
5) etc
6) For the hardest ones (e-factor 1.3 - 1.6) I will consider actual
learning with correct grading, still adjust calendar to emulate
original SM learning history.

BUT! in beta 2.9 I do not find import txt, only import mnem-files and
xml-files.
1) Will txt import be available?
2) Any othe suggestion of porting from SM7 (1995!!!) to mnemosyne 2.0?

Best regards

[Peter Bienstman]

Hi Hendrik,

txt import is not yet in the betas, but it should be in the final version. It
should not be too hard to write, but there's quite a lot on my plate at the
moment :-)

Cheers,

Peter

[Peter Bienstman]

Actually, no need to wait for me until I implement this.

Just temporarily install 1.x, create a new 1.x database, import your stuff
there, then install 2.0 and import that 1.x database there.

Since 1.x and 2.x use different directories, there should be no interference
(although backups never hurt :-) )

Peter

[Henrik in Oslo]

Thanks Peter!
Action ongoing :-) (with 25 000 items subdivided according to
difficulty, it takes time
- future request: possibility of Adding e-factor (or similar) to Q/ A
(and other imports)
/ Henrik

[Peter Bienstman]

On Sunday 22 Jan 2012 01:35:07 Henrik in Oslo wrote:
> Thanks Peter!
> Action ongoing :-) (with 25 000 items subdivided according to
> difficulty, it takes time

OK, good luck!

> - future request: possibility of Adding e-factor (or similar) to Q/ A
> (and other imports)

I don't think it's a good idea to pollute the generic txt import with this,
but a dedicated Supermemo importer would certainly be a good idea.

Peter

[Roel De Coninck]

Henrik,

I planned on writing a SuperMemo importer that takes into account the learning data.  You'd have to export your SM collection as XML instead of text.

Unfortunately, I didn't advance at all the last months :-(

I'm also running SM next to Mnemosyne, and I'd love to ditch SM. So maybe I have to push this project a bit up my priority list...

How long would you be willing to wait?  Let's say that I will try to progress on this quest in February.  If I couldn't make time, I will have to face it and give up.  Would that be ok?

roel

[Henrik in Oslo]

Hi Roel, my hero
I have waited for an SM replacement for 10 years, so I have become a
patient man :-)

My SM is ver 7.5 Feb 20 1995. The text output looks like this:


Q: E-B 2000 phil: Who wrote "The Concept of Mind" (1949)
Q: // Gil____ R____
A: Gilbert Ryle

Q: E-B 2000 phil: "the most distinctive feature of ANALYTIC
Q: PHILOSOPY is its emphasis upon the role that _______ plays in
Q: the creation and resolution of philosophical problems"
A: language

Q: E-B 2000 phil: Who said "Philosophy is a battle against the
Q: bewitchment of the intelligence by means of language."
A: Ludwig Wittgenstein

In other words: My version of SM does NOT export learning data :-(
In SM I am able to extract txt files depending on their difficulty ("e-
factor").
So one learning database in SM I can slit up int let us say 10
different txt-files from "Very easy" (e-factor ~3) to "leeches" (e-
factor ~1.3)

Inside Mnemosyne i will have to twist and tweek to slowly build a
renewed learning history: Establish categories according to easiness
determined by SM e-factor. For the first 6-18 months in I would spend
most of the repetition-time on the hard ones and auto-grade the
easiest (press 5-key until complete).


OR, since you are probably the fortunate user of an SM version that
exports learning history data, we could check whether your version
could "import from SM 7.5" . In other words: we would use your SM
version as a value-adding bridge between my SM ver 7.5 and Mnemosyne
2. That would make me a very happy man.

Best regards from a patient man
Henrik

[Murray James Morrison]

Nice cards, Henrik :-)

-murrayjames

Sent from my HTC phone

--
You received this message because you are subscribed to the Google Groups "mnemosyne-proj-users" group.
To post to this group, send email to mnemosyne-...@googlegroups.com.
To unsubscribe from this group, send email to mnemosyne-proj-u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mnemosyne-proj-users?hl=en.

[Roel De Coninck]

Henrik,

I have a SM version 2008, and in the first place I'll try to import the XML exports I can create, containing all learning data. 

I see two solutions for your collection:

1) as you propose, trying to open your collection in a recent version of SM and export as XML.  I do not know if it's possible to go directly from 7.5 to 2008, but I guess it must always be possible to use intermediate versions untill we arrive at a version with XML export. 

2) start from the exports by e-factor you can produce, and then import them directly in Mnemosyne with a corresponding easyness and interval (I still have to check which parameters are to be set to reflect the learning history of SM).  This would spare you the tweaking and would give you more accuracy.

I think you can still get a demo version of SM 2008 that works for a limited number of repetitions.  I don't think there's any other limitation on the demo, so you could try to download it and open (a backup of) your current collection.  

Do you have anything else then text cards?  You use categories?  

I'm glad you're patient, it's motivating if other users could use the SM import functionality.  I'll do my best :-)

roel

[Gnome]

Hay Henrik, if you don't have access to SM2008 you could take a look
at the XML format created by mnemnosyne 1.x, then try to mimic the
format for your text files using regular expression in an text editor.
By that way you could insert learning data in the XML created.