UPDATED - Seed of Language v1 to v2 Lesson Converter

[Swoka Ikran]

Seed of Language v2 came out recently, using a new INI-like lesson format that's not compatible with the older XML used by v1.xx.

This tool is a converter for the Seed of Language v1 lessons. I wrote it for personal use, so there may still be bugs in it, and it does not support lessons that use images (the images will be not be kept in these lessons, but sound and text will). Testing was performed using lessons from the v1 Eana Eltu exporter and the lessons available from SoL v1.9.2.

Spoiler: Changelog

v0.2.1:
* Bug fix: Lessons weren't being saved in UTF-8, causing issues with ì character.

v0.2:
* Bug fixes, and lots of 'em.
  * Fixed case sensitivity issues
  * Proper conversion of lessons where correct answer is not answer 1.
  * Fixed a bug where lessons missing the XML header line were rejected as invalid.
  * Blank lines in files are now ignored.
  * Several other minor fixes.

* Lesson folder enhancement suggested by Hufwe Atxur
* Command line support. You can now convert by specifying: SolConvert <input XML> <Output folder> on the command line.
* Multiple file support. Select multiple input files by using Ctrl or Shift.
* Major under-the-hood improvements.
  * Rewrote tag handler.
  * App now deals with malformed lessons better.
  * Significant code clean up.

Download it here: Download v0.2.1

Old versions: v0.1.1, v0.1.0

Supports Windows 2000/XP/Vista/7. Likely Linux and Mac compatible using WINE, but untested. Just extract and run. Usage is self-explanatory.

Comments (and bug reports) welcome.

[Swoka Ikran]

Updated to fix a bug reported in the Seed of Language topic.

It now handles questions with empty answer fields correctly.

[Hufwe Atxur]

This is great, so I don't need to rewrite all lessons! Works fine here, but for some reason it did not copy the audio-examples. Also it would be nice, if the program automatically creates a directory in the SoL-lesson-folder chosen with the lessons name as its name ("Select the Seed of Language lesson-folder").

[Swoka Ikran]

This is great, so I don't need to rewrite all lessons! Works fine here, but for some reason it did not copy the audio-examples.

Glad to see it works for you

Also it would be nice, if the program automatically creates a directory in the SoL-lesson-folder chosen with the lessons name as its name ("Select the Seed of Language lesson-folder").

Simple enough to change. When I update it next, I'll add this. It'd certainly make it easier to use.

[Blue Elf]

I just trying to convert lessons form Sol 1.9.2, but not every lesson can be converted. I tried four of them with this result:
- nl-5lv-311010.xml (WB: 5-letter verbs) - succeed
- nl-ae1-290910.xml (WB: Common 2 letter words)- error, file is not a valid lesson file (huh, it normally works in Sol 1.9.2)
- nl-ae2-051010.xml (WB: Three letter words) - same problem
- nl-objects-030910.xml (Objects) - suceed

What can be wrong? Even worse, I tried Objects lesson and it doesn't work well - it considers correct answers as wrong for some questions.

EDIT: found an error!! Old format uses <correntanswer> tag to point out what answer is correct. If id=1, it works. But if id is different, it goes wrong way as in new format correct answer must be ALWAYS on the first position (Ax_1).
Can you fix it?

[Swoka Ikran]

I just trying to convert lessons form Sol 1.9.2, but not every lesson can be converted. I tried four of them with this result:
- nl-5lv-311010.xml (WB: 5-letter verbs) - succeed
- nl-ae1-290910.xml (WB: Common 2 letter words)- error, file is not a valid lesson file (huh, it normally works in Sol 1.9.2)
- nl-ae2-051010.xml (WB: Three letter words) - same problem
- nl-objects-030910.xml (Objects) - suceed

Can you send me the "2 letter words", "Three Letter words", and "Objects" lessons rutxe? I can't download 1.9.2's lessons. You can either upload them somewhere and post a link, or email them to me (use the email in my forum profile).

The "Error: File is not a valid lesson file" is because the <numeliuti> tag has something wrong with it or is missing. Make sure it's there, and is in all lowercase. Also check that there is a <lessionData> tag (this should not be case-sensitive).

Case sensitivity of <numeliuti> will be removed in 0.2...it's a bug.

What can be wrong? Even worse, I tried Objects lesson and it doesn't work well - it considers correct answers as wrong for some questions.

EDIT: found an error!! Old format uses <correntanswer> tag to point out what answer is correct. If id=1, it works. But if id is different, it goes wrong way as in new format correct answer must be ALWAYS on the first position (Ax_1).
Can you fix it?

The lesson parser in this app is not the most robust thing in the world...so it's going to be hard to fix this. I'll take a look at it. Can I get the XML from Objects?

For v0.2 so far:
* Bug fixes, and lots of 'em.
  * Case sensitivity issue in lesson headers
  * Improper conversion of lessons where correct answer is not answer1
  * Case sensitivity issue in vocab headers
  * Minor fixes and improvements
* Lesson folder enhancement suggested by Hufwe Atxur
* A lot of internal clean up.

EDIT: OK. Thanks Blue Elf for the lessons The cause of the erroneous "Not a valid lesson" messages is the XML schema tag is missing. v0.2 will ignore this tag.

[Swoka Ikran]

Updated to v0.2.

@Blue Elf: Irayo nìmun for the lesson files You shouldn't have any trouble converting those now.

[Blue Elf]

Nìprrte' 
Meanwhile I have developed my own converter (simple ugly console app, but able to do batch conversion - no more than an hour of work), so I think at this time is no need of any converter  - I'll change all lessons into new format and send them to Hufwe Atxur for release.
But I'll try your converter too - mine only converts definition files, no additions related to file copying etc.

[Swoka Ikran]

Nìprrte' 
Meanwhile I have developed my own converter (simple ugly console app, but able to do batch conversion - no more than an hour of work), so I think at this time is no need of any converter  - I'll change all lessons into new format and send them to Hufwe Atxur for release.
But I'll try your converter too - mine only converts definition files, no additions related to file copying etc.

I added batch conversion to mine in v0.2 You can select multiple files using Control or Shift keys when you select the XML. My original (unreleased) version was a console app, but it did single files only.

[Blue Elf]

Good job, ma Ikran!
I tested your converter, seems to work well, but I found small problem - it sometimes changes ì to i (mine works correctly  )
For example, lesson "Numbers 1-16" is converted well, but "Conversational Words #1" not

[Swoka Ikran]

Good job, ma Ikran!
I tested your converter, seems to work well, but I found small problem - it sometimes changes ì to i (mine works correctly  )
For example, lesson "Numbers 1-16" is converted well, but "Conversational Words #1" not

Conversational words #1 works fine here, as does Numbers 1-16 and several others. I can't reproduce it (and I tried converting them several times).

What OS are you using?

[Blue Elf]

Windows XP, SP3

[Swoka Ikran]

Windows XP, SP3

Weird. Same thing I'm using.

Can you try this version please? http://swokaikran.skxawng.lu/files/solconv021-utf.zip It has some minor changes with the way it reads UTF-8, and now saves output in UTF-8.

Also, is your WinXP a regular English version, or an international/non-English one? (If it's international, there's a few bugs in Autoit that could be to blame).

[Blue Elf]

Tested on PC in my job (English XP), works well. I'll try also at home (Czech XP). But I expect problem is solved, as ì is correctly encoded into 2 bytes and file contains Byte Order Mark (yesterday's version didn't have it).

[Swoka Ikran]

Tested on PC in my job (English XP), works well. I'll try also at home (Czech XP). But I expect problem is solved, as ì is correctly encoded into 2 bytes and file contains Byte Order Mark (yesterday's version didn't have it).

Good to see it's working now. Let me know how it goes on the Czech XP.

If it works, I'll put in the first post as version 0.2.1.

[Blue Elf]

Seems to work correctly even on my Czech XP, so this release is hopefuly the last one

[Swoka Ikran]

Seems to work correctly even on my Czech XP

Good to hear I'll update the first post with it. Updated the first post with it.

this release is hopefuly the last one

Let's hope so