This is more a MWS question than a question about Gatherer Extractor.
To get correctly fullscans in MWS, you need to save all the pictures as cardname.full.jpg in the correct subfolder (whose name MUST be the set code for MWS) in the Pics folder.
Default parameters in Gatherer Extractor are set for Magic Workstation. The only thing you have to do is to choose the Magic Workstation pics folder path in Scan Options. Then you may download scans and the program will generate all the fullscan pictures in the correct subfolder of the MagicWorkstation pics folder, as the following format : Name.full.jpg
Thanks for your help,
I think I did that, but MWS is still not showing me the scans,
Here's a screenshot of my folder structure
Inside of those folders the card names look like this:
I haven't loaded a new "master base", I just downloaded scans with Gatherer Extractor... any ideas of what could be going on here ?
thanks for your screenshots. It really helps!
Your Pics folder shouldn't be in the Data folder, but directly in the Magic Workstation folder. The correct path should be, for example, for [4E] Abomination: .../Magic Workstation/Pics/4E/Abomination.full.jpg
the OCTGN export is a little tough because there's actually two xml files that are used -- the standard database file with all the card's stats/properties, and another one (the RELS file) which links the card image filenames to a GUID (which is an additional card property). If you know what the Open Packaging Conventions is, apparently that's how OCTGN's set files are constructed.
Is the programming language used for the extractor capable of generating unique IDs? That would make life a LOT easier.
Thanks Gaspare for your PM.
Everything is fine in the xml file except the following things :
— Booster packaging are not handled for the moment with Gatherer Extractor (I'm planning to add it very soon for personal use). Thus, the "packaging" table will not be missing in the xml generated by Gatherer Extractor for Octogon, but has to be set in the options, becaue set data will not be generated automatically.
— How Card IDs are generated. I mean, not the Multiverse_ID, but for example, id = 001d9638-76f4-4827-85d1-adfb1e4ef95e for Karn Liberated. Where does it come from ?! How to retrieve/reconstruct this ID ?!
— <set name="New Phyrexia" id="2566f33a-8472-4ee7-b37b-abbd24a1a5a5" gameId="A6C8D2E8-7CD8-11DD-8F94-E62B56D89593" gameVersion="2.0.4" version="3.1">
Where do you get all those IDs again (for set ID and game ID ?) Game version and version can be easily setable, but the IDs...
VB.NET can create unique IDs, but if I use those IDs for all the IDs required, won't it mess all Octogon up ?! Is that what you are talking about, brine ?
Packaging is not a problem, it's the same structure for every set, the only things that change are guids :). That can be done manually without problems.
Brine refers to the relationships file (which extension is namefile.xml.rels) which contains just relationships between cards defined in the xml file and pictures. I'll attach you an example of .rels file to clarify (forgot to do in yesterday's pm, shame on me :/)
the only id that change is the set id (first one). The "gameId" is fixed and refers to the id of mtg game definition, which is equal for all the sets and must not change. Also "gameVersion" must not be changed, unless a new game definition is released, but in that case there will be a patch that will update every set accordingly :), so no need to touch it too.
I tried to be as much clear as possible, if you have questions feel free to ask
OK. I have now all the material to code an exportation to Octogon2. Token cards (I'm assuming the 4 last cards, T1->T4 are token cards in the attached file you provided) will be missing.
I'll will PM you, Gaspare when the work is done for you to bêta-test the export routine, if you don't mind. You seem to be much more confortable with Octogon than I am (I personally don't use it).
For GUIDs, VB.NET can generate them with the native System.Guid.NewGuid method.
Theres a few specific conventions that we've been using in our set files. The mana symbols found in mana costs and rules text are surrounded by {} like {3}{R}. Converted mana costs don't do this. Also, the cards color is written out in words, like green, or multicolor green blue.
Also, guids have dashes in the XML file but no dashes in the rels, and the rels guids always start with a C.
Hybrid color combinations are {W/U} {B/R} {R/G} {U/B} {G/W} {G/U} {B/G} {W/B} {U/R} and {2/W} {2/U} {2/B} {2/R} {2/G} so for example reaper king will have a cost {2/W}{2/U}{2/B}{2/R}{2/G}, while deus of calamity will have a cost like {R/G}{R/G}{R/G}{R/G}{R/G}
Split cards names are in the form name // name and there are 2 copies for each one, with 2 different guids (example: fire // ice (fire) and fire // ice (ice), although the (fire) and (ice) parts are not displayed in the editor)
flipped cards are treated as single normal cards. No need to touch them for now
Multicolor cards color (either hybrid or standard multicolor) is described as "Multicolor color1 color2 color3... etc."
So taking again the examples i made in my previous post, deus of calamity color is "Multicolor Red Green", while reaper king color is "Multicolor White Blue Black Red Green"
Private Mod Note
():
Rollback Post to RevisionRollBack
On the Internet you can be anything you want.
It's strange that so many people choose to be stupid...
There's a few other nitpicky details about our set conventions :). I plan on using your extractor to update the oracle text on all the sets though, so there will be consistency with whatever you're able to do (so don't worry about any that are too complicated).
Due to XML's limitations, certain characters cannot be used:
- quotation marks have to be replaced with "
- ampersands (&) are replaced with +
- Line breaks are supported by OCTGN's deck editor, but must be indicated using - The "Ae" character for the Aether cards, I'm not 100% sure what you use to indicate them, I've just been entering them with the normal a and e characters.
Split cards are separated with the line breaks, so that their card stats each show up on a different line. For example, Fire//Ice would be typed as Fire
//
Ice and each of its other stats would be separated in a similar way.
Planeswalkers have each ability separated with a line break, and the rules text looks like [+2]: Each player draws a card. Also, a Planeswalker's loyalty goes into the power/toughness property.
Hybrid cards use the same color wording as gold cards - Just use Multicolor Blue Red. We do it like this so that when a player wants to sort by color, all the hybrid and multicolor cards are separated from the monocolored cards.
Finally, the MTG plugin for OCTGN recognizes the following card properties. It's important that the property names are exactly written as shown:
- Name
- Cost
- Converted Mana Cost
- Type
- Subtype
- Rarity
- Rules
- Flavor
- Power
- Toughness
- PT Box (Note: This is the power and toughness put together, and looks like 3 / 2 for creatures. Planeswalker loyalty also uses this property, there is no seperate loyalty property.)
- Artist
- Number
- MultiverseId
If a card doesn't have a property (such as instants not having power or toughness), it doesn't really matter if the extractor keeps the property and makes its value blank, or ignores the property completely. Ignoring the value would save some filesize, but its not really that big a deal if it makes the coding difficult.
There are two more properties, "Tags" and "Faction", but don't worry about those.
Brine, Gaspare, I sent you a PM where you can find a link to the new version of Gatherer Extractor. The exportation to OCTGN2 has been implemented. May you please test it, and report here all the bugs you may find. Please tell me if the generated xml & xml.rels are OK (dual cards, color, format etc.)
Hi, i've tried and successfully exported commander set as a first test, but i found an issue in the .rels file. Last 20 card names are wrong, some numbers just repeat 4 times. Probably it's only a problem of commander set, since it's not on magiccards.info yet, because I've tested other large sets (M11 and Mirrodin) and the issue hasn't repeated.
The export works, but the first line of data is wrong.
[noparse]
<card id="62e8565e-a946-4f25-a4f6-8757e238dbc1" Name="Artisan of Kozilek">
[/noparse]
should be
[noparse]
<card name="Artisan of Kozilek" id="62e8565e-a946-4f25-a4f6-8757e238dbc1">
[/noparse]
I guess the problem is related to "Name" that should be "name", all in lower case. Infact when i try to install the set on octgn, it says "The 'Name' attribute is not declared", and if i change all "Name" instances with "name", the set installs correctly
Another problem: "Color" property is present but empty on all cards, besides multicolor cards, where it's present and correct.
All the bugs you reported should be fixed in the new release (Name -> name, color fixed). May you please check it quickly ? http://www.mediafire.com/?3akc6mtapnar2
The first problem you mentioned, Gaspare, was due to the fact that the Commander page was missing on magiccards.info. Now the page is available and the extraction is clean.
I had some concerns about making these extractors public, as other users may try to export their own sets instead of downloading ours. Since the GUIDs won't match between the sets, there's going to be conflicts in the program and players won't be able to see each other's cards.
I'm going to create and upload a database of sorts which associates all GUIDs we currently have registered in OCTGN to the cards and sets they belong to. If you have the extractor connect to that webpage like it does to gatherer, we should be able to get the GUIDs to match with any extracted set that already exists in our database. Would this be easy to implement as well?
I see your problem. But your suggestion doesn't seem to solve the problem.
It can harmonize all the GUIDs for all previous cards, but not for new ones : when a new set is released and people try to extract files for OCTGN 2 from Gatherer Extractor, if the GUID page on your webpage has not been updated, new GUIDs will be created. That's pretty fine for you to create your own patch (you don't have to worry about GUIDs anymore), but for common user, his or her own patch will differ from yours due to different GUIDs.
A suggest you to inform users (with a reminder somewhere in OCTGN) to use only YOUR set-packs to ensure a clean database. and you may use Gatherer Extractor to create your packs.
We have other developers who have asked for this kind of database for things like deck converters, so I plan on keeping it updated as often as possible.
We typically release incomplete sets and incremental patches as new sets are spoiled, so we'll likely have the entire GUID database for new sets online by the time the set arrives on gatherer. This actually makes it easier on us as well, as we used to type out the entire card information by hand as the cards are revealed (since we didn't have good extractors). Now we can simply record the name and GUID, and have the extractor fill in the rest once the set's released.
Private Mod Note
():
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Thanks for your help,
I think I did that, but MWS is still not showing me the scans,
Here's a screenshot of my folder structure
Inside of those folders the card names look like this:
I haven't loaded a new "master base", I just downloaded scans with Gatherer Extractor... any ideas of what could be going on here ?
Your Pics folder shouldn't be in the Data folder, but directly in the Magic Workstation folder. The correct path should be, for example, for [4E] Abomination: .../Magic Workstation/Pics/4E/Abomination.full.jpg
I hope it will work for you now
It's strange that so many people choose to be stupid...
But could you please link me an example file of an Octogon2 xml file please ?! I'll add the option as soon as possible.
Thanks for looking into it
It's strange that so many people choose to be stupid...
Is the programming language used for the extractor capable of generating unique IDs? That would make life a LOT easier.
Everything is fine in the xml file except the following things :
— Booster packaging are not handled for the moment with Gatherer Extractor (I'm planning to add it very soon for personal use). Thus, the "packaging" table will not be missing in the xml generated by Gatherer Extractor for Octogon, but has to be set in the options, becaue set data will not be generated automatically.
— How Card IDs are generated. I mean, not the Multiverse_ID, but for example, id = 001d9638-76f4-4827-85d1-adfb1e4ef95e for Karn Liberated. Where does it come from ?! How to retrieve/reconstruct this ID ?!
— <set name="New Phyrexia" id="2566f33a-8472-4ee7-b37b-abbd24a1a5a5" gameId="A6C8D2E8-7CD8-11DD-8F94-E62B56D89593" gameVersion="2.0.4" version="3.1">
Where do you get all those IDs again (for set ID and game ID ?) Game version and version can be easily setable, but the IDs...
VB.NET can create unique IDs, but if I use those IDs for all the IDs required, won't it mess all Octogon up ?! Is that what you are talking about, brine ?
guids are generated randomly using sites like this http://www.guidgenerator.com/online-guid-generator.aspx
Brine refers to the relationships file (which extension is namefile.xml.rels) which contains just relationships between cards defined in the xml file and pictures. I'll attach you an example of .rels file to clarify (forgot to do in yesterday's pm, shame on me :/)
About this line
<set name="New Phyrexia" id="2566f33a-8472-4ee7-b37b-abbd24a1a5a5" gameId="A6C8D2E8-7CD8-11DD-8F94-E62B56D89593" gameVersion="2.0.4" version="3.1">
the only id that change is the set id (first one). The "gameId" is fixed and refers to the id of mtg game definition, which is equal for all the sets and must not change. Also "gameVersion" must not be changed, unless a new game definition is released, but in that case there will be a patch that will update every set accordingly :), so no need to touch it too.
I tried to be as much clear as possible, if you have questions feel free to ask
It's strange that so many people choose to be stupid...
I'll will PM you, Gaspare when the work is done for you to bêta-test the export routine, if you don't mind. You seem to be much more confortable with Octogon than I am (I personally don't use it).
For GUIDs, VB.NET can generate them with the native System.Guid.NewGuid method.
Thanks for all your suggestions !
I'll be eager to test the routine, i'll wait for the PM impatiently
Thanks for your support
It's strange that so many people choose to be stupid...
Also, guids have dashes in the XML file but no dashes in the rels, and the rels guids always start with a C.
I would need further information about :
— Colors of hybrid cards ?
— How are splits cards handled ?
— What about flipped cards ?
Split cards names are in the form name // name and there are 2 copies for each one, with 2 different guids (example: fire // ice (fire) and fire // ice (ice), although the (fire) and (ice) parts are not displayed in the editor)
flipped cards are treated as single normal cards. No need to touch them for now
It's strange that so many people choose to be stupid...
But is the color written out in words, like hybrid red green for deus of calamity?
So taking again the examples i made in my previous post, deus of calamity color is "Multicolor Red Green", while reaper king color is "Multicolor White Blue Black Red Green"
It's strange that so many people choose to be stupid...
Due to XML's limitations, certain characters cannot be used:
- quotation marks have to be replaced with "
- ampersands (&) are replaced with +
- Line breaks are supported by OCTGN's deck editor, but must be indicated using
- The "Ae" character for the Aether cards, I'm not 100% sure what you use to indicate them, I've just been entering them with the normal a and e characters.
Split cards are separated with the line breaks, so that their card stats each show up on a different line. For example, Fire//Ice would be typed as Fire // Ice and each of its other stats would be separated in a similar way.
Planeswalkers have each ability separated with a line break, and the rules text looks like [+2]: Each player draws a card. Also, a Planeswalker's loyalty goes into the power/toughness property.
Hybrid cards use the same color wording as gold cards - Just use Multicolor Blue Red. We do it like this so that when a player wants to sort by color, all the hybrid and multicolor cards are separated from the monocolored cards.
Finally, the MTG plugin for OCTGN recognizes the following card properties. It's important that the property names are exactly written as shown:
- Name
- Cost
- Converted Mana Cost
- Type
- Subtype
- Rarity
- Rules
- Flavor
- Power
- Toughness
- PT Box (Note: This is the power and toughness put together, and looks like 3 / 2 for creatures. Planeswalker loyalty also uses this property, there is no seperate loyalty property.)
- Artist
- Number
- MultiverseId
If a card doesn't have a property (such as instants not having power or toughness), it doesn't really matter if the extractor keeps the property and makes its value blank, or ignores the property completely. Ignoring the value would save some filesize, but its not really that big a deal if it makes the coding difficult.
There are two more properties, "Tags" and "Faction", but don't worry about those.
Lots of work to do ! I'll keep in touch with you guys. I will PM you when the OCTGN Exportation will be coded in order to test it.
About Packaging, I have only an example for Boosters. May I have an example for starters ?
The export works, but the first line of data is wrong.
should be
I guess the problem is related to "Name" that should be "name", all in lower case. Infact when i try to install the set on octgn, it says "The 'Name' attribute is not declared", and if i change all "Name" instances with "name", the set installs correctly
Another problem: "Color" property is present but empty on all cards, besides multicolor cards, where it's present and correct.
It's strange that so many people choose to be stupid...
All the bugs you reported should be fixed in the new release (Name -> name, color fixed). May you please check it quickly ?
http://www.mediafire.com/?3akc6mtapnar2
The first problem you mentioned, Gaspare, was due to the fact that the Commander page was missing on magiccards.info. Now the page is available and the extraction is clean.
I'm going to create and upload a database of sorts which associates all GUIDs we currently have registered in OCTGN to the cards and sets they belong to. If you have the extractor connect to that webpage like it does to gatherer, we should be able to get the GUIDs to match with any extracted set that already exists in our database. Would this be easy to implement as well?
It can harmonize all the GUIDs for all previous cards, but not for new ones : when a new set is released and people try to extract files for OCTGN 2 from Gatherer Extractor, if the GUID page on your webpage has not been updated, new GUIDs will be created. That's pretty fine for you to create your own patch (you don't have to worry about GUIDs anymore), but for common user, his or her own patch will differ from yours due to different GUIDs.
A suggest you to inform users (with a reminder somewhere in OCTGN) to use only YOUR set-packs to ensure a clean database. and you may use Gatherer Extractor to create your packs.
We typically release incomplete sets and incremental patches as new sets are spoiled, so we'll likely have the entire GUID database for new sets online by the time the set arrives on gatherer. This actually makes it easier on us as well, as we used to type out the entire card information by hand as the cards are revealed (since we didn't have good extractors). Now we can simply record the name and GUID, and have the extractor fill in the rest once the set's released.