Thursday, July 1, 2010

New Parser - MM3 support (pre compendium)

The 4.0.89 parser is out. It handles all the released MM3 content, but until the compendium gets updated and I can give it more MM3 data to process, I won't say MM3 is fully supported.

It can be found at the usual address

Version History:
Note: to use the NPC Database features the Microsoft Office or the ACE drivers are required.
ACE drivers can be downloaded at the following address:

****************************Public Release: 7/1/2010 *****************
* General: Hopefully fixed an out of range exception when clearing some menu items

* Trap: Null perception check.

* MM3: Power style flag for ruleset formatting

* MM3: MM3 Action/Power format support

* MM3: Support for MM3 senses

* MM3: Support for different cascading caused by MM3 formatting

* Disease: Support for diseases based on skills other than Endurance

* MM3: Built MM3 style section of the fluff output routine to more closely match the MM3.

* NPC: Corrected a problem where an extra space was being stored with the stats when sending to the database.

* MM3: Support for ____ In Combat header of fluff. To be entered as Z ____ In Combat. For example Z Rot Grubs In Combat

* NPC: Fixed Power parse issue when a power includes range, but it isn't a range clause

* NPC: Allowed cascade from Origin to XP

* Scrape: Fixed Z4.gif image translation

* NPC Database: Compiled as DLL

* NPC Database: Fixed "Query too complex" error

* NPC Database: Added Encounter file Filter and processing

* NPC Database: Fixed Keyword search

* NPC Database: Added Filter settings for XP and Keywords

* NPC Database: Added Keywords and XP to NPC database

* NPC Database: Got Permutation function working

* NPC Database: Added GetAvailable

* NPC Database: Switched to access database

* NPC Database: First pass

* Trap: Zeph can't build trap files fix.

* Trap: Fixed traps where WOTC was sloppy and didn't include the expected fields.

* Trap: Countermeasures and triggers processed into different nodes. Trigger -> Initiative transistion allowed

* Catalog: Corrected bug created when feature only lists were added.

* PHB2: Requirement->Trigger cascade allowed

* Trap: Corrected a bug with multiple additional skills. Upgrade to Elite processed into different nodes.

* Powers: "Pact Boon" now recognized and functions the same as "Attack" and "Utility"

* Trap: Additional skill data processed into different nodes.

* Trap: Multiple Additional Skills fixed

* Trap: Perception data processed into different nodes.


Olodrin said...

Thank you!

Also, NPC database? What?

Zap said...

Thanks for the update. I'll give it a try at some point.

J said...

It ties in with the module workshop. I believe there's a video of it on the fgii site

Juan said...

Just a question... for compendium mode. Say we've scraped the D&D Insider Compendium for local use, which can be accessed at e.g. "http://localhost/dndinsider/compendium/" but no longer have a current insider account. Any way to parse to the local compendium?

J said...

No. I recommend keeping your subscription current.

Mortani said...

Great updates! It would appear the database process doesn't like module names with apostrophe's though.

J said...

No one likes sources with apostrophes!

Anonymous said...

Yeah, I think in general the parser doesn't like anything with an "interesting" high-order character in the title, since the nodes get named with those characters and that isn't acceptable xml.

Tenian, can I ask if the scraper stops at Dungeon 176 because the block format changed, or are the sourcebooks hardcoded? I'm trying to scrape NPCs from more recent adventures.

J said...

The compendium doesnt go past 176 currently. Dunno why..wotc getting lazy I assume

CopyMonkey said...

Thanks so much for continuing to work on this - much appreciated. I look forward to trying this post compendium update.

Czarisyn said...

I get "Compendium returned a login screen. Account validation failed" when scraping, any ideas on how to fix it?

J said...

Verify that you have entered your email address and password correctly?

Czarisyn said...

I get the error during the scrape, the email and password works fine but as its "Processing" it brings that error. I've seen it do it a lot after 20 or so items have been processed.

Olodrin said...

Hey J,

I think they posted the MM3 stuff into the compendium today. How did the parser do? Is it good to go, or should I wait for the next version?


J said...

It doesn't handle all the format changes. The HTML behind the stat blocks changed significantly.

Even the phr Riesling book required modifications as they've added a page reference after the links

Natedog said...

I am new to your great program and have run into a few strange issues. First, the new version didn't scrape some of the Compendium content that older versions did, specifically from PHB (Weapons, Armor, to name two).
Second, My group sees two dif ver of the PHB when linked to our DM. Some of us see an older ver while others see the new ver. I see the old when connected to him as a player, but the new when logged in as DM. The DM sees the new ver. Any idea why?

Natedog said...

Oh, I also meant to ask what is the optimal type of mod to create for player material: common or client type files? once we had all of the books, it slowed our game down to a crawl. players couldn't get maps from the DM for hours, etc.

J said...

The compendium is having issues with items. Go to the website and attempt to search by source and pick the Adventurers Vault. You'll see you get a whole mess of data from other sources.

For your module issue, it sounds like the host is using a common module. This will be used instead of the local copy when you are connected to that host.

Common modules are downloaded from the host. Large common modules. Will degrade performance. Fgii manages keeping them in sync if they change

Stray Neutrino said...

MM3 scrape is fine.
When parsing, the XML check throws an error "Name cannot begin with '>' character".

Since it's the first entry of the MM3 xml file, I imagine all entries will be broken.

J said...

The fact that the scrape completes does not mean it's fine. It just means it does not hit any fatal errors. The .89 version produces an invalid input file due to the changes made in the compendiums data structure.

Attempting to extract the data with that version of the parser is a complete waste of time

Stray Neutrino said...

@J - Does this mean that all the Monster Manuals are now, effectively, broken and unable to be parsed (since they changed the format across the board) ?

DM Drazulfel said...

Yep, I've been scraping and parsing all evening. I did 29 books, tonight, and MM3 is the only one that didn't work properly. The raw text file comes out looking like XML, and the parser just says no. But hey... 29 books! Great tool, thank you!