Is there any kind of a topic limit?

[EightBitz]

Originally Posted by Acenoid

Like he wrote in the 1st post: he reads the csv and creates an RW XML for import
probably it is being made especially for this task I wont be of much use, however I agree, it could be great to create all kinds of lists / topics / things via csv and import them.

Let's see what EightBlitz says

I'm doing my best to make the code easily adaptable, but you'll still have to be familiar with PowerShell scripting. That being said, there are a few pending issues.

1. There is currently a bug with RealmWorks where tag values assigned to snippets are instead assigned to topics. I'm told by Farling that this is a known bug which LWD says it's working on.

2. The CSV file I'm working is from pathfindercommunity.net, but I've had to clean it up a bit.

3. There's the issue with the view size limitation. I have yet to modify the code for the suggested workaround.

4. The CSV file has a column for a plain-text description and one for an HTML-formatted description. The formatted description is obviously preferable, but is giving me issues. At least one spell (#134 of ~2700) is giving me an error on import. I say at least one, because it stops there and doesn't go beyond.

5. I've had to modify the tags in the Pathfinder structure to add spell descriptors, schools and subschools. So if I distribute the code, I'd have to distribute it with my Pathfinder structure and my sanitized version of the CSV file, otherwise errors.

Once I get a clean, flawless import, I'll be willing to share. Right now, my copy of RW is supposedly deleting the imported spells so I can have a clean realm to start testing things again.

Except for the import error, all these things have workarounds. For example, I can clone the Spell category, and in the clone, replace the tag snippets with labeled text snippets. Or stick information in the annotation fields instead of assigning tags.

But that one import error, I gotta find out what's going on there, and right now, I'm clueless. I've submitted a support request.

[HippyCraig]

Awesome looking forward to seeing the final result.

[daplunk]

Yeah this is bloody exciting.

[Farling]

Originally Posted by EightBitz

5. I've had to modify the tags in the Pathfinder structure to add spell descriptors, schools and subschools. So if I distribute the code, I'd have to distribute it with my Pathfinder structure and my sanitized version of the CSV file, otherwise errors.

There is no need to modify the tags. All of these are already in the default Pathfinder structure.

I've got all the 4 CSV loading quite happily except for the tags issue. I decided to use the basic text instead of the HTML-ised text. I don't know if the HTML-ised text is from the raw PRD or someone has added from another source.

My solution uses XSLT scripts rather than PowerShell scripts.

[EightBitz]

Descriptors in the spreadsheet that were not in the Pathfinder structure: Disease, Draconic, Meditative, Phantasm, Ruse, Travel

Spell Levels: Adept 0, 1, 2 and 3
Antipaladin 5 and 6
Bloodrager 5 and 6

Spell ranges: Several.

[EightBitz]

Originally Posted by Farling

There is no need to modify the tags. All of these are already in the default Pathfinder structure.

I've got all the 4 CSV loading quite happily except for the tags issue. I decided to use the basic text instead of the HTML-ised text. I don't know if the HTML-ised text is from the raw PRD or someone has added from another source.

My solution uses XSLT scripts rather than PowerShell scripts.

I apologize if I'm a bit cranky today, my brother set me off by bringing up an old argument that I thought had been settled. Anyway ...

I'm wondering if we're using different spreadsheets or if maybe XSLT works differently than Powershell, or at least the way I'm using Powershell.

What I'm doing with tags, for example, is that when there's a value in "School" or "Subschool" (etc) in the spreadsheet, I search for that text in the tag domain, and I assign the corresponding tag_id for whatever is found. If there is no matching text, the tag_id ends up being an empty value, so the resulting value in the export file becomes:

tag_id=""

When I try to import that, it throws an error.

So then, I can do two things. I can open the CSV file in Excel, click on the "Data" tab, then click the "Filter" option.

That makes each header a drop-down list. So I can click the drop-down list for school, subschool, descriptor or range, and see all the values that exist in each column.

Or, I can tell Powershell that whenever it finds an empty tag_id, append the tag name to a file. When it's all done, I can open that file and see all the tags that weren't recognized as existing in the structure file.

I'm wondering if with XSLT, if it doesn't recognize a tag, it doesn't even try to assign a value, so you never get an error, and you end up with tags that are just missing.

I don't know for sure, but if we're using the same spreadsheet and the same software, there has to be something to explain the disparity of our experiences, and the only thing that's different (so far as I know) is Powershell vs. XSLT.

[Farling]

Originally Posted by EightBitz

I apologize if I'm a bit cranky today, my brother set me off by bringing up an old argument that I thought had been settled. Anyway ...

I'm wondering if we're using different spreadsheets or if maybe XSLT works differently than Powershell, or at least the way I'm using Powershell.

What I'm doing with tags, for example, is that when there's a value in "School" or "Subschool" (etc) in the spreadsheet, I search for that text in the tag domain, and I assign the corresponding tag_id for whatever is found. If there is no matching text, the tag_id ends up being an empty value, so the resulting value in the export file becomes:

tag_id=""

When I try to import that, it throws an error.

So then, I can do two things. I can open the CSV file in Excel, click on the "Data" tab, then click the "Filter" option.

That makes each header a drop-down list. So I can click the drop-down list for school, subschool, descriptor or range, and see all the values that exist in each column.

Or, I can tell Powershell that whenever it finds an empty tag_id, append the tag name to a file. When it's all done, I can open that file and see all the tags that weren't recognized as existing in the structure file.

I'm wondering if with XSLT, if it doesn't recognize a tag, it doesn't even try to assign a value, so you never get an error, and you end up with tags that are just missing.

I don't know for sure, but if we're using the same spreadsheet and the same software, there has to be something to explain the disparity of our experiences, and the only thing that's different (so far as I know) is Powershell vs. XSLT.

I have an <xsltherwise> clause for when there is no match, so that I can set the "Other" tag and then put the string into the <annotation> child element.
(The emoji is supposed to be a colon followed by the word 'otherwise'.)

[EightBitz]

Originally Posted by Farling

I have an <xsltherwise> clause for when there is no match, so that I can set the "Other" tag and then put the string into the <annotation> child element.
(The emoji is supposed to be a colon followed by the word 'otherwise'.)

Ah. I just went ahead and either entered the tags or corrected the spreadsheet. For example, for range, "Medium (100 ft. + 5 ft./level)" was entered about 5 or 6 different ways. Sometimes with different spacing around the +, sometimes with "levels" instead of "level", sometimes with "feet" instead of "ft.". I tried to make things as consistent as I could on both sides.

I'm still having an issue with Detect Evil. I'm getting it to load now, but the table headers on the last table are shifted one cell to the right. I don't see anything obvious in the HTML that would cause that. Again, not sure if this is a Realm Works bug, or if the HTML is still wonky. All I can say is that a few other tables I looked at seem OK.

[kbs666]

Originally Posted by EightBitz

I'm still having an issue with Detect Evil. I'm getting it to load now, but the table headers on the last table are shifted one cell to the right. I don't see anything obvious in the HTML that would cause that. Again, not sure if this is a Realm Works bug, or if the HTML is still wonky. All I can say is that a few other tables I looked at seem OK.

Can you post the HTML code here? I could take a look and see if I catch an issue.

[EightBitz]

Originally Posted by kbs666

Can you post the HTML code here? I could take a look and see if I catch an issue.

This includes the basic HTML from the spreadsheet, the resultant XML that my script creates, a screenshot of what the basic HTML looks like in IE (with everything properly lined up), and a screenshot of what it looks like after importing the XML into RW, with the headers shifted to the right.