moratnz / makepdfbook Goto Github PK

So we can close the div in the chapterinfo template and still have the styling work

Hopefully these can be included as transcluded pages

e.g., group all the 'conventions - ' sections together

There's big wacks of javascript that MakePdfBook uses that's being manually copied into MediaWiki:Common.js in installs we're using MakePdfBook in.

That's dumb, It should be included as a resource in MakePdfBook.

Refactor things to use getBooks() rather than making lots of separate DB calls

It's currently alphabetical

These are currently global variables being set in LocalSettings.

They should be config items being set in extension.json.

Move them, and clean up references to them

Need to account for US letter and A4

Links in the pdf are rendering as relative links (as they should within the wiki), which breaks in a pdf.

For links that are internal to the pdf, it'd be great to work out how to get them to render as actual internal pdf links. For links that are actually external to the pdf, they should be prefixed with the wiki host name.

And the handbook page shouldn't be included in the chapter set that's sent to things like the pdf generator.

options include giving the handbook page a magic sort key, e.g., 'handbook', or 'contents', or move the stuff that's currently held on the handbook page into the category page.

As part of the work to automate the navigation sidebar in use in the wikis using this extension, it would be useful to be able to hand the font end a json object describing the handbooks in the wiki and their contents.

An example of the desired data structure:
[
{
namespace: "Armored_Combat",
label: "Armored Combat",
webVersion: "/index.php/Armored_Combat:Handbook",
pdfVersion: "/index.php/Special:MakePdfBook?category=Armored_Combat_Handbook",
chapters: [
[
"Introduction",
"/index.php/Armored_Combat:Introduction"
],
[
"Change Log",
"/index.php/Armored_Combat:Change_Log"
],
[
"1. Combat Authorization Requirements",
"/index.php/Armored_Combat:Combat_Authorization_Requirements"
],
[
"2. Rules of the Lists",
"/index.php/Armored_Combat:Rules_Of_The_Lists"
],
[
"3. Conventions of Combat",
"/index.php/Armored_Combat:Conventions_Of_Combat"
],
[
"4. The Use of Weapons and Shields",
"/index.php/Armored_Combat:Use_Of_Weapons_And_Shields"
],
[
"5. Acknowledgement of Blows",
"/index.php/Armored_Combat:Acknowledgement_Of_Blows"
],
[
"6. Armour Requirements",
"/index.php/Armored_Combat:Armor_Requirements"
],
[
"7. Weapon Standards",
"/index.php/Armored_Combat:Weapons_Standards"
],
[
"8. Siege Combat",
"/index.php/Armored_Combat:Siege_Combat"
],
[
"9. Authorization of Marshals",
"/index.php/Armored_Combat:Procedures_For_The_Authorization_Of_Marshals"
],
[
"10. Marshaling Wars",
"/index.php/Armored_Combat:Procedures_For_Marshaling_Wars"
],
[
"11. Marshaling Requirements",
"/index.php/Armored_Combat:Marshaling_Requirements"
],
[
"12. Combat Injury Procedures",
"/index.php/Armored_Combat:Combat_Injury_Procedures"
],
[
"13. Marshaling on the Field",
"/index.php/Armored_Combat:Guidelines_For_Marshaling_On_The_Field"
],
[
"14. Combat Authorization Procedures",
"/index.php/Armored_Combat:Combat_Authorization_Procedures"
],
[
"15. Equipment Inspection Guidelines",
"/index.php/Armored_Combat:Equipment_Inspection_Guidelines"
],
[
"16. Experimental Weapons and Materials",
"/index.php/Armored_Combat:Experimental_Weapons_And_Materials_Procedures"
],
[
"17. Marshal Responsibilities, Chain of Command, and Reporting",
"/index.php/Armored_Combat:Marshal_Responsibilities_Chain_Of_Command_And_Reporting"
],
[
"18. Grievances and Sanctions",
"/index.php/Armored_Combat:Procedures_For_Grievances_And_Sanctions"
],
[
"Glossary",
"/index.php/Armored_Combat:Glossary"
]
]
}
]

The namespace bits included in MakePdfBook are potentially of wider interest than the pdf generation bits.

Look at spinning them off into their own extension and publishing it

Will use H1 headings as Chapter titles instead (can remove {{DISPLAYTITLE:}} which is tripping security stuff in Lochac live)

In testing I've observed:

• fetching all content with BookSet->fetchChapters and BookSet->fetchTitlepages results in the right chapters in the right place.
• fetching single book's content with Book->fetchChapters & Book->fetchTitlepages results in the same content being put into multiple books.

The difference between these is that the latter has an additional WHERE statement in its DB call; that's about it.

For the moment I'm sidestepping this issue by doing all content loading with BookSet->fetch..., but it'd be nice to sort this so content can be loaded in a more targeted fashion.

Add title page creation links to the output of generateCategoryListPage, such that if a title page for category foo isn't found, a redlink for foo_titlepage?create is inserted in the table there

Feature addition:

Add option to render command to create an archive copy of the current pdf

The archive should be named something like <category_name>-<tag>.pdf where tag is either a date or version number.

It should be stored into the appropriate place in the file tree, and registered appropriately in the DB.

Parameters:

generateArchive=tag: here the specified tag gets put into the archive file.

Stretch goal:
Add dynamic pages to show all the historical versions of the handbooks.

We're currently applying inline styles to the pdf icons to avoid the delightfully named FOUT (flash of unsettled text).

This is somewhat filthy.

The correct solution is probably to include an appropriately licensed SVG in the extension images.

To avoid draft content from being indexed by google, we should implement 'noindex' meta tags on all draft pages.

This should be a very short hook that can be attached anywhere we have access to the Output object with $out->addMeta()

The result meta tag we're after is:
<meta name="robots" content="noindex, nofollow" />.