Thanks, Mark. I'm working on a perl script to manage this (since I don't
know JavaScript, but vaguely remember perl). It would seem that it will be
necessary to rely on the willingness of the wikipedia editors to maintain
their page in standard structure. And every wikipedia page will be
The original request was to split off the WP article into separate tids. If
you replace the BulletList section of the code I previously provided with
this:
function BulletList(items)
local buffer = {}
local name = ""
local cnt = 0
for _, item in pairs(items) do
buffer = {}
Ciao Mark S.
I found it interesting. Though I shouldn't pretend I really understand the
lua code.
I had a look at WikiPedia HTML. Its a lot cleaner than a lot. I wouldn't
say it was easily human readable but it doesn't have the extreme
complexities/redundancies(?) so many big traffic sites
Here's where I'm at with using lua filters inside of Pandoc. The script is
attached. It can be invoked by:
pandoc -f html -t TW5.lua myfile.html -o myfile.tid
or
env basename="/mylocaldir/" pandoc -f html -t TW5.lua myfile.html -o
myfile.tid
... if you want to specify a basename
Hi Steve,
I've been experimenting with text slicer and Sublime text editor.
Basically I am learning to use Sublime to help prepare cut and pasted text
for text-slicer.
Using multiply cursers, adding spaces at the end of paragraph so that the
text slicer makes sense of the spaces
Alex
On 24
Hi Steven,
originally you've also mentioned the text-slicer plugin and I'd anyway like
to figure it out and make it's interactions clearer
I was also thinking about creating something on top of it, but before that
I'd have to figure out what all the things are one can do with it,
which input
The lua filter approach might be made to work. But there's a lot of
problems matching HTML from Wikipedia and TW5 markup.
For one thing, WikiPedia (WP) uses anchors and id's to move around the
page. You can't use this approach in TW5. So the table of contents becomes
effectively static text.
This is quite an interesting conversation, tho not exactly sure where we go
with it. I'll look into regexp, though that is a bit complex (both for me,
and for my students). I'll update here as I make progress...//steve.
On Friday, February 23, 2018 at 11:54:53 AM UTC-5, Mark S. wrote:
>
> This
Ciao Mark S.
Totally agree about regex, though I love it, its difficult to use 100%
reliably the more generalised the input case is. "Debugging" regex can be
seriously difficult. Its easier for "construction-up" as you can have a
CLEAR starting point than "strip-drown" of formats that may have
Baby steps. The original post was about converting Wikipedia (HTML) into TW
format.
For my own use, I have a html2tw regex-based javascript macro that I use
inside BJ's web clipper. This works really well with my target sites
(though there's always a little clean-up and conversion from remote
This just in from John MacFarlane, the author of Pandoc:
Your best bet is to make a custom lua writer.
>
> pandoc --print-default-data-file sample.lua
>
> will give you a sample lua writer that basically
> imitates pandoc's HTML writer. You can modify that
> as you see fit. See the manual
Steve, Mark.S, TonyM, BTC & all.
Returning to the theme. I'm sure Pandoc to TW would be a great plus to ease
migration to TW. In terms of conversion TWC & TW5 variants would be
awesome. TWC is still live and significant.
My concern about it is NOT Pandoc to TW, its TW to Pandoc. BOTH ways
>
> Ok, so we'll see you back in, um, 20 minutes?
>
> uhm...
> I used org-mode for awhile. But it was too easy to accidentally bleed one
> entry into another, and there are were (are?) no good Android apps.
>
> For publicity, what's needed is to discover that some prominent politician
> or
Ok, so we'll see you back in, um, 20 minutes?
I used org-mode for awhile. But it was too easy to accidentally bleed one
entry into another, and there are were (are?) no good Android apps.
For publicity, what's needed is to discover that some prominent politician
or celebrity uses TW.
-- Mark
I'm on Linux, now the next 10 minutes I'm doing
https://wiki.haskell.org/Learn_Haskell_in_10_minutes
Then for those five org-mode users it'll be time :D
No, seriously, I think also if there won't be many users, having it on the
list could still be good in terms of "showing that there's also tw"
It kind of looks like you would need to take an existing pandoc writer
(maybe markdown) and rewrite it as a new TiddlyWiki module. But ... it's
all written in Haskell. It would be quite a time investment to get up to
speed with Haskell (unless of course you already know it) enough to make
the
As much as I like tiddlywiki to be self sufficent, including its wikitext rules
in a utility makes sence because it enables conversion between many different
formats.
Tony
--
You received this message because you are subscribed to the Google Groups
"TiddlyWiki" group.
To unsubscribe from
Hi Steve,
I recently took a look at the pandoc github repository to see what's needed
to maybe include tiddlywiki syntax
I didn't have that much time to get into it...
I think in a collaborative effort we could gather informations at one place
that help make it clear a) if it's possible and if
Steve
Footnotes to last.
1 - You could use Text-Slicer or TiddlyClip for the first step. Then
finesse with Flexity?
2 - I believe BJ is working on a new version of TiddlyClip that will
include greater scope as a customised "screen-scraper". I don't know the
details.
Best wishes
Josiah
--
Steve
IMO here you are talking about a specificity of import. Text slicer alone
is unlikely to be able to do that as its a generic tool for effective slice
on normal universally repeated elements.
But a specific Wikipedia page de-reconstructor needs pay attention to its
*specific
layout*. I
20 matches
Mail list logo
