Indexing: a wiki holds search indexes to other wikis

[Mohammad]

• Tiddlywiki is great at small to moderate number of tiddlers
• It is easier to keep few subjects in a Tiddlywiki, for example a student can have one wiki per course
• I see Tiddlywiki like a MsWord file but with tons of features and suprior flexibility
• I see Tiddlywiki as a webpage, but simple to create, customize and use
• I see Tiddlywiki as a database
• RESULT
• People using Tiddlywiki have several separate Tiddlywiki
• ISSUE
• Searching and finding something gets difficult and makes headache

One solution

• create a master Tiddywiki or a main wiki
• use a dataTiddler per wiki to index few fields per tiddler like tiddler titles, caption and keywords field
• put all these indexes in the master Tiddlywiki
• Use the standard or advanced search to look in these indexes and return a link to the correct wiki

Howto

• a simple script to create or update the index in the wiki
• export the updated index to master wiki

I know about Twedration, but this is a simple solution for whom use single html Tiddlywiki.

1. What do you think?
2. Have you seen such solution before?
3. Can you share any tools, idea?

--Mohammad

[TonyM]

Mohammad,

I considered this path for my learning tiddlywiki edition. Basically I created a tiddler containing every regular tiddler then generated a link to access it https://tiddlywiki.com/#HelloThere all you need to do it prefix every tiddler name with https://tiddlywiki.com/# to construct the link You could place this path in a tiddler field, in a tiddler that contains a list of tiddler names for each wiki.

A quick brainstorm

The above would work well for reference back to tiddlywiki.com because the title is often enough info in the titles, however you could also construct a link to open tiddly wiki.com with the search string you started with.

I favor the use of titles and a keyword field to intentionally name the keywords and being able to do a keyword only search, ie you design tiddlers to state exactly which keywords are relevant in addition to the title.

However if we wanted to get tricky we may be able to capture and extract the indexes that tiddlywiki builds internally for performance, but allow them, to be packaged and dropped into another wiki.

A more sophisticated approach would be to generate a serial number of every tiddler, then every word you find is recorded and given its own serial number, then you generate a list of every word found and list the tiddlers in which it occurs, by recording the tiddler serial number against the word serial number. A special search could then use this database to all tiddlers containing one or more words and give you a link to it.

This makes me wonder if we could use a compression technique that we could still search within.

More simply we could generate a kind of usable site index for wikis, which can be copied into another wiki, if the search finds an item in that index you know within which wiki it resides.

It may be possible to open a wiki in an iframe and feed the search string into the window once you identify the wiki in which the keyword belongs.

Regards

Tony

[Mohammad Rahmani]

Best wishes

Mohammad

On Tue, Jan 28, 2020 at 9:18 AM TonyM <anthony...@gmail.com> wrote:

Mohammad,

I considered this path for my learning tiddlywiki edition. Basically I created a tiddler containing every regular tiddler then generated a link to access it https://tiddlywiki.com/#HelloThere all you need to do it prefix every tiddler name with https://tiddlywiki.com/# to construct the link You could place this path in a tiddler field, in a tiddler that contains a list of tiddler names for each wiki.

I think this is the simplest!

1. a button 
1. to list all NON system tiddlers
2. to add a prefix to refer to wiki and create the link
3. write into a dataTiddler called xxx.index where xxx stands for wiki name like tiddlywiki.com.index
4. export it as .tid file

NOW

1. import the xxx.file
2. use a custom searchbox to search in dataTiddler
3. return the key/value as a link

A quick brainstorm

The above would work well for reference back to tiddlywiki.com because the title is often enough info in the titles, however you could also construct a link to open tiddly wiki.com with the search string you started with.

I favor the use of titles and a keyword field to intentionally name the keywords and being able to do a keyword only search, ie you design tiddlers to state exactly which keywords are relevant in addition to the title.

However if we wanted to get tricky we may be able to capture and extract the indexes that tiddlywiki builds internally for performance, but allow them, to be packaged and dropped into another wiki.

This is great! If one can export this to another wiki and use it for search external wikis!

A more sophisticated approach would be to generate a serial number of every tiddler, then every word you find is recorded and given its own serial number, then you generate a list of every word found and list the tiddlers in which it occurs, by recording the tiddler serial number against the word serial number. A special search could then use this database to all tiddlers containing one or more words and give you a link to it.

This makes me wonder if we could use a compression technique that we could still search within.

More simply we could generate a kind of usable site index for wikis, which can be copied into another wiki, if the search finds an item in that index you know within which wiki it resides.

Yep, this would be very useful.

--
You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/5e6fa555-7728-43df-877c-fcbeab8bfbd6%40googlegroups.com.

[TiddlyTweeter]

Ciao Mohammad

One thought I had was to use Powershell to generate an "Index wiki". 

This would parse single file wikis you have and extract index data and compile into a new wiki.

I done a bit of work on it and it looks doable.

Thoughts

TT

[Mohammad]

Hello Josiah,

Very interesting idea!

go ahead!

--Mohammad

[bimlas]

Mohammad,

Guided by a similar idea, I created this: https://gitlab.com/bimlas/userscript-combine-tw5-and-search-engine-results

This only works with the Node wiki and for the time being it requires a modified core tiddler until this pull request getting merged, so DO NOT USE ON REAL WIKIS, just for testing.

Anyway, your ideas seems to be more "TiddlyWikian", I will follow this thread closely. :)

[David Gifford]

When you guys get this made, with buttons and more detailed explanations on how to do it, let me know, I would be interested in using this. I have so many failed attempts at 'one-file-for-everything' TiddlyWikis, each with hidden/lost information. This could help me round up my lost sheep...

[Mohammad Rahmani]

We need some of Tiddlywiki heros come and help.

Cheers

--

You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/29ef0a25-9bc1-4ba8-80d7-5fa013f35eb7%40googlegroups.com.

[Mohammad Rahmani]

Hi bimlas

This is interesting. It off course need browser add on and node.js

But it worth.

How about single files?

Mohammad

--

You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/c1afe416-73a6-4319-81d8-7821dc548e7a%40googlegroups.com.

[bimlas]

Mohammad,

 

How about single files?

Because the solution uses WebServer API, it does not work with HTML files. 

[Mark S.]

On Monday, January 27, 2020 at 9:14:46 PM UTC-8, Mohammad wrote:

Use the standard or advanced search to look in these indexes and return a link to the correct wiki

Creating a data tiddler doesn't sound too hard.

But you would need to make a special search tool to show the contents of the data tiddler and reconstruct a link.

Hmm. If you're on a local file system, can you link to another tiddler on the local system?

> I know about Twedration, but this is a simple solution for whom use single html Tiddlywiki.

Twederation used/uses single files. It had some over-arching concerns re long-term storage, sizing, and security.

[Mohammad]

Thank you bimlas,

Your solution works great for who can use Node.js

--Mohammad

[Mohammad]

On Tuesday, January 28, 2020 at 7:25:49 PM UTC+3:30, Mark S. wrote:

On Monday, January 27, 2020 at 9:14:46 PM UTC-8, Mohammad wrote:

Use the standard or advanced search to look in these indexes and return a link to the correct wiki

Creating a data tiddler doesn't sound too hard.

But you would need to make a special search tool to show the contents of the data tiddler and reconstruct a link.

Hmm. If you're on a local file system, can you link to another tiddler on the local system?

Yes, TW-Scripts has a search tool to open other wikis, but it does not have an index file to search them just creates a link! 

> I know about Twedration, but this is a simple solution for whom use single html Tiddlywiki.

Twederation used/uses single files. It had some over-arching concerns re long-term storage, sizing, and security.

Mark,

Do you mean Twedration can search multiple wikis? 

[Mark S.]

On Tuesday, January 28, 2020 at 10:40:02 AM UTC-8, Mohammad wrote:

Mark,

Do you mean Twedration can search multiple wikis? 

In Twederation, your TiddlyWiki file pulls in the content of other wikis. So, then your instance has a copy of tiddlers from other wikis. So in that sense it does search the other wikis. Obviously you couldn't do this indefinitely without running into space problems.

[Mohammad]

I did not know this. It will give a try!

So, why it is not popular?

--Mohammad 

[TonyM]

Folks,

Perhaps a helpful feature would be to allow keywords to be selected from the content of a tiddler. A View Template addon that parses the whole text field, even other fields like description caption etc.. and lists them. With a simple button or checkbox allowing yo you select which works to add to a keyword field (perhaps stored off tiddler) with the words you select. This makes the keywords "curated" and purposefully selected, a little more work but possibly much more meaningful.

Ideally we could list the words given how often they are used, excluding stop words and/if/the/etc..., exclude widgets etc and perhaps even favoring longer words as these can be more specific.

 

Regards

Tony

On Tuesday, January 28, 2020 at 4:14:46 PM UTC+11, Mohammad wrote:

[Mohammad]

Good idea,

 I think Josiah Powershell script can scan a wiki or wiki folder and create an index dataTiddler, then import to a master wiki to do the search!
In my opinion, this might be simple approach at the cost of having Powershell installed and run it.

--Mohammad

[James]

Hi Mohammad,

Do you have any workable example?

It sounds interesting, I have multiple TWs and I am looking for ways to filter tiddlers across the TWs.

[Mohammad]

Hello James,

 Not yet, but I will share if you have any working solution!

By the way the Twedration seems to do the job, when you are on node.js

See Mark post: https://groups.google.com/d/msg/tiddlywiki/8qyPUgmpDYY/v4-fPn53AgAJ

--Mohammad

[oleghbond]

Dear tiddlymates,

With similar thoughts I started my project TW Farm bearing in mind this idea from the very beginning and giving an appropriate name. I have still been successfully using it for developing my TW farm. The main limitation of my realization is absence of cross searching between separate wikis. So, at this point I totally support Mohammed in raising the proper task.

The Master TW of my farm is at https://protw.github.io/, the JS code is located at https://github.com/protw/twfarm and its user description is at https://protw.github.io/twfarm/ (though in Ukrainian).

My JS knowledge level and limited time are not sufficient for developing the idea to full extent. But I believe that it is an issue of very soon time.

As a basis of the TW farm on my computer I use BobEXE TW server. For publication at the Web of my TW farm I use git instruments and web resources together with my JS code (https://github.com/protw/twfarm).

My vision that BobEXE could be a very good native platform for developing the idea once more expressed by Mohammed.

Olegh

вівторок, 28 січня 2020 р. 07:14:46 UTC+2 користувач Mohammad написав:

[Mark S.]

Here's code that will create an index for the current wiki. It uses the wiki title as the basis for the index name. It saves title, caption, and tags.

The next step would be to modify the search code so that it knows to look in the index files.

\define indexvalue() $(title)$|$(caption)$|$(tags)$
<$vars indextitle={{{ [{$:/SiteTitle}addprefix[index-for-]] }}}>
<$button>Make Index
<$list filter="[all[tiddlers]!is[system]]">
<$vars title={{!!title}} caption={{!!caption}} tags={{!!tags}} >
<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/>
</$vars>
</$list>
</$button>
</$vars>

On Monday, January 27, 2020 at 9:14:46 PM UTC-8, Mohammad wrote:

[Mohammad]

Hi Mark,

 This is wonderful!

The next step is to create a customize sarchbox. This way one can search and find direct link to the other wiki!

--Mohammad

p.s

Tested on TW5.1.22pre creates a 106kB file!

[Mohammad]

Olegh,

 For documentation purpose and future reference, is there any readme or tutorial help to setup such Tiddly farm?

 

--Mohammad

[Mohammad]

This is a first attempt to make search in an index. NO link is created noe

\define thisSearchMacro(search:"", dataTid:"exmp39/acronyms/web")

<dl>

<$list filter="[[$dataTid$]indexes[]regexp:title[(?i)$search$]]">

<dt><$text text=<<currentTiddler>>/></dt>

<dd><$text text={{{ [[$dataTid$]getindex<currentTiddler>] }}}/></dd>

</$list>

</dl>

\end

<$edit-text tiddler=test tag=input default="no search"/>

<$macrocall $name="thisSearchMacro" search={{test}} dataTid="index-for-TiddlyWiki Pre-release"/>

The thisSearchMacro was developed by Stephan Hradek.

On Friday, January 31, 2020 at 9:50:23 PM UTC+3:30, Mark S. wrote:

[oleghbond]

I'll try to translate https://protw.github.io/to farm as sooner

[David Gifford]

Where does this save to? I don't see a new tiddler in the wiki, nor a new file in the same folder.

[Mark S.]

 Here's an improved version which does a better job constructing the output tiddler so it doesn't insert square brackets.

It also attempts to guess at the working file name of the current tiddler and use that for a field "tiddlywiki", based on the only two fields that

are available for that function. These guesses could be wrong. The user can make sure they're "right" by manually setting

the tiddler value "$:/config/SaveWikiButton/Filename" to the file name of that tiddler.

\define indexvalue() """$(title)$|$(caption)$|$(tags)$"""

<$wikify name=indextitle text={{{ [{$:/SiteTitle}addprefix[index-for-]] }}} select="0">

<$button>Make Index
<$list filter="[all[tiddlers]!is[system]]">
<$vars title={{!!title}} caption={{!!caption}} tags={{!!tags}} >
<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/

>
</$vars>
</$list>
<$wikify name=tiddlywiki text={{{ [{$:/SiteTitle}addsuffix[.html]][{$:/config/SaveWikiButton/Filename}]+[!is[blank]last[]] }}} >
<$action-setfield $tiddler=<<indextitle>> $field="tiddlywiki" $value=<<tiddlywiki>>/>
</$wikify>
</$button>
</$wikify>

[Mark S.]

It should pop right up under the "recent" tab.

[David Gifford]

[Mark S.]

Be sure to try the newer version I posted.

Is "ROAM-y" online somewhere?

I assume that if it's breaking that it has to do with one of your titles. Which is amazing, considering that none of the 1300 tiddlers on tiddlywiki.com broke it.

Thanks!

[Mat]

Mark S. wrote:

\define indexvalue() """$(title)$|$(caption)$|$(tags)$"""
<$wikify name=indextitle text={{{ [{$:

/SiteTitle}addprefix[index-for-]] }}} select="0">
...

Detail: that last select="0" should be removed.

<:-)

[Mat]

further, when I do copy-paste of Marks thing, I find that this line is split up:

...

<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/
>

...which causes a bug. Make sure that last angle bracket is where it should be so to get only a button showing.

<:-) 

[Mat]

BTW, Mark, that is just a really slick way to make an index.

<:-)

[Mohammad]

David,

before saving the tiddler hosts Mark code, look at the lines

<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/
>

and correct it as

<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/>

Mohammad

[Mohammad]

Hi Mark,

 I added an export button so, one can index and export!

\define indexvalue() """$(title)$|$(caption)$|$(tags)$"""
<$wikify name=indextitle text={{{ [{$:/SiteTitle}addprefix[index-for-]] }}} select="0">


<$button>Make Index
<$list filter="[all[tiddlers]!is[system]]">
<$vars title={{!!title}} caption={{!!caption}} tags={{!!tags}} >
<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/>
</$vars>
</$list>
<$wikify name=tiddlywiki text={{{ [{$:/SiteTitle}addsuffix[.html]][{$:/config/SaveWikiButton/Filename}]+[!is[blank]last[]] }}} >
<$action-setfield $tiddler=<<indextitle>> $field="tiddlywiki" $value=<<tiddlywiki>>/>
</$wikify>
</$button>

<$button>Export
<$action-sendmessage $message="tm-download-file" $param=<<indextitle >> filename={{{[<indextitle>addsuffix[.tid]]}}}    />
</$button>
</$wikify>

A tiddler containing above buttons are attached. download drag and drop into your wiki.

--Mohammad

[Mohammad]

Hi Mark,

 I pushed a proof of concept of index external wikis and search from another wiki on tiddlyspot

see: http://kookma.searchwikis.tiddlyspot.com/

I also developed a searchbox to search in title, caption and tags of indexed wiki.

As an example, I have indexed tiddlywiki.com and one can search and work with created links.

Please have a look and let me know how we can improve it.

@Mat

I have corrected the Mark code based on your comments. I also removed """ from indexvalue macro.

--Mohammad

[David Gifford]

Hi Mohammad

Wow, this is great! I can see myself using this a lot. Observations:

1) Instructions step 3 should specify to save it as a .tid file.

2) Instructions step 5.1 should say that you also need to change the link in wikiUrl

3) The CSS needs to change for the search results. I am doing the search in a file from empty.html and the resulting links are an illegible bright yellow.

4) So, the limitation to this is only being able to search the tiddlers of one external file? I tried changing dataTid to "[tag[indexes]]" (which is how I tagged the indexes imported from the two files I wanted to index and search), but it gave me an error message. I think this tool would be more helpful if there were a way to search the indexes of multiple other 'branch' files from one 'trunk' file.

Thanks again for this.

[Mat]

Mohammad, nice proof of concept.

David Gifford wrote:

4) So, the limitation to this is only being able to search the tiddlers of one external file? 

I don't see why it would be limited to only one external wiki. maybe that is a limitation in Mohammads proof-of-concept demo, but there should be no theoretical such limitation. IMO the severe limitation is rather that changes in the external file are obviously not reflected in the index. 

<:-)

[David Gifford]

I could live with the severe limitation you mention. Just schedule a weekly updating and drag-and-drop of the index tids of the branch files you edited during the week (would require keeping a log). Small price to pay for having the ability to search a gazillion tiddlers from a central TiddlyWiki. But my point 4 would be the missing ingredient.

[Mark S.]

Hi Mohammad,

That's looking really nice!

My idea is that each index tiddler will have a field like index-source where the base path to the source wiki is found.

Then in the search engine, it would look at ALL tiddlers that have index-source field, and it would use that instead

of "data-tid". That way you could search through more than one index-tiddler. This may require the user to

adjust the index field depending on the location of the target wiki file. A small price to pay to be able to search

anywhere. Maybe.

I would not worry about presenting the tags and caption fields -- it's enough to know that some field contains the searched-for value.

Finally, I think Mat has some example code where he added additional search fields to the standard search box. I would like the

index searches to eventually end up there, sort of down below the existing search hits. That way searching other

wikis becomes "almost" as transparent as a regular wiki search. Or am I dreaming?

Once people begin to use an indexer, it might be good to provide for a "keywords" field, that people can complete

with terms that may not actually occur in the original text.

Thanks!

[Mat]

Mark S. wrote:

Finally, I think Mat has some example code where he added additional search fields to the standard search box. I would like the

index searches to eventually end up there, sort of down below the existing search hits. That way searching other

wikis becomes "almost" as transparent as a regular wiki search. Or am I dreaming?

I would think you are dreaming (tho not 100% sure). I think I've come with proposals and perhaps included images but not actually implemented something that involves modifying the search engine. Maybe it was someone else?

<:-)

[Mark S.]

Well, if it wasn't you, it should have been ;-)

[Mat]

haha!

<:-)

[Mohammad]

Many thanks all.

@Mark

 I created a second solution implement your idea of getting Url from a field. The code still needs some polishing.

@Mat

I add the multi index search.

@David

I corrected the wiki to address your comments.

I think:

The indexer shall be corrected to reflect the changes

The searchbox should improved

The number of output should be controlled some way

and documentation shall be added.

Latest update: http://kookma.searchwikis.tiddlyspot.com/

--Mohammad

p.s: Solution resulted from collaboration of several interested people are the best solution :-)

[Mat]

Hm, I now vaguely recall @pmario made something with extra search fields.... it was relatively recently, I think... 'or am I dreaming'?

Now I did make searchwords which does not have extra search fields but it does present extra results in the search results list. It's not this you're referring to, are you Mark?

<:-)

[David Gifford]

Woooooooooowwwwwww

This has been a good week for me. And for TiddlyWiki.

[David Gifford]

Ideas for improving the search results:

1) Change <h2> to <br> in file links to reduce vertical overhead of header text

2) Change the file link so the file sitetitle stands as an alias for the file URL. As it stands, it renders really long links to filepaths for me.

I did the first change in my own file, but am not sure how to do the second.

On Saturday, February 1, 2020 at 2:28:14 PM UTC-6, Mohammad wrote:

[oleghbond]

As I promised, I translated my documentation for TW-farm into English. So, please find it where it was before: 

• https://protw.github.io/twfarm/

Olegh

пʼятниця, 31 січня 2020 р. 22:51:51 UTC+2 користувач Mohammad написав:

[Mohammad]

Hi Olegh,

 Many thanks!

I will add a link in TW-Scripts to this.

@David

This worth to be listed in Tiddlywiki Toolmap.

--Mohammad

[Mark S.]

It looks like the translation may have changed <<alert-info into << alert-info (note the space), making it

not render correctly?

[Mohammad]

Mark,

A problem encountered is making link for tiddler title have : in their name.

Look at http://kookma.searchwikis.tiddlyspot.com/#Search%20Indexes%20of%20External%20Wikis%20II

and click on Hidden Setting: ViewTemplate and EditTemplate

The second item found for Search results in: https://tiddlywiki.com

It seems Tiddlywiki makes links differently. There is no issue for \, /, space, $, ., ...

--Mohammad

On Friday, January 31, 2020 at 9:50:23 PM UTC+3:30, Mark S. wrote:

Here's code that will create an index for the current wiki. It uses the wiki title as the basis for the index name. It saves title, caption, and tags.

The next step would be to modify the search code so that it knows to look in the index files.

\define indexvalue() $(title)$|$(caption)$|$(tags)$

<$vars indextitle={{{ [{$:/SiteTitle}addprefix[index-for-]] }}}>

<$button>Make Index
<$list filter="[all[tiddlers]!is[system]]">
<$vars title={{!!title}} caption={{!!caption}} tags={{!!tags}} >
<$action-setfield $tiddler=<<indextitle>> $index=<<currentTiddler>> $value=<<indexvalue>>/

>
</$vars>
</$list>
</$button>
</$vars>

On Monday, January 27, 2020 at 9:14:46 PM UTC-8, Mohammad wrote:

[Mohammad]

Further input,

 The strategy I used in https://kookma.github.io/TW-Scripts//#Search%20in%20%20Official%20Tiddlywiki%20Site

is to open tiddlywiki.com and search again in opened page, but I think if we use this strategy here, like we do search two times.

--Mohammad

[Mohammad]

@Olegh

1. look at https://protw.github.io/twfarm/#Organization%20of%20work%20with%20wiki-farm%20folders as Mark said, it seems << alert-info should changes to <<alert-info the alert does not work now
2. I changed the sidebar layout to Fluid story, as it does not show correctly the sidebar right now
3. I noticed also in https://protw.github.io/twfarm/#What%20and%20how%20TW%20Farm%20solves there is undefined widget cuases by extra space in calling << $macrocall

Any way thank you for this great tutorial and wonderful methodology you created to host multiple wiki on the internet!

I am wondering why it is listed in tiddlywiki.com community tools and plugins.

--Mohammad

[Olegh Bondarenko]

@Mohammad

Thanks for your comments. Points 1 & 3 are done, please check. Point 2 is not clear. Would you please explain it to me in a bit more details.

Olegh Bondarenko
https://bit.ly/oleghbond

пн, 3 лют. 2020 о 09:33 Mohammad <mohammad...@gmail.com> пише:

--
You received this message because you are subscribed to a topic in the Google Groups "TiddlyWiki" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tiddlywiki/8qyPUgmpDYY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to tiddlywiki+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/1537788a-047a-4124-9134-0e274f91c4ab%40googlegroups.com.

[oleghbond]

You are absolutely correct - as a result of the translation extra spaces were added. I deleted them, now it's fine. Thanks

понеділок, 3 лютого 2020 р. 06:43:13 UTC+2 користувач Mark S. написав:

[oleghbond]

Hello, Mohammad,

Thanks to you and your wonderful Shiraz - I included it in the Template wiki for my wiki-farm. Please have a look at https://protw.github.io/templ.

понеділок, 3 лютого 2020 р. 06:23:57 UTC+2 користувач Mohammad написав:

[Mohammad]

Hi Olegh,

 Thanks. It is fine now!

The second comment was regarding sidebar size, on my desktop I see a small sidebar and I have to scroll left and right to work with sidebar!

I think your settings for tiddler, story river and sidebar width are non consistent.

I changes Story layout to fixed story-fluis sidebar and it works fine on my desktop.

--Mohammad

On Monday, February 3, 2020 at 1:14:46 PM UTC+3:30, oleghbond wrote:

@Mohammad

Thanks for your comments. Points 1 & 3 are done, please check. Point 2 is not clear. Would you please explain it to me in a bit more details.

Olegh Bondarenko
https://bit.ly/oleghbond

пн, 3 лют. 2020 о 09:33 Mohammad <mohamma...@gmail.com> пише:

@Olegh

1. look at https://protw.github.io/twfarm/#Organization%20of%20work%20with%20wiki-farm%20folders as Mark said, it seems << alert-info should changes to <<alert-info the alert does not work now
2. I changed the sidebar layout to Fluid story, as it does not show correctly the sidebar right now
3. I noticed also in https://protw.github.io/twfarm/#What%20and%20how%20TW%20Farm%20solves there is undefined widget cuases by extra space in calling << $macrocall

Any way thank you for this great tutorial and wonderful methodology you created to host multiple wiki on the internet!

I am wondering why it is listed in tiddlywiki.com community tools and plugins.

--Mohammad

--
You received this message because you are subscribed to a topic in the Google Groups "TiddlyWiki" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tiddlywiki/8qyPUgmpDYY/unsubscribe.

To unsubscribe from this group and all its topics, send an email to tiddl...@googlegroups.com.

[Olegh Bondarenko]

@Mohammad

Not clear yet.

I made the following settings in the section Theme tweaks:

• Story right: 1170px
• Story width: 1170px
• Tiddler width: 1086px
• Sidebar breakpoint: 1060px

Please tell me what I shall do instead.

Olegh Bondarenko
https://bit.ly/oleghbond

пн, 3 лют. 2020 о 11:58 Mohammad <mohammad...@gmail.com> пише:

To unsubscribe from this group and all its topics, send an email to tiddlywiki+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/c5b07f98-4232-4105-b0ff-520dbdf77ecf%40googlegroups.com.

[Mohammad]

What I see on my monitor is attached as an image.

--Mohammad

On Monday, February 3, 2020 at 2:02:09 PM UTC+3:30, oleghbond wrote:

@Mohammad

Not clear yet.

I made the following settings in the section Theme tweaks:

• Story right: 1170px
• Story width: 1170px
• Tiddler width: 1086px
• Sidebar breakpoint: 1060px

Please tell me what I shall do instead.

Olegh Bondarenko
https://bit.ly/oleghbond

пн, 3 лют. 2020 о 11:58 Mohammad <mohamma...@gmail.com> пише:

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/c5b07f98-4232-4105-b0ff-520dbdf77ecf%40googlegroups.com.

[oleghbond]

@Mohammad

I've changed fixed story/fluid sidebar to fluid story/fixed sidebar. Please check now how it works for you and let me know.

понеділок, 3 лютого 2020 р. 13:01:58 UTC+2 користувач Mohammad написав:

[Mohammad]

Works fine!

Thank you Olegh!

--Mohammad

[David Gifford]

Yes of course, I will put both the Farm and your indexing thing in the Toolmap in a few moments.

[David Gifford]

I added both the TW5 Farm Technique and the Kookma SearchWiki in the toolmap, under a new category, "Multiple TiddlyWikis, connecting and integrating"

https://dynalist.io/d/zUP-nIWu2FFoXH-oM7L7d9DM

Let me know if you think there are other categories in which it would be important to include them.

[Mohammad Rahmani]

Thank you David!

Sounds good to me.

Best wishes

Mohammad

--
You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/45180034-1a32-4b9e-823e-5952bd8fb766%40googlegroups.com.

[ILYA]

# Another approach

Every wiki would have tiddlers.json. This JSON could be any of (I think we need to support all 4 methods):
- a file served by regular web server (github.io or apache or caddy or whatever you want)
- served by bob
- served by nodejs tw web server
- regular tiddler

The file would be in the same format as expected by lazy loading mechanism present in TW5. IRC the JSON structure is a list where every item of the list is a map which includes all the fields of the tiddler except text. I.e. it would have tags, fields, title, caption and so on.

I think the main wiki could have links to all locations where subject wiki are located. Here is how it might work from user perspective.

1. We can have special tag to mark dependent wiki sources. Something like $:/plugins/multiwiki/Source
2. We tag new tiddler (let's call this kind of tiddler SourceWiki) with tag $:/plugins/multiwiki/Source. This tiddler would have _canonical_uri which points to tiddlers.json
3. We would have a filter which finds all tiddlers with that tag.
4. I could extend split-web plugin to support
- multiple tiddlers.json sources
- use SourceWiki tiddler name as a namespace. I.e. as prefix for each imported tiddler.
5. In this case all external tiddlers would appear to be local. We would need to modify view template to:
- provide visual indication that it is external read only tiddler
- add visual element to open external tiddler in new browser window or tab (in the context of subject wiki)

The reason I think this approach is viable is because it uses existent TW5 lazy loading mechanism. It also makes external tiddlers first class citizens (the only difference is they are read only).

Best regards,
ILYA

Tiddlywiki is great at small to moderate number of tiddlers

It is easier to keep few subjects in a Tiddlywiki, for example a student can have one wiki per course

I see Tiddlywiki like a MsWord file but with tons of features and suprior flexibility

I see Tiddlywiki as a webpage, but simple to create, customize and use

I see Tiddlywiki as a database

RESULT

People using Tiddlywiki have several separate Tiddlywiki

ISSUE

Searching and finding something gets difficult and makes headache


One solution

create a master Tiddywiki or a main wiki

use a dataTiddler per wiki to index few fields per tiddler like tiddler titles, caption and keywords field

put all these indexes in the master Tiddlywiki

Use the standard or advanced search to look in these indexes and return a link to the correct wiki

Howto

a simple script to create or update the index in the wiki

export the updated index to master wiki

I know about Twedration, but this is a simple solution for whom use single html Tiddlywiki.


What do you think?

Have you seen such solution before?

Can you share any tools, idea?

--Mohammad

--

On 28 January 2020 05:13:01 GMT-08:00, David Gifford <dgif...@crcna.org> wrote:

When you guys get this made, with buttons and more detailed explanations on how to do it, let me know, I would be interested in using this. I have so many failed attempts at 'one-file-for-everything' TiddlyWikis, each with hidden/lost information. This could help me round up my lost sheep...

On Monday, January 27, 2020 at 11:14:46 PM UTC-6, Mohammad wrote:

• Tiddlywiki is great at small to moderate number of tiddlers
• It is easier to keep few subjects in a Tiddlywiki, for example a student can have one wiki per course
• I see Tiddlywiki like a MsWord file but with tons of features and suprior flexibility
• I see Tiddlywiki as a webpage, but simple to create, customize and use
• I see Tiddlywiki as a database
• RESULT
• People using Tiddlywiki have several separate Tiddlywiki
• ISSUE
• Searching and finding something gets difficult and makes headache

One solution

• create a master Tiddywiki or a main wiki
• use a dataTiddler per wiki to index few fields per tiddler like tiddler titles, caption and keywords field
• put all these indexes in the master Tiddlywiki
• Use the standard or advanced search to look in these indexes and return a link to the correct wiki

Howto

• a simple script to create or update the index in the wiki
• export the updated index to master wiki

I know about Twedration, but this is a simple solution for whom use single html Tiddlywiki.

1. What do you think?
2. Have you seen such solution before?
3. Can you share any tools, idea?

--Mohammad

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[ILYA]

# Another approach

Every wiki would have tiddlers.json. This JSON could be any of (I think we need to support all 4 methods):
- a file served by regular web server (github.io or apache or caddy or whatever you want)
- served by bob
- served by nodejs tw web server
- regular tiddler

The file would be in the same format as expected by lazy loading mechanism present in TW5. IRC the JSON structure is a list where every item of the list is a map which includes all the fields of the tiddler except text. I.e. it would have tags, fields, title, caption and so on.

I think the main wiki could have links to all locations where subject wiki are located. Here is how it might work from user perspective.

1. We can have special tag to mark dependent wiki sources. Something like $:/plugins/multiwiki/Source
2. We tag new tiddler (let's call this kind of tiddler SourceWiki) with tag $:/plugins/multiwiki/Source. This tiddler would have _canonical_uri which points to tiddlers.json
3. We would have a filter which finds all tiddlers with that tag.
4. I could extend split-web plugin to support
- multiple tiddlers.json sources
- use SourceWiki tiddler name as a namespace. I.e. as prefix for each imported tiddler.
5. In this case all external tiddlers would appear to be local. We would need to modify view template to:
- provide visual indication that it is external read only tiddler
- add visual element to open external tiddler in new browser window or tab (in the context of subject wiki)

The reason I think this approach is viable is because it uses existent TW5 lazy loading mechanism. It also makes external tiddlers first class citizens (the only difference is they are read only).

Best regards,
ILYA

>>

>> - Tiddlywiki is great at small to moderate number of tiddlers
>> - It is easier to keep few subjects in a Tiddlywiki, for example a

>> student can have one wiki per course

>> - I see Tiddlywiki like a MsWord file but with tons of features
>and
>> suprior flexibility
>> - I see Tiddlywiki as a webpage, but simple to create,
>customize
>> and use
>> - I see Tiddlywiki as a database
>> - RESULT
>> - People using Tiddlywiki have several separate Tiddlywiki
>> - ISSUE
>> - Searching and finding something gets difficult and makes
>headache
>>
>>
>> One solution
>>
>> - create a master Tiddywiki or a main wiki
>> - use a dataTiddler per wiki to index few fields per tiddler like

>> tiddler titles, caption and keywords field

>> - put all these indexes in the master Tiddlywiki
>> - Use the standard or advanced search to look in these indexes and

>
>> return a link to the correct wiki
>>
>> Howto
>>

>> - a simple script to create or update the index in the wiki
>> - export the updated index to master wiki

>>
>> I know about Twedration, but this is a simple solution for whom use
>single
>> html Tiddlywiki.
>>
>>

>> 1. What do you think?
>> 2. Have you seen such solution before?
>> 3. Can you share any tools, idea?

[oleghbond]

Ilya's idea looks complementary to TW farm approach (https://protw.github.io/twfarm/). Every wiki of a wiki-farm collection can be presented as external read-only JSON file in the central wiki (catalog or directory). Every wiki must have a single typical tiddler with fields containing information necessary for identification and web-location of each wiki.

Of course the central wiki (directory) needs a plugin for reading, searching and revealing structured information from those external JSON files. 

IMHO BobEXE suits very well for this role.

вівторок, 4 лютого 2020 р. 05:26:38 UTC+2 користувач ILYA написав:

[Mark S.]

My guess ...

You may need to uri encode the titles. The browser can handle most everything, even if it's not uri-encoded, but colons are special.

It might be worth it even to only uri encode the titles containing colons, since the encoded links are hard to read.

[David Gifford]

Hi all

This thread has grown quiet, but it feels like there are a few items pending, so I am going to dump here, for whomever might be interested:

1) http://kookma.searchwikis.tiddlyspot.com/, the real solution is the "Solution ii". But it says "[in progress]". What is in progress or left to be done from your perspective, Mohammad?

2) I am going to re-mention my idea of using the SiteTitle of the files searched as an alias so that the search results will not be long ugly filepaths. If Mohammad doesn't like that idea, maybe someone could at least show me how I could do it for my copy.

I really think this is an exciting possibility for me. Not only could I use it to search my own notes, I could also put one up online, I think, for my site users to search my multiple online wikis.

[David Gifford]

Addendum:

Another item that would enhance this would be for the search results to NOT show the "Search results in..." line if there are no results.

[David Gifford]

Oh, and one addendum more:

Check out https://giffmex.org/wiki/, the busqueda tab, to see this plugin, slightly tweaked, in action for multiple online files in my website!

On Sunday, February 9, 2020 at 10:12:05 AM UTC-6, David Gifford wrote:

[Mohammad]

Hi David,

 I am working on dynamic tables in Shiraz plugin and hopefully will finish soon!

After that I come back to searchwikis.

On Sunday, February 9, 2020 at 7:42:05 PM UTC+3:30, David Gifford wrote:

Hi all

This thread has grown quiet, but it feels like there are a few items pending, so I am going to dump here, for whomever might be interested:

1) http://kookma.searchwikis.tiddlyspot.com/, the real solution is the "Solution ii". But it says "[in progress]". What is in progress or left to be done from your perspective, Mohammad?

The index file needs some more fields. The search tools needs some polishing like open all results from one wiki in the same tab (or at least this can be configurable)

Examples and documentation are also left for last step 

[David Gifford]

On Sunday, February 9, 2020 at 12:53:08 PM UTC-6, Mohammad wrote:

Hi David,

 I am working on dynamic tables in Shiraz plugin and hopefully will finish soon!

After that I come back to searchwikis.

So did you finish the dynamic tables? Have you come back to this? 

[David Gifford]

At the very least, the indexes created by the 'make index' button should have an 'index-source' field, to save that manual step, even if the build-and-export is not able to grab the actual file location or URL of the present file and place it in the field.

[Mohammad]

Hi David,

 Sure, I will come back to this and finalize it! Hopefully by tomorrow!

 I will take your comments here!

--Mohammad

[David Gifford]

Hi Mohammad,

Two comments:

1. Today is my birthday, so hearing you will work on it is a great birthday present! Thanks!

2. It would be great if for each file the user could have a system tiddler like $:/searchwiki/filepath where to insert the URL or filepath. Then have have build-and-export 'Make index' button grab that filepath and insert it into an index-source field. That way the updating process only involves opening the file, clicking Make index > tid file, and dragging the created tid file into a central index file. As it is now, the user must also create an index-source field manually to each created tid file, then insert manually the URL or filepath into the index-source field.

[Mohammad]

David!

On Thursday, February 27, 2020 at 8:29:14 PM UTC+3:30, David Gifford wrote:

Hi Mohammad,

Two comments:

1. Today is my birthday, so hearing you will work on it is a great birthday present! Thanks!

Happy your birthday!!

Live many years with full health and happiness!

 

2. It would be great if for each file the user could have a system tiddler like $:/searchwiki/filepath where to insert the URL or filepath. Then have have build-and-export 'Make index' button grab that filepath and insert it into an index-source field. That way the updating process only involves opening the file, clicking Make index > tid file, and dragging the created tid file into a central index file. As it is now, the user must also create an index-source field manually to each created tid file, then insert manually the URL or filepath into the index-source field.

Good suggestion! I will try to implement and push updates and discuss these then

[David Gifford]

Thanks and thanks, Mohammad!

[Julio Peña]

Hello David,

Just read the post...Happy Birthday...wish you many more years and much health!

Felicidades hermano.

all the best,

Julio

[David Gifford]

¡Muchas gracias Julio!

[Mohammad]

Hi David,

 I thought to make a plugin for searchwikis and

1. add a tab automatically to AdvancedSearch
2. use the AdvancedSearch feature
3. add wiki title, url and indexing date to the indexTiddler
4. a flat interface

I will push an update soon!

Let me know if you have any other points

--Mohammad

[David Gifford]

Sounds great to me!

[Mohammad]

The plugin is published here

https://groups.google.com/d/msg/tiddlywiki/BFiSyOpBNHA/3E05qSTDAQAJ

[Mohammad]

To continue this thread see

https://groups.google.com/d/msg/tiddlywiki/BFiSyOpBNHA/3E05qSTDAQAJ

This thread resulted in Searchwikis plugin.

--Mohammad

[Olegh Bondarenko]

Dear Mohammad and David, 

the plugin TW-Searchwikis completely fits to idea of my tool TW Farm (https://protw.github.io/twfarm)! I've immediately started thinking of combining these two.

For this I need your advice how to generate the index data tiddler and then copy/move it into another (central) wiki using command line only.

Grateful in advance.

Olegh Bondarenko
https://bit.ly/oleghbond

вт, 3 бер. 2020 о 21:34 Mohammad <mohammad...@gmail.com> пише:

--
You received this message because you are subscribed to a topic in the Google Groups "TiddlyWiki" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tiddlywiki/8qyPUgmpDYY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to tiddlywiki+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/f986e559-d396-47cd-9182-df7b6fbd510f%40googlegroups.com.

[Mohammad]

Olegh,

 The plugin, has a tiddler $:/plugins/kookma/searchwikis/ui/index-export generates the index

But I have not used command line! I know Josiah and Mark were working on Powersell and 

they may help!

--Mohammad

On Wednesday, March 4, 2020 at 10:06:04 AM UTC+3:30, oleghbond wrote:

Dear Mohammad and David, 

the plugin TW-Searchwikis completely fits to idea of my tool TW Farm (https://protw.github.io/twfarm)! I've immediately started thinking of combining these two.

For this I need your advice how to generate the index data tiddler and then copy/move it into another (central) wiki using command line only.

Grateful in advance.

Olegh Bondarenko
https://bit.ly/oleghbond

вт, 3 бер. 2020 о 21:34 Mohammad <mohamma...@gmail.com> пише:

To continue this thread see

https://groups.google.com/d/msg/tiddlywiki/BFiSyOpBNHA/3E05qSTDAQAJ

This thread resulted in Searchwikis plugin.

--Mohammad

--
You received this message because you are subscribed to a topic in the Google Groups "TiddlyWiki" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tiddlywiki/8qyPUgmpDYY/unsubscribe.

To unsubscribe from this group and all its topics, send an email to tiddl...@googlegroups.com.