Hi again, I'm testing my luck in Stackoverflow to see if I can get more eyes and keep the conversation going: http://stackoverflow.com/questions/25215103/building-a-tree-outliner-like-gr... This community is very responsive but I feel I'm not understanding quickly/properly enough the main logic of tree-like browsers on Moose. So, any extra help is welcomed. Cheers, Offray On 08/07/2014 03:28 PM, Offray Vladimir Luna Cárdenas wrote:
Hi Doru and Community :-),
In the screenshot at [1] you can see my explorations. I took the code from treeWithChildrenByeLevel in the GLMBasicExamples and modified it until I got this:
[1] http://www.enlightenment.org/ss/e-53e3dee6777744.68598023.jpg
So I have now a browser which shows a tree made of UbakyeNodes (a tree like data structure I defined), but I would like not to show the Ubakye Nodes, but the titles of each node (headers) and the contents (bodies) of them when selected. With your help I have understood that I need to pass the collection of all children (not just the headers of them), but I don't know who to sellect something particular in that collection to be shown on the browser, like headers of nodes in the #tree panel or bodies in the #body panel.
I would like to change also the size of each panel to be more like the shown in the screenshot, instead of the default one and be relative to window size. Is this possible?
Thanks,
Offray
Below is the code, for easiness of reading:
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- " Another exploration of Outliners provided by Glamorous Toolkit and browsers. This code was obtained by running 'GLMBasicExamples open' and then browsing until 'treeWithChildrenByLevel'. Some code was modified to open explicitely the browser on the world and starting to hack on it. "
| browser mainTree |
mainTree := UbakyeNode new. mainTree becomeDefaultTree.
browser := GLMTabulator new. browser column: #tree; column: [ :c | c row: #body; row: #plugins ]. (browser transmit) to: #tree; andShow: [ :a | (a tree) title: mainTree header; children: [ :eachNode | eachNode children. ] "Children must return a collection" ]. (browser transmit) to: #body; from: #tree; andShow: [ :a | a text title: 'Cuerpo | Body ' ]. (browser transmit) to: #plugins; from: #tree port: #selectionPath; andShow: [ :a | a text title: 'Plugins | Extensiones' ].
browser openOn: mainTree children. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
On 08/05/2014 12:19 AM, Tudor Girba wrote:
Hi,
You probably need this:
explorer transmit to: #tree; andShow: [:a | a tree display: headers; *children: [:eachNode | eachNode theMessageYouUseToGoToTheChildrenOfANode ]*].
A tree is a recursive structure, and to describe it you need: - a way to construct the root elements. This is the result of applying display: to the input object. So, display: either takes a collection or a block that will return a collection when executed. - a way to define the children for each node. This is the result of applying children:
You should also take a look at the examples from: GLMBasicExamples open
Does this help now?
Cheers, Doru
On Sun, Jul 27, 2014 at 4:59 PM, Offray Vladimir Luna Cárdenas <offray@riseup.net <mailto:offray@riseup.net>> wrote:
Hi,
Answering to myself: I have solved the code that selects the headers of the main tree. The key is to create a new collection containing only node names. Here is the code:
"*************************" | mainTree node1 node2 explorer headers |
mainTree := UbakyeNode header: 'Arbol raÃz' body: ''.
node1 := UbakyeNode header: 'Nodo 1' body: 'Texto 1'.
node2 := UbakyeNode header: 'Nodo 2' body: 'Texto 2'.
mainTree addNode: node1; addNode: node2.
explorer := GLMTabulator new title: (mainTree header). explorer column: #tree; column: #body.
headers := (mainTree children) collect: [:node | node header].
explorer transmit to: #tree; andShow: [:a | a tree display: headers].
explorer openOn: mainTree.
"*************************"
Now I need to make the children sellectable, and that all the contents of the tree can be updated with a shortcut.
I will keep you posted.
Cheers,
Offray
On 07/26/2014 09:01 PM, Offray Vladimir Luna Cárdenas wrote:
Hi again,
I will be using this thread to update my advances and questions about how to build an outliner in Pharo Smalltalk. If there is a better method like starting a new thread for particular questions, or a less narrative style, please let me know.
The idea is to use the tools provided by Moose to build a quick outliner that can be extended to suit my needs on academical writing. This is kind of a strange approach in the sense that I'm not following the tutorials with a predefined problems (make a game and so) but trying to start with a real (in the sense of closer) problem (making an outliner) and to see which knowledge I need to solve this necessity. In that sense is more like the Freire's alphabetization of adults in Brazil.
So, the things I have done so far was to search for a good model to start with. Something already done that can be used as scaffolding for my outliner. The Help System seems like a good start for an outliner (in fact it is already one), so I have taken the Help-Core system and start to use it as a base for my project.
After that I have used the Moose browsers to build a simple interface, as seen here:
http://mutabit.com/offray/__static/blog/output/galleries/__objetos/ubakye-br...
<http://mutabit.com/offray/static/blog/output/galleries/objetos/ubakye-browse...>
The part I want to deal with is this:
===============
explorer := GLMTabulator new title: (mainTree header). explorer column: #tree; column: #body.
explorer transmit to: #tree; andShow: [:a | a tree display: mainTree children ].
explorer openOn: mainTree.
===============
So, instead of "display: mainTree children" I need something that takes the get names (headers) of the two nodes and the contents in the right panel. For that I think that I need to learn some iterators. I have already a "header" method for the nodes. Any clue would be appreciated and I will keep you posted on my advances.
Cheers,
Offray
On 07/21/2014 12:58 PM, Offray Vladimir Luna Cárdenas wrote:
Hi Damien,
Thanks for your answer. Comments below.
On 07/21/2014 11:09 AM, Damien Cassou wrote:
On Sat, Jul 19, 2014 at 2:47 AM, Offray Vladimir Luna Cárdenas <offray@riseup.net <mailto:offray@riseup.net>> wrote:
The first idea that comes to mind is using STON for storage nodes and tree information, so I can interchange it with the flatland files world and keep it readable. Sounds that reasonable?
without more information, it is hard to stay. Try with STON and change if that does not work :-). We have XML and JSON generators as well.
This is a kind of raw preview of I'm talking about:
http://www.enlightenment.org/__ss/e-53cd4f36f021e9.68569046.__jpg
<http://www.enlightenment.org/ss/e-53cd4f36f021e9.68569046.jpg>
Of course in this case, it is just a Help browser with a Playground window over it, but I would like to have something like Playgrounds inside the help browser. I was trying to build a custom browser with Glamour, but seems that Help Browser already has the machinery I'm looking for.
So my first question is how to use the Help Browser class as a template for my outliner class? And can I put a Playground where the plain text is located right now?
The second thing I would like to do is to add pandoc's markdown inside comments, but I don't like the syntax of comments in Smalltalk because single quotes are fairly easy to find in light markup language like markdown. Is difficult to change it to create something more like python (with """) or Lua (with -[]- )?
There is only one syntax for comments in Pharo. Instead of Markdown, you might want to have a look at Pillar which is implemented in Pharo and can generate Markdown (and html, and pdf) : https://github.com/pillar-__markup/pillar-documentation/ <https://github.com/pillar-markup/pillar-documentation/>
I have seen Pillar. Seems really interesting, but Pandocs markdown support academic citation in several formats and I have already long docs wrote on that format integrated in my workflow from Zotero and even there is a growing community working on Scholarly Markdown[1][2] so I would like to stick with it as much as I can for my own writing. That being said. I would like also a better integration between Smalltalk outliners and all the academic publication work flow, including working better with pandoc as a external library.
[1] https://github.com/scholmd/__scholmd/wiki <https://github.com/scholmd/scholmd/wiki> [2]
http://blog.martinfenner.org/__2013/06/29/metadata-in-__scholarly-markdown/
<http://blog.martinfenner.org/2013/06/29/metadata-in-scholarly-markdown/> [3]
http://programminghistorian.__org/lessons/sustainable-__authorship-in-plain-...
<http://programminghistorian.org/lessons/sustainable-authorship-in-plain-text...>
Thanks again, this conversation with people in the community is very valuable to me,
Offray
_________________________________________________ Moose-dev mailing list Moose-dev@iam.unibe.ch <mailto:Moose-dev@iam.unibe.ch> https://www.iam.unibe.ch/__mailman/listinfo/moose-dev <https://www.iam.unibe.ch/mailman/listinfo/moose-dev>
-- www.tudorgirba.com <http://www.tudorgirba.com>
"Every thing has its own flow"