[Pharo-project] Investigating PluggableTextMorph - Moose to the rescue?
I am trying to wrap my mind around PluggableTextMorph and its internals. I am using the latest 1.4, where there has been some cleanup done (yay!) So, PluggableTextMorph, TextMorphForEditView, Text, Paragraph, TextComposer, CharacterScanner, Stylers (in Shout), TextAttributes etc. I need to understand this for my little game based on text and attributes (there is nothing like a 'real' project to cut my teeth on a myriad of issues). I wondered if I could make use of Moose for that. There are things like in setText that use the textMorphClass instance variable and I am not sure I can get structural information from parametrized selectors and classnames done that way. Would Moose be helpful there? Or am I better doing a Debug It (like I am doing now) and drawing Class and Interaction diagrams ? Truth be told I am a VAR of Sparx Enterprise Architect (and heavy user). Maybe is it possible to have a bridge between Pharo and Sparx EA to get some documentation done easily. I can import metadata through scripting. So, remixing FAMIX/Moose/Pharo and some investigations may end up providing some nice figures to help with PBE2. So, tell me what you think and/or where I should go. There are no reverse engineering tools for pharo but these should be easy to build for a ton of things. Has anybody done an XMI exporter for Pharo classes? Phil
phil@highoctane.be wrote:
Truth be told I am a VAR of Sparx Enterprise Architect (and heavy user). Maybe is it possible to have a bridge between Pharo and Sparx EA to get some documentation done easily. I can import metadata through scripting. So, remixing FAMIX/Moose/Pharo and some investigations may end up providing some nice figures to help with PBE2.
So, tell me what you think and/or where I should go.
There are no reverse engineering tools for pharo but these should be easy to build for a ton of things. Has anybody done an XMI exporter for Pharo classes?
Phil
There is a mention of XMI at http://www.moosetechnology.org/about/history but it seems this has been removed at some point. I had previously asked a similar question here... http://forum.world.st/forward-engineering-UML-models-into-Smalltalk-code-td3... http://comments.gmane.org/gmane.comp.lang.smalltalk.pharo.devel/52388 ...with only a few answers. You actually seem to be looking for the opposite of what I wanted - moving a model into Pharo rather than extracting one out. But you may anyway be interested.... that I then noticed that when Enterprise Architect opened its EAP model file, it created an LDB file next to the EAP file - which was a lot like Microsoft Access. So if you rename the EAP extension to MDB you can open the file in Microsoft Access. You can read MDB files into Pharo using ConfigurationOfODBC as described http://blog.openinworld.com/2011/12/pharo-odbc-working-on-windows-7/ After that, by using Microsoft Access to browse the raw EA model, I worked out which data I needed to load the data into Pharo. Its very rough since this was just a hack for my own use, but I have uploaded it to http://ss3.gemstone.com/ss/Steap.html in case it is of interest. I had hacked some stuff to convert that data into actual classes but I've lost that for the moment. btw, it may take a few minutes to load a large model, during which the image locks. cheers -ben
participants (2)
-
Ben Coman -
phil＠highoctane.be