Minimum spanning tree plugin#138
Conversation
…n-plugin:migrate`
builds successfully.
…n org.gephi:gephi-maven-plugin:validate`
anything other than null.
attempting to load test graphs to verify.
…g with multiple graph loading.
equality documentation.
aren't part of a minimum spanning tree.
already uses integers. Might change both test data and implementation back to boolean in the future.
'com.carlschroedl.gephi.plugin.minimumspanningtree'
|
It looks like the only conflict is in the enumeration of plugins in the pom: Is it better for me to resolve this, or for you? |
| * number of edges and nodes. | ||
| * All other node and edge attributes are ignored in this comparison. | ||
| */ | ||
| public class GraphTopologyEquals { |
There was a problem hiding this comment.
I created this because I wanted my plugin's tests to assert that the result of running the algorithm was the same as a manually-verified graph loaded from file. Since I'm new to Gephi, I'm wondering if this is necessary. Are there any alternatives built in to Gephi for determining if graphs have equal topologies?
There was a problem hiding this comment.
I don't think there are builtin methods for this, but this is a small class. Great that you have tests!
There was a problem hiding this comment.
Ok. Cool. Thanks for confirming that I'm not duplicating existing Gephi functionality.
modules/MinimumSpanningTree/cover.sh
Outdated
| @@ -0,0 +1,12 @@ | |||
| #!/bin/bash | |||
There was a problem hiding this comment.
This script is a workaround so that I can view code coverage in NetBeans while developing this plugin. I'm not sure it needs to be included. What do you think?
There was a problem hiding this comment.
This file is not necessary but I can omit it in the merge, no problem.
There was a problem hiding this comment.
Ok. That works. If it's easier for you, I have no problems with (A): git rm-ing the file on my end, or, if you prefer, I can (B): altogether eliminate the file from the history of this branch and force push to update the pull request. Of course, I don't want to rewrite history if you or another Gephi team member have already pulled this branch down somewhere and started tinkering with it :) Please let me know if I should do either (A) or (B) otherwise I will assume you will exclude it from the merge.
|
|
||
| <groupId>com.carlschroedl.gephi</groupId> | ||
| <artifactId>spanning-tree-plugin</artifactId> | ||
| <version>1.0.0</version> |
There was a problem hiding this comment.
Is it ok that this is a release version instead of a SNAPSHOT?
There was a problem hiding this comment.
No problem, your plugin can declare any version number you want. Just future updates will need to increase it.
| <plugin> | ||
| <groupId>org.codehaus.mojo</groupId> | ||
| <artifactId>cobertura-maven-plugin</artifactId> | ||
| <version>2.6</version> |
There was a problem hiding this comment.
Unfortunately, when triggering code coverage collection in the NetBeans UI, it disgregards the version of the coverage plugin that is specified in the pom. Instead, NetBeans uses the most recent stable release of the plugin. In this case, there are integration problems between NetBeans 8.2 and the most recent stable release of the cobertura-maven-plugin (2.7). By running the cover.sh script I included above, a user can at least view their code lines painted in red, yellow, or green depending on test coverage. While you can't trigger the collection of code coverage data via the NetBeans UI, you can still view the results in NetBeans.
| + "<br/> The minimum spanning tree can be visualized by creating a Filter on the Attribute \"" + ST_COL_NAME + "\" with value = \"1\""; | ||
| } | ||
|
|
||
| // <editor-fold defaultstate="collapsed" desc="Inner Helper Classes"> |
There was a problem hiding this comment.
When I was a student I made the unfortunate decision to write many of the following classes as inner classes. I hope to convert these into standalone public classes in a subsequent refactor (carlschroedl#4).
| * workspace or project after loading. | ||
| * @param file The file containing a graph | ||
| * @return the GraphModel | ||
| */ |
There was a problem hiding this comment.
I wrote this method so that I could concisely and reliably load graphs from file into memory for comparison during my plugin's tests. Is there functionality built-in Gephi to accomplish something like this?
|
Thanks for the pull request! I will review it as soon as possible. |
|
Thanks for volunteering to resolve the conflict! |
|
Apologies, this pull request includes several test resources that are not used. The only ones used are:
The following are not used:
I believe it is prudent to delete the unused test resources. Just like I mentioned above, I can do that by either (A): simply |
|
I should mention that I've rewritten history many times in the past with native git as well as with BFG Repo-Cleaner, so there's no need to worry about me accidentally destroying the good parts of the history while eliminating the bad parts. In other words, don't be afraid that I will screw up option (B) and delete the plugin :) |
|
Hi, Actually, remove unnecessary files if you can please. If the files are just small examples or similar we can merge them. By the way, it's possible to include demo screenshots for your plugin that will appear in the new plugins site. Just drop them in src/img folder like https://github.com/gephi/gephi-plugins/tree/geolayout-plugin/modules/GeoLayout/src/img |
|
Thanks! I will rewrite history to remove the unnecessary files. I didn't know about the |
|
The earliest I can resume work on this pull request is January 5th. |
073b440 to
d6ac100
Compare
|
Closing this in case there are stale refs. Have opened #140 for the cleaned history. |
Hello Gephi team! Thanks for the excellent instructions on how to contribute plugins in version 0.9. This pull request upgrades my 0.8-compatible plugin (formerly called SpanningTree) to be compatible with 0.9. Please let me know what you think!