No description, website, or topics provided.
Ruby
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
lib
spec
.gitignore
README.rdoc
Rakefile

README.rdoc

delta-flora

DESCRIPTION:

Delta-flora is set of classes and functions which enable interactive analysis of Ruby code histories in an interactive Ruby shell.

It consists of a class named Repository which builds up an intermediate representation of all of the histories of all Ruby class and method changes in a git repo.

Here is an example use of delta-flora in irb:

2.0.0p0 :001 > load 'repository.rb'
2.0.0p0 :002 > es = Repository.new('/Users/joe-shmoe/Projects/rails').events

The value held in es after this call is an array of `method events'. Each method event describes an occurrence in the lifetime of a method. There are three types of method event: added, changed, deleted. Regardless of their type, each event contains the following information:

  • type - added, changed, or deleted

  • commit - sha1 of the git commit for the code change

  • date - date of the commit

  • file_name - name of the file containing the method

  • committer - author of the commit

  • class_name - name of the class containing the method

  • method_name - name of the method (fully-qualified by class and module name)

  • start_line - start line of the method at that commit

  • end_line - end line of the method at that commit

The file analytics.rb contains a set of functions that can be used to analyze histories. Here is function which produces a frequency histogram of the classes by the number of methods they contain:

def class_method_count_freq es
  es.group_by(&:class_name)
    .map {|_,v| v.map(&:method_name).uniq.count }
    .freq
end

The method takes the list of method events and groups them into bins by the names of the classes they reside upon. It then takes the names of all of the methods that have existed over the history of that class and counts them. Afterward, it uses an array extension method named 'freq' to produce frequencies of those counts.

Let's do that analysis:

2.0.0p0 :003 > load 'analytics.rb'
2.0.0p0 :004 > class_method_count_freq(es)

Here are the results for Rails at the time of this writing. The format for the histogram is bin-number followed by frequency:

=> [[1, 1393], [2, 937], [3, 576], [4, 442], [5, 371], [6, 253], [7, 208], [8, 176], [9, 149], [10, 115], [11, 101], [12, 86], [13, 84], [14, 74], [15, 51], [16, 42], [17, 43], [18, 51], [19, 37], [20, 35], [21, 40], [22, 24], [23, 24], [24, 17], [25, 20], [26, 20], [27, 19], [28, 18], [29, 13], [30, 17], [31, 15], [32, 20], [33, 18], [34, 10], [35, 12], [36, 7], [37, 7], [38, 14], [39, 12], [40, 10], [41, 9], [42, 4], [43, 4], [44, 5], [45, 11], [46, 1], [47, 7], [48, 6], [49, 4], [50, 3], [51, 4], [52, 5], [53, 3], [54, 5], [55, 4], [56, 4], [57, 6], [58, 6], [59, 4], [60, 2], [61, 4], [62, 1], [63, 1], [64, 2], [67, 3], [68, 1], [71, 3], [72, 1], [73, 2], [74, 3], [75, 2], [76, 1], [78, 1], [79, 1], [80, 2], [81, 1], [82, 2], [83, 3], [84, 2], [87, 3], [88, 2], [90, 2], [91, 1], [92, 3], [94, 1], [95, 1], [96, 1], [98, 2], [101, 2], [103, 1], [105, 1], [109, 1], [111, 2], [112, 3], [115, 2], [116, 2], [117, 1], [118, 1], [119, 1], [122, 1], [125, 1], [129, 2], [131, 1], [132, 1], [135, 1], [137, 1], [138, 1], [139, 1], [144, 1], [146, 1], [155, 1], [156, 1], [166, 1], [181, 1], [192, 2], [196, 1], [199, 1], [208, 1], [210, 1], [235, 1], [253, 1], [268, 1], [311, 1], [342, 1], [410, 1]]

Method events are created from the repository using the –topo-order flag on git log. This ordering puts branches one after another rather than using strict date ordering. This allows us to do simple analyses like seeing how method lengths have changed over time without the complications that we would have with strict date ordering. Although branch information is disgarded in this 'linear' history, you can expect 'runs' of events within branches to be date ordered.

USAGE:

Use of delta-flora is easy. The steps in the description should get you started. But, it's important to note that the first time you run delta-flora on a repo it takes considerable amount of time rip the repo and produce method events.

To make use easier, the Repository class has been designe to check for a file named methodevents.csv in the repository directory. If it exists and there are no commits with a later datestamp in the repo, then methodevents.csv is assumed to be current and it is loaded. If methodevents.csv does not exist or it is out of date, Repository rips the repo and produces a methodevents.csv file.

Since 'es = Repository.new('some repostory path').events is a bit of a mouthful, delta-flora supplies a convenience method that has the same effect:

es = load_events('some repository path')

NAMING:

`Delta Flora' is the name of an album by Hughscore - a group formed by the late Hugh Hopper of Soft Machine. The name struck me because it literally means the flowering/bountiful mouth of a river. That seemed like a good name for a tool that produces useful information from repositories. Aside from that, the pun on the word `delta' with regard to version control systems was too good to pass up.

REQUIREMENTS:

  • Ruby 2.0

LICENSE:

(The MIT License)

Copyright © 2015 Michael Feathers

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.