Code Maat Importer

The Code Maat importer turns the coupling analysis of Code Maat into a cc.json file. You see which files change together: every coupled pair becomes an edge carrying pairingRate and avgCommits. Code Maat reads version control history, so this works for code in any language.

Use it when you want to find hidden dependencies, such as two files in different folders that always change in the same commit.

Quick start

# 1. Write the git history in the format Code Maat's "git" parser reads
git log --pretty=format:'[%h] %an %ad %s' --date=short --numstat > project.log

# 2. Let Code Maat compute the coupling (or run it with java -jar code-maat-<version>-standalone.jar)
maat -c git -l project.log -a coupling > edges.csv

# 3. Import the coupling and turn the edge attributes into metrics on every file
ccsh codemaatimport edges.csv -o edges.cc.json
ccsh edgefilter edges.cc.json.gz -o visual_edges.cc.json

# 4. Merge the result with a map of the code itself
ccsh unifiedparser . -o metrics.cc.json
ccsh merge visual_edges.cc.json metrics.cc.json.gz -o merged.cc.json

Open merged.cc.json.gz in the Web Studio.

Options

Option	Description
FILE...	one or more Code Maat coupling CSV files, each with its header
-o, --output-file=<outputFile>	output file; prints to stdout when omitted
-nc, --not-compressed	write plain JSON instead of a gzip file
-h, --help	show the help and exit

Usage: ccsh codemaatimport [-h] [-nc] [-o=<outputFile>] FILE...

Metrics

Both metrics are edge metrics. They sit on the edge between two files.

Metric	Code Maat column	Type	Description
pairingRate	degree	relative	how often, in percent, the two files are committed together
avgCommits	average-revs	absolute	average number of commits the two files appear in

The entity and coupled columns name the two files. Every file named there becomes a node of the map, even when it carries no metric yet.

Example

A coupling CSV from Code Maat looks like this:

entity,coupled,degree,average-revs
app/codeCharta.html,app/codeCharta.scss,56,10
app/testVille.html,app/codeCharta.html,42,8

ccsh codemaatimport writes each row as an edge in the dependency lens (excerpt):

"dependency": {
  "edges": [
    { "fromId": "6b185a8cc9433de8", "toId": "d515f754245533a8", "attributes": { "pairingRate": 56, "avgCommits": 10 } },
    { "fromId": "4c6b71df479538b7", "toId": "6b185a8cc9433de8", "attributes": { "pairingRate": 42, "avgCommits": 8 } }
  ],
  "attributeTypes": { "pairingRate": "relative", "avgCommits": "absolute" }
}

After ccsh edgefilter, app/codeCharta.html, which takes part in both pairs, carries pairingRate 49 (the average) and avgCommits 18 (the sum). See the Edge Filter for how it aggregates.

Related

• Edge Filter: turn edge attributes into metrics on every file
• Merge Filter: combine the coupling with other maps
• Git Log Parser: coupling and churn metrics straight from git, without Code Maat