Emilio - The Helpful Style Checking Butler for CouchDB

Emilio is a fairly opinionated style check for Erlang which looks to promote consistency, readability, and portability between developers.

While Emilio attempts to check a large number of common cases of styles it is by no means exhaustive. Just because Emilio does not report any errors does not mean a particular source file has no style issues. Although hopefully the number of checks that do exist will lead developers to making similar decisions in the uncovered situations. And obivously there can always be more checks added in the future.

Building Emilio

$ git clone https://github.com/cloudant-labs/emilio
$ cd emilio/
$ make

This produces a self contained emilio executable which you can then place anywhere on your path as desired.

Usage

Usage: ./emilio [-h <help>] [-l <list>] [-e <explain>] [-c <config>]
                [-i <ignore>] [-j <jobs>] [-f <report_formatter>]
                [-C <context>] [-w <whitelist>]
                [--debug-stats <debug_stats>] path [path ...]

  -h, --help       Show this help message
  -l, --list       List all error codes with a short description
  -e, --explain    Show explanation for the given error code
  -c, --config     The config file to use
  -i, --ignore     Ignore any file path matching the specified glob
  -j, --jobs       Number of files to process in parallel [default: 4]
  -f, --format     Set the output format [default: text]
  -C, --context    Number of contextual lines to display with the text 
                   formatter [default: 0]
  -w, --whitelist  A CSV file of filename,line,column,code reports to 
                   ignore
  --debug-stats    Display timing information with the text formatter
  path             Paths to process, directories are searched recursively

Implemented Checks

• 1xx - Indentation
• 2xx - Whitespace
• 3xx - Blank Lines
• 4xx - Exports and Imports
• 5xx - Line Length
• 6xx - Anti-Patterns

Code	Good	Bad	Description
111	good	bad	Indentation should be a multiple of the configured indentation_count
112	good	bad	Indentation should not increase by more than two levels
120	good	bad	Indentation should match for corresponding tokens
121	good	bad	Indentation should increase one level for clauses
122	good	bad	Indentation for when clauses should increase two levels
123	good	bad	Indentation for line breaks at operators should increase by two levels
124	good	bad	Indentation for line breaks at operators should increase by two levels
125	good	bad	Indentation for clause bodies should increase by at least one level
126	good	bad	Indentation for expression spans should increase by at least two levels
127	good	bad	Clause bodies should be all inline or multiline
130	good	bad	Exports should be indented one level
141	good	bad	Inline funs as arguments should start on the same line as the call
142	good	bad	Inline funs as arguments should end on the last line of the call
201	good	bad	Whitespace should not contain tabs
202	good	bad	Whitespace should not contain carriage returns
203	good	bad	Line endings should not include carriage returns
204	good	bad	Whitespace should not contain other non-printable characters
210	good	bad	Source files should end with a single newline.
221	good	bad	No spaces before commas
222	good	bad	Commas should have a trailing space
223	good	bad	No spaces before semicolons
224	good	bad	No whitespace before dot
225	good	bad	Attribute declarations should start in the first column of the line
230	good	bad	No spaces after left parenthesis
231	good	bad	No spaces before right parenthesis
240	good	bad	No spaces after left square brace
241	good	bad	No space before right square brace
242	good	bad	Pipe operators should have preceding whitespace
243	good	bad	Pipe operators should have trailing whitespace
250	good	bad	No space after left curly brace
251	good	bad	No space before right curly brace
301	good	bad	No more than two consecutive blank lines at module level
302	good	bad	Functions should be separated by two blank lines
310	good	bad	Use zero or one blank lines between function clauses
311	good	bad	Use a consistent number of blank lines between function clauses
401	good	bad	Only put module and behavior attributes before exports
402	good	bad	Do not use imports
410	good	bad	Exports should be grouped correctly
411	good	bad	Do not use arbitrary groups of exports
412	good	bad	Export groups should follow the order of declared behaviors
413	good	bad	Order known behavior callbacks consistently
420	good	bad	Function definitions should follow the same order as exports
421	good	bad	Do not mix private functions with exported functions
501	good	bad	Limit source lines to 80 columns
601	good	bad	Do not re-use underscore prefixed variable names
602	good	bad	Cases should have more than one clause

Adding a Check

To add a new check to emilio you'll need to first implement your check as a module. The existing emilio_check_* modules should provide decent guidance on the implementation. To include the check you should add the module name to the list in emilio.hrl so that its executed.

Once your check is implemented you'll then want to add entries in priv/documentation/ for each error code it generates. Make sure to follow the existing style for description.md so that the README.md generation works correctly.

Exercising Emilio

You can exercise Emilio on a number of open source projects easily by runninng:

make check-projects

The list of projects can be found here.