Copyright | (c) 2014, Andrew G. Seniuk |
---|---|
License | BSD-style (see the file LICENSE) |
Maintainer | Andrew Seniuk <rasfar@gmail.com> |
Stability | provisional |
Portability | portable |
Safe Haskell | Safe-Inferred |
Language | Haskell2010 |
Details concerning the seqaid.config
file.
The purpose of the seqaid.config
file is to control
the extent of coverage of the automatically injected
instrumentation harness.
At this early stage, additional Cabal flags and even some
per-module CPP switches are used to further vary the
behaviour of seqaid. These alternatives will become
documented as they find a reflection in the seqaid.config
.
Also, see Seqaid.Ann for a few user annotations, which
will also be deprecated soon...
The format of seqaid.config
is a seqence of lines, which
can be one of three kinds:
- Blank lines (containing only whitespace)
- Comment lines (first non-whitespace character is
#
) - Configuration lines
Configuration lines begin with a keyword, and are followed by one or more arguments. Arguments are comma-separated. Layout is used when keywords generate nested structure.
For instance, here is the seqaid.config
file from the
leaky package:
package leaky module Main binds duty types Types.TA instances Types.TA, Types.TB, Types.TC
(The extra whitespace after the keywords is purely cosmetic.)
More documentation is pending, but for additional explanations about this particular example, please refer to this document.