-*- mode: org -*-
* Overview
Geiser is a generic Emacs/Scheme interaction mode, featuring an
enhanced REPL and a set of minor modes improving Emacs' basic scheme
major mode. The main functionalities provided are:
- Evaluation of forms in the namespace of the current module.
- Macro expansion.
- File/module loading.
- Namespace-aware identifier completion (including local bindings,
names visible in the current module, and module names).
- Autodoc: the echo area shows information about the signature of
the procedure/macro around point automatically.
- Jump to definition of identifier at point.
- Access to documentation (including docstrings when the
implementation provides it).
- Listings of identifiers exported by a given module.
- Listings of callers/callees of procedures.
- Rudimentary support for debugging (list of
evaluation/compilation error in an Emacs' compilation-mode
buffer).
- Support for inline images in schemes, such as Racket, that treat
them as first order values.
NOTE: if you're not in a hurry, Geiser's website
(http://www.nongnu.org/geiser/) contains a nicer manual, also
included in Geiser tarballs as a texinfo file (doc/geiser.info).
* Requirements
Geiser needs Emacs 23.2 or better, and at least one of the
supported scheme implementations:
- Guile 2.2 or better.
- PLT Racket 6.0 or better.
- Chicken 4.8.0 or better.
- MIT/GNU Scheme 9.1.1 or better.
- Chibi 0.7.3 or better.
- Chez 9.4 or better.
* Installation
The easiest way is to use MELPA, and just type
=M-x package-install RET geiser= inside emacs.
Geiser can be used either directly from its uninstalled source tree
or byte-compiled and installed after performing the standard
configure/make/make install dance. See the INSTALL file for more details.
** Chicken Addendum
These steps are necessary to fully support Chicken Scheme, but are
not required for any other scheme.
- Install the necessary support eggs:
#+BEGIN_EXAMPLE
$ chicken-install -s apropos chicken-doc
#+END_EXAMPLE
- Update the Chicken documentation database:
#+BEGIN_EXAMPLE
$ cd `csi -p '(chicken-home)'`
$ curl http://3e8.org/pub/chicken-doc/chicken-doc-repo.tgz | sudo tar zx
#+END_EXAMPLE
* Basic configuration
The loading invocations above install all supported Scheme
implementations. You can list explicitly the ones that you want by
setting the variable =geiser-active-implementations= *before* loading
geiser.el. For instance:
#+BEGIN_SRC elisp
(setq geiser-active-implementations '(chicken guile))
#+END_SRC
On opening a scheme file, Geiser will try to guess its Scheme,
defaulting to the first in the list. Use =C-c C-s= to select the
implementation by hand (on a per file basis).
Check the geiser customization group for some options with:
#+BEGIN_EXAMPLE
M-x customize-group RET geiser RET
#+END_EXAMPLE
In particular, customize =geiser-<impl>-binary=, which should point
to an executable in your path.
To start a REPL, =M-x geiser=.
** Completion with company-mode
Geiser offers identifier and module name completion, bound to
=M-TAB= and =M-`= respectively. Only names visible in the current
module are offered.
While that is cool and all, things are even better: if you have
[[http://company-mode.github.io/][company mode]] installed, Geiser's completion will use it. Just
require company-mode and, from then on, any new scheme buffer or
REPL will use it.
* Quick key reference
** In Scheme buffers:
|---------------+--------------------------------------------------|
| =C-c C-s= | Specify Scheme implementation for buffer |
| =C-c C-z= | Switch to REPL |
| =C-c C-a= | Switch to REPL and current module |
|---------------+--------------------------------------------------|
| =M-.= | Go to definition of identifier at point |
| =M-,= | Go back to where M-. was last invoked |
| =C-c C-e m= | Ask for a module and open its file |
| =C-c C-e C-l= | Add a given directory to Scheme's load path |
| =C-c C-e [= | Toggle between () and [] for current form |
|---------------+--------------------------------------------------|
| =C-M-x= | Eval definition around point |
| =C-c C-c= | Eval definition around point |
| =C-c M-e= | Eval definition around point and switch to REPL |
| =C-x C-e= | Eval sexp before point |
| =C-c C-r= | Eval region |
| =C-c M-r= | Eval region and switch to REPL |
| =C-c C-b= | Eval buffer |
| =C-c M-b= | Eval buffer and switch to REPL |
|---------------+--------------------------------------------------|
| =C-c C-m x= | Macro-expand definition around point |
| =C-c C-m e= | Macro-expand sexp before point |
| =C-c C-m r= | Macro-expand region |
|---------------+--------------------------------------------------|
| =C-c C-k= | Compile and load current buffer |
| =C-c C-l= | Load scheme file |
| =C-u C-c C-k= | Compile and load current buffer, restarting REPL |
|---------------+--------------------------------------------------|
| =C-c C-d d= | See documentation for identifier at point |
| =C-c C-d s= | See short documentation for identifier at point |
| =C-c C-d i= | Look up manual for identifier at point |
| =C-c C-d m= | See a list of a module's exported identifiers |
| =C-c C-d a= | Toggle autodoc mode |
|---------------+--------------------------------------------------|
| =C-c <= | Show callers of procedure at point |
| =C-c >= | Show callees of procedure at point |
|---------------+--------------------------------------------------|
| =M-TAB= | Complete identifier at point |
| =M-`=, =C-.= | Complete module name at point |
| =TAB= | Complete identifier at point or indent |
| | (If =geiser-mode-smart-tab-p= is t) |
|---------------+--------------------------------------------------|
** In the REPL
|----------------+----------------------------------------------------|
| =C-c C-z= | Start Scheme REPL, or jump to previous buffer |
| =C-c M-o= | Clear scheme output |
| =C-c C-q= | Kill Scheme process |
| =C-c C-l= | Load scheme file |
| =C-c C-k= | Nuke REPL: use it if the REPL becomes unresponsive |
|----------------+----------------------------------------------------|
| =M-.= | Edit identifier at point |
| =TAB=, =M-TAB= | Complete identifier at point |
| =M-`=, =C-.= | Complete module name at point |
| =M-p=, =M-n= | Prompt history, matching current prefix |
|----------------+----------------------------------------------------|
| =C-c C-m= | Set current module |
| =C-c C-i= | Import module into current namespace |
| =C-c C-r= | Add a given directory to scheme's load path |
|----------------+----------------------------------------------------|
| =C-c C-d C-d= | See documentation for symbol at point |
| =C-c C-d C-m= | See documentation for module |
| =C-c C-d C-a= | Toggle autodoc mode |
|----------------+----------------------------------------------------|
** In the documentation browser:
|--------------+----------------------------------------------|
| =f= | Next page |
| =b= | Previous page |
|--------------+----------------------------------------------|
| =TAB=, =n= | Next link |
| =S-TAB=, =p= | Previous link |
| =N= | Next section |
| =P= | Previous section |
|--------------+----------------------------------------------|
| =k= | Kill current page and go to previous or next |
| =g=, =r= | Refresh page |
| =c= | Clear browsing history |
|--------------+----------------------------------------------|
| =.=, =M-.= | Edit identifier at point |
| =z= | Switch to REPL |
|--------------+----------------------------------------------|
| =q= | Bury buffer |
|--------------+----------------------------------------------|
** In backtrace (evaluation/compile result) buffers:
- =M-g n=, =M-g p=, =C-x `= for error navigation.
- =q= to bury buffer.