| Age | Commit message (Collapse) | Author |
|---|
|
We were expanding the path of files to be loaded at the wrong place in
the wrong way. This should be better and address bug #196.
|
|
An attempt to address #194.
|
|
`shell-command` assumes Bourne-shell-compatible quoting, which
doesn't work when the user isn't using a Bourne-compatible shell.
Instead of futzing about with quoting, we can just use `process-lines`
to execute a process and pass it arguments directly.
|
|
Let's see if i finally got this right...
|
|
We could probably be even more robust, but i am being a bit lazy
instead.
|
|
Fixes #95. This is @kovrik's patch, with 80-columns max formatting.
|
|
|
|
I kind of dislike completion on symbols, because a quote reads to me as
'stop evaluating', and a symbol per se has infinite possible
conversions. But, on the other hand, not completing has no practical
advantage, and, moreover, we're already completing symbols inside quoted
lists (e.g. try M-TAB next to `'(defi`)), so my prejudices are not even
consistent. So here we go!
|
|
Move general indentation rules to "geiser-syntax".
|
|
|
|
Use this function instead of repeating the same code in each
implementation.
|
|
Apparently, the nesting level returned by emacs's syntax parser can be
negative (presumably when it gets confused), and we were not avoiding
calling backward-up-list when that happened.
Could or could not address issue #41...
|
|
|
|
Follow the convention for `with-' procedures.
|
|
Guile's `define-once' allows defining a variable only once, but its
syntax is different from `define', so its highlight is different.
|
|
And, if you happen to be launching it all the time, a way of skipping
them via a customizable variable.
Should address issue #15.
|
|
|
|
We add the paths in geiser-guile-load-path also to %load-compiled-path,
and new directories added to the load path via geiser-add-to-load-path
are added to both %load-path and %load-compiled-path.
Here's hope Ludovic will like all these additions!
|
|
|
|
|
|
|
|
By default, keywords are now not fontified in Scheme buffers unless
they have the correct (lower) case. This behaviour can be altered by
new, per-implementation customization variables.
Thanks to Diogo F. S. Ramos for pointing this out.
|
|
Splitting better the specially indented forms between our two
implementations, so that users of a single one don't get weird
indentations for froms without a special meaning in their scheme.
Ideally, we should make these indentation rules buffer-local, so that
when a user is in a, say, Guile buffer, module+ has no special
indentation (as is the case now if that user also has activated
support for Racket).
|
|
In my debian machine, the info nodes for guile live in the "guile-2.0"
node, rather than plain "guile". A new customizable variable,
geiser-guile-manual-lookup-nodes, lets now specify additional names,
and we only add indexes to the info-lookup mode definition when the
node actually exists.
|
|
If we didn't find a define-module form after the cursor, or an
enclosing R6RS library form, we search forward for a module
definition. That way, things like C-c C-a work also from the top of
the file.
|
|
Thanks Jon!
|
|
The new custom variable, geiser-guile-load-init-file-p, will be gone
once Guile adquires the ability to specify the path to its init file.
|
|
|
|
|
|
This reverts commit 801422d1558f488059ede4f9abab5163ca610900.
We cannot blindly substitute make-symbol for intern in the scheme
reader, because we rely on symbol equality elsewhere, often. The fix
will have to be much more careful.
|
|
We were calling `intern' instead of `make-symbol', polluting emacs'
obarray.
|
|
Was a real bug, after all, and quite reproducible. Sending an ,use
metacommand was not returning a prompt, and we were waiting for ever
to start (or almost). Now, connect-to-guile is not only right, but
spiffy again.
|
|
Today, W was seeing errors when connecting to Guile, which of course
immediately disappeared when we tried to reproduce them and get some
logs. I'm logging Guile's initialisation unconditionally, to make sure
the problem doesn't repeat. Much easier than fixing the bug.
|
|
|
|
|
|
|
|
Spinning up from correct fontification of [else in this brave Racket
world.
I'm keeping the list of extra keywords lean and mean, but making it
customizable in both Racket and Guile.
|
|
Inferior schemes weren't really a good idea, were they? With remote
connections one can launch an external scheme to debug Geiser anyway.
And everything is (ahem, will be) simpler when we add new
implementations.
|
|
Separate connections for the REPL and Geiser commands was kind of
neat, but it had the problem of synchronising the current namespace
for both connections. A quick fix would have been to ask the scheme
for the current namespace for every Geiser command in the REPL, but
that, besides clunky, would add potentially prohibitive overhead for
(real) remote connections.
As it happens, using a single connection turned out to be not that
difficult and relatively clean code-wise. We could even turn back to
not use inferior schemes, and the net result of this refactoring would
be the replacement of comint-redirect (which wasn't able to match the
whole EOT token if it didn't arrive all at once) by transaction queues
(which also makes geiser-connection's implementation cleaner).
But using an inferior scheme has a dog-food value, and allows external
processes to connect to the scheme being used by Geiser without
further ado, which could be useful for debugging (although this is a
lame excuse: nothing prevents you from starting a REPL server from
emacs if you want). We'll see.
|
|
|
|
|
|
Or the importance of EOL. Switching to a transaction queue for
communication with the Scheme process means that i had to care about
sending eols in the queries... Guile was waiting for ever reading a
metacommand taking a variable number of arguments. Argh: this has
taken me a few hours -- i'm getting old.
|
|
|
|
|
|
Conflicts:
elisp/geiser-guile.el
|
|
|
|
|
|
|
|
|
|
|
