#226 closed enhancement (fixed)
Generate list of symbols for autoloading at build-time
| Reported by: | ehuelsmann | Owned by: | ehuelsmann |
|---|---|---|---|
| Priority: | major | Milestone: | |
| Component: | build | Version: | |
| Keywords: | Cc: | ||
| Parent Tickets: |
Description
Currently, our list of symbols to be set up for auto-loading is
a manual maintenance effort which makes ABCL fragile with respect
to refactoring.
My idea would be to have an auto-collected list, with two (hand-coded) lists: one to include additional symbols and the other to exclude symbols (e.g. to reduce the size of the symbol list).
Change History (8)
comment:1 by , 14 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
comment:2 by , 14 years ago
| Resolution: | fixed |
|---|---|
| Status: | closed → reopened |
comment:3 by , 14 years ago
comment:4 by , 14 years ago
comment:5 by , 14 years ago
Next step:
Move EXPORT forms to the respective source files
defining the function and macro bindings.
comment:6 by , 14 years ago
We're nearly there: I noticed we don't detect structure slot accessors when
compiling the macro expansion of a defstruct, due to its use of SETF SYMBOL-FUNCTION.
Expanding to FSET and detecting that at compile-time would solve it.
comment:7 by , 14 years ago
| Resolution: | → fixed |
|---|---|
| Status: | reopened → closed |
comment:8 by , 14 years ago
FSET detection can always be added later. For now, I'm assuming structures will be created before being accessed, meaning I'll rely on timely initialized accessors for now.
(In [14035]) Fix #226 (Invocation of an undefined function in a fresh ABCL crashes):
make sure the PRINT-OBJECT generic function exists
before defining methods on it.