Also modify dependency generation to be a side-effect of compilation.
It's still not quite perfect in picking up some changes (in particular
just after a commit, no files have changed but the git identification
is now different).
This is most relevant in implied .WORD directives which are caused by an
attempt to call a macro (which happens to be undefined) with arguments
that don't parse as valid expressions.
several compilers:
- gcc version 4.5.3 (NetBSD nb2 20110806)
- gcc version 4.9.2 (Ubuntu 4.9.2-10ubuntu13)
- Ubuntu clang version 3.6.0-2ubuntu1 (tags/RELEASE_360/final) (based on LLVM 3.6.0)
The warnings were mostly about local variables shadowing others,
unused function parameters, and C++ style comments. Some variables were
indeed used before set.
Note that on Linux, using -std=c99 does stupid things like *remove* the
declaration of strdup() from <string.h>. Therefore I've reluctantly used
-std=gnu99.
They behave observably different from generic string parsing and trying
to account for them generically just gets in the way.
.rept is treated the same as a macro.
this is how it should be according to page 6-42 of the manual.
Kermit does this ugly thing to backpatch some .word value:
435 000070 dial$time 35.
1 000070 modval 35.,dial.time
1 .save
2 000070 .psect modinf
3 000006 . = $$current + dial.time
4 000006 000043 .word 35.
5 000070 .restore
436 000070 wake$string <^E^M>
where the current program section already is modinf. So the .save stores
DOT, it's changed to a lower value (6) to backpatch, and .restore sets DOT
back to 70.
The original intent of the code, when faced with something like
.MACRO A
.MACRO B
.ENDM A
is to terminate the outer macro definition immediately.
However some macros redefine themselves, and that broke.
For now, detect self-redefinition and disable the early
end if there is one.
Since this does not match what's described on page 7-3,
this will change if it doesn't break my test cases.
The code for nested and unnested strings was mixed, thereby being too
simple for the nested case. I separated them, which is simpler than
the fixed combined version.
Detected by the macro call
..EMIT <$FAC=^D<fb$rea>>
Also, to make Kermit sources work, only increase the local symbol block
counter if there actually was a local symbol used in the block.
This way, conditional inclusions (which include source text only in the
first pass) will have less potential for de-synchronisation between the
passes. After all, if the generated internal local symbol names do
not match, phase errors will result (showing themselves as strange label
redefinition problems).
