Avery Pennarun
39ef065443
Refactor all.do and test.do in various directories.
...
Now 'redo test' runs the tests, but 'redo t' just builds the programs.
Also removed wvtest stuff; we're not really using it properly anyway and
it's not helping our testing right now. It might come back later.
2010-11-21 05:47:48 -08:00
Avery Pennarun
27407f8f8e
t/all.do: make sure the subdir all.do's always run.
...
...not just when their dependencies are dirty. Some of them want to do some
explicit checks of redo's behaviour.
2010-11-21 05:39:00 -08:00
Avery Pennarun
b19a918894
Test for the previous bugfix.
...
This fails if you make test *twice* without the preceding patch.
Unfortunately I couldn't find a good way to make it fail if you only make
test once.
2010-11-21 04:41:03 -08:00
Avery Pennarun
2f5814c0fe
Add curse/all to t/all, now that it passes.
2010-11-19 07:33:45 -08:00
Avery Pennarun
eae3e7cdef
Add t/example, a basic example build environment suitable for a tutorial.
2010-11-18 22:48:46 -08:00
Avery Pennarun
b9853d3858
minimal/do: handle nonzero exit codes and set $1/$2 correctly.
...
Also add some tests to confirm this.
2010-11-17 19:07:27 -08:00
Avery Pennarun
c7585558ef
If the .do script deletes $3, don't die.
2010-11-17 17:55:16 -08:00
Avery Pennarun
c1f09f564b
Support for default.*.do rules.
...
I *think* this was the last missing part from djb's spec. Certainly it's an
important one for any real project.
2010-11-16 03:04:11 -08:00
Avery Pennarun
7505048093
redo-if*.py: import fewer things from helpers.
...
Just to keep track of how many helper functions we actually are using. In
case I get brave and try to convert to C sometime.
2010-11-13 01:40:01 -08:00
