| Age | Commit message (Collapse) | Author | Files | Lines | |
|---|---|---|---|---|---|
| 2010-04-24 | Fix how the source directory is defined. | Daniel Burrows | 1 | -1/+1 | |
| A dict apparently can't be appended, but a tuple (var, value) works. | |||||
| 2010-04-24 | Directly depend on .o files to eliminate the scons warning. | Daniel Burrows | 1 | -1/+1 | |
| Apparently scons is happy to depend on the output of another environment; the problem was that I was implicitly defining a second build rule for the .o in a new environment. | |||||
| 2010-04-24 | Start adding extra dependencies into src/ to the test build files. | Daniel Burrows | 1 | -2/+11 | |
| This causes a problem (currently nonfatal but should be fixed), because scons doesn't like that I ask for some .cc files with "different" environments. Need to look into how best to fix this. | |||||
| 2010-04-24 | Define SRCDIR for the Boost test. | Daniel Burrows | 1 | -0/+3 | |
| 2010-04-24 | Deal with the fact that BOOST_TEST_NO_MAIN might be defined by a compiler flag. | Daniel Burrows | 1 | -0/+2 | |
| 2010-04-24 | Try to build and run unit tests, and only build aptitude by default. | Daniel Burrows | 1 | -0/+39 | |
| TODO: probably should add a UnitTest builder. | |||||
| 2010-04-21 | Fix a second type error in the parsers code. | Daniel Burrows | 1 | -1/+1 | |
| 2010-04-21 | Fix type errors caused by a change to the return type of fusion::make_vector. | Daniel Burrows | 1 | -3/+10 | |
| 2010-04-21 | Fix the unit tests to #include <loggers.h> for the new namespace. | Daniel Burrows | 2 | -0/+4 | |
| 2010-04-21 | Rename the log4cxx namespace to logging. | Daniel Burrows | 5 | -22/+22 | |
| There are various problems with log4cxx, especially that it seems to be unreliable and leads to random crashes. It might be a good idea to eventually move to a different framework or roll our own. Decreasing the number of explicit dependencies on log4cxx is a good way to do that. | |||||
| 2010-04-20 | Allow "conflict" to be specified as a safety level again, with its old ↵ | Daniel Burrows | 1 | -0/+12 | |
| behavior, and fix setting safety levels using the old names. aptitude's behavior should now be fully compatible with pre-0.6.2 releases (by default). The safety level change fixes the fact that you could raise safety levels using the old names but not lower them. | |||||
| 2010-04-19 | Test the planned new convention around NULL expressions (they're treated as ↵ | Daniel Burrows | 1 | -0/+97 | |
| "true"). Currently this crashes. | |||||
| 2010-04-19 | Write tests of the "not" expression. | Daniel Burrows | 1 | -0/+58 | |
| 2010-04-19 | Write out tests of "or" that are dual to the tests of "and". | Daniel Burrows | 1 | -0/+276 | |
| 2010-04-19 | Add tests checking what happens if one input to a doubleton is *lowered*. | Daniel Burrows | 1 | -0/+81 | |
| 2010-04-19 | Un-swap the first and second case of the doubleton test, and add a test that ↵ | Daniel Burrows | 1 | -1/+59 | |
| raising just one input to an AND has no effect. | |||||
| 2010-04-19 | Expect the child_modified call that comes from an "and" expression to refer ↵ | Daniel Burrows | 1 | -4/+4 | |
| to the expression itself, not to the child that triggered the call. | |||||
| 2010-04-19 | Print the child that was modified when printing a child_modified_call object. | Daniel Burrows | 1 | -1/+2 | |
| 2010-04-19 | Add the missing registration call for the new test suite. | Daniel Burrows | 1 | -0/+2 | |
| 2010-04-19 | Add a first test of "and" over two elements. | Daniel Burrows | 1 | -0/+33 | |
| 2010-04-19 | Actually run the new tests. | Daniel Burrows | 1 | -0/+6 | |
| 2010-04-19 | Add some basic tests of "and" expressions. | Daniel Burrows | 1 | -0/+91 | |
| 2010-04-19 | Add tests of the expression weak reference subsystem. | Daniel Burrows | 1 | -0/+33 | |
| 2010-04-19 | Start a suite of unit tests for the incremental expression module of the ↵ | Daniel Burrows | 2 | -0/+273 | |
| resolver. | |||||
| 2010-04-18 | Implement the test that placing a cost on breaking a soft dependency has an ↵ | Daniel Burrows | 1 | -2/+84 | |
| effect. | |||||
| 2010-04-18 | Add a utility routine to make it simpler to test resolver runs, by viewing ↵ | Daniel Burrows | 1 | -0/+24 | |
| their output as a list. | |||||
| 2010-04-17 | Fix the test of the temporary file system to match the new behavior. | Daniel Burrows | 1 | -4/+4 | |
| All you can count on about the new temporary file name now is that it starts with the prefix. | |||||
| 2010-04-09 | Terminology change: replace "tier operation" with "cost" everywhere. | Daniel Burrows | 3 | -216/+218 | |
| The only remaining references to tiers are in backwards-compatibility code; the "safety" cost component is composed of several "tiers" whose values can be configured. | |||||
| 2010-04-09 | Restructure the source tree to move tier_ related code into files whose name ↵ | Daniel Burrows | 3 | -5/+5 | |
| is based on "cost" instead. | |||||
| 2010-04-08 | Yank out the "tier" type and just use operations instead. | Daniel Burrows | 1 | -134/+1 | |
| Although the concept of a "tier" is useful (to distinguish changes in the tier from the things being changed), the actual data type isn't needed. We can just store an operation that computes the tier of a step given a minimal tier. Throwing out tier objects yields several benefits. It makes the code somewhat simpler by reducing the number of complex object types floating around. It will make it easier to clean up the nomenclature by talking about costs instead of tiers. And it will make it easier to impose a "cost ceiling", just by virtue of making the cost-tracking subsystem simpler. | |||||
| 2010-04-08 | Add a placeholder for a future test that the break-soft-dep cost does the ↵ | Daniel Burrows | 1 | -0/+6 | |
| right thing. | |||||
| 2010-04-08 | Fix the tests to compile with the new resolver constructor signature. | Daniel Burrows | 1 | -13/+39 | |
| 2010-04-04 | Nomenclature change: manyOne -> manyPlus, sepByOne -> sepByPlus, skipManyOne ↵ | Daniel Burrows | 1 | -26/+26 | |
| -> skipManyPlus The old nomenclature was based on the Haskell library Parsec, which in turn draws on idioms from that community. But C++ has no such idiom, and the literal text of the function name makes it sound like it recognizes only one instead of *at least* one. Change to "plus", which I think is much clearer (and of course references the EBNF suffix "+" operator). | |||||
| 2010-04-04 | Fix the expected return values in the cost settings parse test. | Daniel Burrows | 1 | -2/+2 | |
| 2010-04-04 | Implement a new system of "container parsers" that allows much more flexible ↵ | Daniel Burrows | 1 | -10/+239 | |
| specification of parser that recognize various types of sequences. | |||||
| 2010-03-21 | Add an explicit test that trailing separators don't confuse the parse library. | Daniel Burrows | 1 | -0/+12 | |
| 2010-03-21 | Add a test verifying that "max" can't be used as a cost component name. | Daniel Burrows | 1 | -1/+2 | |
| 2010-03-21 | Write a parser that fails unconditionally. | Daniel Burrows | 1 | -0/+21 | |
| 2010-03-21 | More diagnostic information when a fail test fails (also output the ↵ | Daniel Burrows | 1 | -2/+9 | |
| structure it supposedly parsed to). | |||||
| 2010-03-21 | Be more informative about which input didn't pass the test in the test of ↵ | Daniel Burrows | 1 | -1/+14 | |
| things that should fail to parse. | |||||
| 2010-03-21 | With the new operator== and operator<<, I can use CPPUNIT_ASSERT_EQUALS and ↵ | Daniel Burrows | 1 | -69/+28 | |
| get a more sensible error message. | |||||
| 2010-03-20 | Capture parse exceptions at the main entry point to the library and turn ↵ | Daniel Burrows | 1 | -24/+12 | |
| them into a typed return value using boost::variant. | |||||
| 2010-03-20 | Write code to test the error position reporting and parsing Boost ranges. | Daniel Burrows | 1 | -0/+42 | |
| The Boost range parsing is now a free function instead of a member of parser_base. Without some more metaprogramming I couldn't get it to stop conflicting with the other single-argument parse routine (the one that takes an input object). | |||||
| 2010-03-20 | Write a "between" convenience routine that sandwiches a parser between two ↵ | Daniel Burrows | 1 | -0/+53 | |
| side parsers. | |||||
| 2010-03-20 | Terminology change: skip -> skipMany. | Daniel Burrows | 1 | -24/+24 | |
| I felt it needed to be more clear that this *does* match multiple copies of the input. | |||||
| 2010-03-20 | Write a lexeme convenience wrapper that adds "drop trailing whitespace" to ↵ | Daniel Burrows | 1 | -0/+31 | |
| anything in the obvious way. | |||||
| 2010-03-20 | Fix the expected input position after the foreach failure. | Daniel Burrows | 1 | -1/+1 | |
| 2010-03-20 | Test the failure case for both foreach and many. | Daniel Burrows | 1 | -0/+24 | |
| As it turns out, foreach is buggy and doesn't fail. | |||||
| 2010-03-20 | Implement a sepBy parser for lists of separated values (e.g., ↵ | Daniel Burrows | 1 | -0/+70 | |
| comma-separated lists). | |||||
| 2010-03-20 | Change the internal protocol of the parser library to support tracking line ↵ | Daniel Burrows | 1 | -1/+1 | |
| and column numbers, and to support any Boost-style range. The big change here is to use an abstraction for the input rather than directly passing an iterator pair. The abstraction handles counting lines and columns as the iterator is advanced through the input. Still need to test that lines and columns are coming out right. | |||||
 |
index : aptitude | |
| [no description] | Dyson |
| summaryrefslogtreecommitdiff |