is and it modifiers have to be first class citizens

[luke-biel]

Existence of hamcrest_assertions is usefull, but requires addition of an extra crate. It would be beneficial if user could write test_case(args => is ...) without extra work.

For first release, for all assertions provided by hamcrest2 lib we should either cover it with our own syntax or mention replacement/why it's not implemented.

Later we'll be able to expand functionality of is|it on our own.

General matchers

• eq - already covered via arg => value is equal_to X
• not eq
• (not) lt is less_than X
• (not) leq is less_or_equal_than X
• (not) gt is greater_than X
• (not) geq is greater_or_equal_than X
• (not) type_of - do we need that?
• Add it (not) matches_regex ... #80

Numerical matchers

• (not) close_to is almost_equal_to X

Filesystem matchers

• (not) path_exists is existing_path
• (not) file_exists is file
• (not) dir_exists is dir

Option and Result

• has - matches Ok(5) can be used
• negated has - matches Ok(v) if v != 5
• (not) ok - matches Ok(_) & matches Err(_)
• (not) err - same
• (not) some - same
• (not) none - same

Collection matchers

• (not) contains it contains X
• (not) contains_exactly - matches vec![1, 2, 3] & matches [1, 2, 3]
• (not) contains_in_order it contains_in_order [X, Y, Z]
• Add is (not) len ... #81
• Add is (not) empty #82

Compound matchers

• (not) all
• (not) any

Misc

• is(bool) - supported via arg => value
• anything - useless in test_case context

[luke-biel]

All cases that have to be added (preferably in 2.1 release) were converted into separate issues. Rest seem to be unimportant, unless someone turns up with use case.
Closing this. In case something compes up, let's create separate issue and not block 2.0 release.