Age | Commit message (Collapse) | Author |
|
|
|
By keeping tracks of modified types the joining of maps ha become
significantly faster.
|
|
|
|
|
|
|
|
|
|
Variable substitution was not generalizing any unknown variables.
|
|
|
|
|
|
|
|
Enable warnings by default, add two options for suppressing them,
fix warning formatting and update testsuites.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Whenever a variable that could hold one of two or more possible closures was
used in a particular application, the application was assumed to fail if ONE of
the closures would fail in this application. This has been corrected to infer
failing application if ALL possible closures would fail in the particular
application.
|
|
Dialyzer emits warnings like the following "The specification for _
states that the function might also return _ but the inferred return
is _", which are actually underspecifications and not wrong type
specifications. This patch makes sure that they are filed under the
appropriate category.
|
|
|
|
|
|
|
|
No warnings are emitted for funs that are non-returning when the
function that generates them has a contract that specifies that
it will return such a non-returning fun.
The actual bug, reported by Tuncer Ayaz and simplified by Maria
Christakis is included in Dialyzer's tests.
|
|
|
|
|
|
* sa/dialyzer-server-loop-fix:
Fix server loop detection
OTP-9489
|
|
Fixed dialyzer warning occuring on binary comprehension of form
"<< <<>> || {A, B} <- [{a, b}] >>" caused by default clause
inserted by compiler. Since this clause is different from the
case of list comprehension, dialyzer fails to suppress that warning.
|
|
|
|
Warnings about wrong contract ranges are post-processed to be displayed in the
file/line that contains the contract, although they can also be emitted while
processing other modules. For this relocation to succeed the module that
contains the contract should be currently under analysis. If this is not the
case the warning is displayed in the file/line of the call that proves the
discrepancy in the contract.
|
|
Dialyzer does not normally emit warnings for functions that
implement non-terminating server loops. This detection failed
when some of the elements in an SCC terminated normally (being
for example list comprehensions or other generic anonymous
functions that were included in the SCC). This patch fixes that.
|
|
* sa/dialyzer-invalid-spec-fix:
Fix crash related with the contract blame assignment patch
OTP-9219
|
|
The relevant commit is 8342fcf5395133a19d647f2ace606af9b7fc1732.
The old patch could emit warnings even for function that had a
problematic spec even without refinement. This warnings would
consume the relevant invalid spec warnings. This patch takes
care of this by ensuring that normal invalid spec messages are
emitted if the call that triggers the blame contract range
warning is in another module.
|
|
|
|
|
|
* sa/dialyzer-missing-spec:
Add spec to dialyzer_cl_parse:get_lib_dir/1
OTP-9129
|
|
|
|
Applies to the specification of the return value of a function.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* aronisstav/dialyzer/dialyzer_tests/OTP-9116:
Increase timetrap of options1 suite
Write output_plt even when plt_check is ok
Create plt with erts, kernel and stdlib only
Update test results as they currently appear in dev
Major restructure of dialyzer's testsuite
Add 'apps' option to the erlang interface
Update spec file to work with new common test structure
Test suites for Dialyzer
|
|
* sa/dialyzer-fix-guards:
Fix warnings about guards containing not
Fix errors in the handling of 'and'/'or' guards
Cosmetic changes
OTP-9099
|
|
|
|
|
|
This change adds a --fullpath option to Dialyzer, which makes the
warning messages contain the full path of the corresponding file.
Original patch submitted by Magnus Henoch (legoscia) on 15/9/2010
and cooked to death in the 'pu' branch all this time.
The patch was essentially correct and most of it has been used as
is, but there have been some changes to make the code slightly
prettier, avoid some code duplication, and add documentation to
dialyzer's doc files and to its help message.
|
|
The wording of warnings about unsatisfiable guards that used
'not' was incorrect (the 'not' was not mentioned and it appeared
as "Guard test is_atom(atom()) can never succeed").
|
|
Apart from the obvious bug in the negative evaluation of an 'and'
guard, Dialyzer handled dont_know cases rather single-mindedly
towards the positive branch. This patch allows for negative results
as well and does some clever guesses to narrow them down. It was
constructed similarly to the handling of the 'not' guard.
|