summaryrefslogtreecommitdiff
path: root/test
Commit message (Collapse)AuthorAgeFilesLines
...
* Start using Mock to test the command-line code.Ned Batchelder2009-09-091-4/+47
|
* Silly typoNed Batchelder2009-08-201-1/+1
|
* Fix a problem with DecoratorTools fiddling with the trace function and ↵Ned Batchelder2009-08-092-0/+65
| | | | screwing us up. Now the Python trace function is simpler, with no variability of registered trace function. Fixes bugs #12 and #13.
* Narrow the amount of code in a coverage start/stop window in a test so that ↵Ned Batchelder2009-07-121-8/+11
| | | | the differences in 2.x and 3.x scoping rules don't change the results of the test. Also, some commented-out logging of the trace machinery that helped me find the problem. Now all tests pass on Python 3.1!
* Py3k: manipulating PYTHONPATH seems different under Py 3.x, so do it ↵Ned Batchelder2009-07-101-3/+7
| | | | differently.
* Py3k: print statements in test files and gold annotation files have to be ↵Ned Batchelder2009-07-094-6/+8
| | | | switched to a 2-and-3 compatible form that 2to3.py won't change.
* Py3k: some output is bytes instead of string, so convert since it doesn't ↵Ned Batchelder2009-07-092-5/+5
| | | | matter.
* Py3k: fix more print statements.Ned Batchelder2009-07-091-1/+1
|
* Have the tests use the script explicitly until I can figure out how to ↵Ned Batchelder2009-07-091-1/+2
| | | | install a usable script for Py3k
* Py3k: bytes vs. strings, and one more print statement.Ned Batchelder2009-07-082-3/+3
|
* Py3k: exec is a function now, not a statement. Update the tests accordingly.Ned Batchelder2009-07-081-29/+62
|
* Py3k: unify subprocess and popen4.Ned Batchelder2009-07-084-22/+9
|
* Various Py3k fixes: remove gratuitous print, don't test the print statement, ↵Ned Batchelder2009-07-083-32/+46
| | | | deal with __cmp__ ugliness, etc.
* Move the backward-compatibility definitions to a common file. There seems to ↵Ned Batchelder2009-07-031-8/+1
| | | | be no pretty way to do this.
* Add a test that proves 2.3 has problems measuring exceptions. Now to figure ↵Ned Batchelder2009-07-031-0/+101
| | | | out what to do about it...
* Clean up the `set` definitions for 2.3 (these should go someplace common).Ned Batchelder2009-07-031-1/+2
|
* I forget why this test was neutered.Ned Batchelder2009-06-281-1/+1
|
* Epic bug: pyexpat fiddles incorrectly with the systrace function. This is a ↵Ned Batchelder2009-06-281-1/+46
| | | | hack to make it behave correctly with coverage.py. Fixes bug #10.
* When executing files, open them in Universal Newline mode, just as Python ↵Ned Batchelder2009-06-271-0/+11
| | | | itself does. Makes it possible to run Python from Windows on Mac, for example.
* PYTHONPATH might not exist (for example on OS/X)Ned Batchelder2009-06-271-1/+1
|
* Don't measure the coverage package itself. Inspired by, but does not fix, ↵Ned Batchelder2009-06-242-7/+17
| | | | issue #8.
* Comparing files needs to be agnostic to line endings, or testing on Linux ↵Ned Batchelder2009-06-2310-99/+108
| | | | won't work. Now gold files are newline-only, and the comparison function ignores line endings.
* Remove the fixed limit on recursion depth. Fixes issue #9.Ned Batchelder2009-06-231-0/+31
|
* Point to new docs; more doc tweaks.Ned Batchelder2009-06-131-1/+1
|
* Get the flat_rootname right for nested CodeUnits. Thanks, Christian Heimes.Ned Batchelder2009-06-127-0/+65
|
* Slight tweak so I could debug a test problem.Ned Batchelder2009-06-121-4/+6
|
* Expand tabs to spaces before creating HTML output. Fixes bug #6.Ned Batchelder2009-06-102-0/+26
|
* Silence pylint nagging.Ned Batchelder2009-05-131-6/+15
|
* Add some excluded lines to the annotation tests, and write some docstrings ↵Ned Batchelder2009-05-133-0/+6
| | | | for annotate.py
* Give the singleton module interface a way to keep the old behavior of ↵Ned Batchelder2009-05-121-2/+1
| | | | auto-loading and -saving data as needed.
* cover_stdlib wasn't quite right, it controls covering the entire Python ↵Ned Batchelder2009-05-121-2/+2
| | | | library, so cover_pylib it is.
* Reporting on nothing is OK.Ned Batchelder2009-05-111-0/+6
|
* Oops: two tests with the same name means one of them doesn't get run! ↵Ned Batchelder2009-05-111-2/+2
| | | | Thanks, pylint.
* Re-think the api to set the data file name and suffix.Ned Batchelder2009-05-112-4/+47
|
* More docstringsNed Batchelder2009-05-096-1/+20
|
* More docstrings all around.Ned Batchelder2009-05-092-6/+14
|
* Also make __builtins__ available in main programs.Ned Batchelder2009-05-092-1/+6
|
* Change run_python_file again so that it doesn't produce compiled turds in ↵Ned Batchelder2009-05-091-16/+10
| | | | the file system.
* makeFile is more useful if it doesn't append .pyNed Batchelder2009-05-084-18/+34
|
* Treat files not ending in .py properly.Ned Batchelder2009-05-071-1/+5
|
* OS truceNed Batchelder2009-05-071-8/+8
|
* Add a failing test for running non-.py files on the command line.Ned Batchelder2009-05-073-2/+34
|
* Change how data is harvested from the collector to simplify api use.Ned Batchelder2009-05-052-3/+9
|
* Lint lintNed Batchelder2009-05-051-1/+1
|
* Lint clean-ups.Ned Batchelder2009-05-051-8/+13
|
* Use the new clear_exclude() function to improve coverage_coverage.pyNed Batchelder2009-05-041-2/+8
|
* Added clear_exclude() and get_exclude_list()Ned Batchelder2009-05-041-0/+12
|
* Pass values into coverage's contructor rather than set properties.Ned Batchelder2009-05-041-2/+1
|
* Split api tests into their own file.Ned Batchelder2009-05-042-171/+177
|
* Keeping a stack of Collectors makes it possible for Coverage to measure ↵Ned Batchelder2009-05-031-0/+34
| | | | itself (mostly).
View Lorry Controller Status