summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Prepare for release 1.20150829cliapp-1.20150829Lars Wirzenius2015-08-293-4/+4
|
* Update NEWSLars Wirzenius2015-08-291-0/+6
|
* Add subcommand usage getterRichard Ipsum2015-08-292-2/+12
|
* Update debian/changelogLars Wirzenius2015-08-291-0/+7
|
* Update NEWSLars Wirzenius2015-08-291-0/+10
|
* Update manual page wrt string list optionsLars Wirzenius2015-08-291-3/+12
|
* Fix whitespace, line lengthLars Wirzenius2015-08-292-1/+3
|
* Allow comma in StringListSetting valuesJan Gerber2015-08-292-10/+24
| | | | | - Escape values with comma with double quotes in config syntax. - Don't split values passed as arguments.
* Fix meliae useLars Wirzenius2015-07-222-1/+6
|
* Prepare release 1.20150701cliapp-1.20150701Lars Wirzenius2015-07-013-4/+4
|
* Change source package formatLars Wirzenius2015-06-302-1/+2
|
* Prepare release 1.20150630Lars Wirzenius2015-06-303-2/+8
|
* Move memory dumping method to util.pyLars Wirzenius2015-05-314-66/+99
|
* Mark outside test coverageLars Wirzenius2015-05-311-1/+1
|
* Remove logging.debug calls to reduce verbosityLars Wirzenius2015-05-311-2/+0
|
* Include stdout output in exception if command failsLars Wirzenius2015-05-091-1/+1
|
* Don't run pylint if it's too oldLars Wirzenius2015-05-011-1/+5
|
* Whitespace fixes for PEP8Lars Wirzenius2015-05-014-16/+16
|
* Build-depend on pep8, pylintLars Wirzenius2015-05-011-1/+1
|
* Add running of pylint in 'make check' and fix thingsLars Wirzenius2015-04-1824-209/+257
|
* Run pep8 in "make check" and fix so it passesLars Wirzenius2015-04-1818-142/+166
|
* Add ssh_runcmd(..., remote_cwd='...'))Lars Wirzenius2015-03-242-1/+18
|
* Quote user-provided optionsLars Wirzenius2015-03-241-1/+1
|
* Add ssh_options to ssh_runcmdLars Wirzenius2015-03-242-9/+23
|\
| * Update NEWSLars Wirzenius2015-03-241-0/+6
| |
| * Add ssh_options to ssh_runcomdLars Wirzenius2015-03-241-0/+3
| |
| * Document ssh_options in docstringLars Wirzenius2015-03-241-1/+5
| |
| * Refactor ssh_runcmd for code clarityLars Wirzenius2015-03-241-8/+9
|/
* Prepare for releasecliapp-1.20150305Lars Wirzenius2015-03-054-3/+9
|
* Collect all stderrs from pipelineLars Wirzenius2015-02-082-4/+20
|\
| * Update NEWSLars Wirzenius2015-02-081-0/+4
| |
| * runcmd: Collect stderr of whole pipelinebaserock/richardmaw/pipeline-stderr-pipe-v2Richard Maw2015-02-081-4/+16
|/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When pipe_stderr is None (inherit parent fd), subprocess.STDOUT (stderr to stdout) or a specific file, it would put everything to the right place, but when pipe_stderr is subprocess.PIPE the behaviour is less well defined. Before this change, it would happen to return the stderr of the last element of the pipeline, behaving similar to this shell snippet, except it does not need an intermediate file to store the contents of stderr in memory. stdout="$(foo 2>/dev/null | bar 2>/dev/null | baz 2>tmp)" stderr="$(cat tmp)" With this patch, it now behaves more like: stdout="$((foo | bar | baz) 2>tmp)" stderr="$(cat tmp)" It works by duplicating some of the subprocess.PIPE behaviour outside of subprocess.Popen, as there's no better way to make every subprocess have the same stderr, since there's no way to get the write end out of the Popen object after it has been constructed, since it is closed in the constructor, only leaving the read-end available as p.stderr. Only the last element of the pipeline is given the read end of the pipe, since it ought to have only one read pipe, and it makes the most sense for the last element to have the pipe, since you can then think of the pipeline as having a pipe that joins every element for stderr, that comes out at the end, with the stdout of the pipeline.
* Fix runcmd pipeline file descriptor in parentLars Wirzenius2015-01-302-0/+16
|\
| * Update NEWSLars Wirzenius2015-01-301-0/+7
| |
| * Fix non-termination bug in _build_pipelineRichard Maw2015-01-301-0/+9
|/ | | | | This fixes a bug that leads to non-termination in some pipelines that should terminate immediately such as `cat /dev/zero | false`
* Fix test for callbacksLars Wirzenius2014-08-031-2/+7
|
* Add callbacks to cliapp.runcmd for stdout/err dataLars Wirzenius2014-07-315-5/+107
|\
| * Allow runcmd callbacks to mangle the collected dataLars Wirzenius2014-07-312-8/+18
| |
| * Update NEWSLars Wirzenius2014-07-311-0/+7
| |
| * Fix copyright statementLars Wirzenius2014-07-311-1/+1
| |
| * Add tests for runcmd callbacksRichard Ipsum2014-07-311-0/+27
| |
| * Add LoggerApp exampleRichard Ipsum2014-07-312-0/+47
| |
| * Add callbacks to runcmdRichard Ipsum2014-07-311-2/+13
|/
* Prepare release 1.20140719cliapp-1.20140719Lars Wirzenius2014-07-194-3/+9
|
* Make Plugin.disable be a no-opLars Wirzenius2014-04-203-6/+6
|
* Remove some dead codeLars Wirzenius2014-04-151-16/+0
|
* Split logging setup into further overrideable methodsLars Wirzenius2014-04-152-7/+18
|
* Split of logging setup for file into separate methodLars Wirzenius2014-04-152-10/+19
|
* Fix missing get_help_text_formatterLars Wirzenius2014-04-092-1/+9
|
* Prepare release 1.20140315cliapp-1.20140315Lars Wirzenius2014-03-155-6/+6
|
View Lorry Controller Status