summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* ripping out rspec_reset entirelyLamont Granquist2014-01-295-16/+0
|
* fix busted spec exposed by rpsec-2.14-izationLamont Granquist2014-01-291-1/+1
| | | | | | | the behavior it is testing has never worked. it can't really work since you're blowing up in String#[]= long before you could hit the attribute code. changed the test to test something slightly more useful.
* remove some more deprecation warningsLamont Granquist2014-01-296-9/+9
|
* fix broken specs revealed by rspec-2.14-izationLamont Granquist2014-01-291-11/+9
|
* need action for define_resource_requirementsLamont Granquist2014-01-293-0/+3
|
* fix test, define_resource_requirements needs an actionLamont Granquist2014-01-291-0/+1
|
* fix broken file provider specsLamont Granquist2014-01-291-0/+10
| | | | | fixing the should_not raise_error(Whatever) specs exposed that these were broken.
* fix should_not raise_error(SpecificException)Lamont Granquist2014-01-2945-216/+216
|
* this test should have been failingLamont Granquist2014-01-291-8/+8
| | | | | | "root*goo" is now legal for username/groupname. not sure why this test was passing on rspec 2.13.x. changed the name to something which is guaranteed to be invalid.
* the perils of regexp-based code editingLamont Granquist2014-01-291-1/+1
|
* s/mock\(/double(/gLamont Granquist2014-01-29114-363/+363
|
* s/stub!/stub/gLamont Granquist2014-01-29193-1625/+1625
| | | | fix deprecation warnings
* fix obvious derpLamont Granquist2014-01-291-1/+1
| | | | | | the fact that rspec 2.14 accepted this and changed the should_not_receive into a should_receive is a bit of a bug in rspec.
* fix rspec_reset problemLamont Granquist2014-01-295-4/+15
| | | | | | | follows the pattern in the rspec-mocks spec tests of defining a helper module: https://github.com/rspec/rspec-mocks/blob/cee433c89125a3984df33c87eb61985613adce9b/spec/spec_helper.rb
* fix "please stub a default message first"Lamont Granquist2014-01-291-17/+17
| | | | | | | | | | | | fixes this kind of thing: 1) Chef::Application class method: fatal! should log an error message to the logger Failure/Error: Chef::Log.stub!(:fatal).with("blah").and_return(true) #<IO:0x0000010106ab88> received :puts with unexpected arguments expected: ("FATAL: blah") got: (no args) Please stub a default value first if message might be received with other args as well. # /Users/lamont/oc/chef/spec/unit/application_spec.rb:234:in `block (3 levels) in <top (required)>'
* bump to rspec 2.14.xLamont Granquist2014-01-291-1/+1
|
* Merge pull request #1234 from opscode/adamed/oc-10899Adam Edwards2014-01-272-18/+70
|\ | | | | CHEF-4885: Chef::ReservedNames::Win32::Version has invalid methods
| * CHEF-4885: Refactor to add more tests, workaround for Win 8.1 API changeAdam Edwards2014-01-262-20/+66
| |
| * Added rspec test to check dynamic legit method nameameyavarade2014-01-261-1/+7
| |
| * [CHEF-4885] Don't define illegally named methods called :windows_8.1?Julian C. Dunn2014-01-261-1/+1
| |
* | Merge pull request #1227 from opscode/lcg/CHEF-4502lamont-granquist2014-01-2711-293/+487
|\ \ | |/ |/| CHEF-4502: Validate Content-Length Field in HTTP requests
| * move text message to exception classLamont Granquist2014-01-272-2/+6
| |
| * rename ValidateResponse to ValidateContentLengthLamont Granquist2014-01-272-4/+7
| |
| * add real error for bad download sizeClaire McQuin2014-01-233-5/+9
| |
| * validate streaming file sizeClaire McQuin2014-01-239-83/+84
| |
| * add specs for CHEF-4502 behaviorLamont Granquist2014-01-231-6/+32
| |
| * rspec modernizationLamont Granquist2014-01-231-55/+55
| |
| * more test refactoringLamont Granquist2014-01-231-97/+114
| |
| * WIP: partial rewrite of base specs on masterLamont Granquist2014-01-231-170/+226
| |
| * add counting content-length for streaming requestsLamont Granquist2014-01-231-2/+14
| |
| * WIP: hacking on ugly spec testsLamont Granquist2014-01-231-22/+34
| |
| * adding validation of content-length fieldLamont Granquist2014-01-232-0/+59
| |
* | Merge branch 'nowiki'danielsdeleo2014-01-231-36/+57
|\ \ | | | | | | | | | Updates the README to point to docs.opscode.com when possible.
| * | Update copyright footer in the readmedanielsdeleo2014-01-231-1/+1
| | |
| * | Remove most references to wiki from the READMEdanielsdeleo2014-01-231-35/+56
|/ / | | | | | | | | | | | | | | * Replace most references to the wiki with appropriate docs.opscode.com content. * Put a link to learnchef at the very top for new users * Half of the development guide was "how to install from source"; make that its own section instead of linking to wiki.
* | Merge branch 'CHEF-4983'danielsdeleo2014-01-236-26/+96
|\ \ | |/ |/|
| * Add public API for expanding run_list in Chef::Clientdanielsdeleo2014-01-235-25/+92
| | | | | | | | | | External projects like chefspec need to do this separately so they can inject custom behavior before/after.
| * Return node from Chef::Client#build_nodedanielsdeleo2014-01-232-1/+4
|/
* Merge branch 'policyfile-support'danielsdeleo2014-01-238-467/+1265
|\ | | | | | | https://tickets.opscode.com/browse/CHEF-4984
| * Fix Policyfile tests for ruby 1.8danielsdeleo2014-01-221-0/+6
| |
| * Fix incompatibilites between node expansion and policyfiledanielsdeleo2014-01-222-26/+66
| |
| * Update error messaging in Policyfile policy builderdanielsdeleo2014-01-222-5/+18
| |
| * Allow policyfile feature to be enabled in configdanielsdeleo2014-01-222-1/+15
| |
| * Add Policyfile-based policy builder for chef-clientdanielsdeleo2014-01-223-0/+662
| | | | | | | | | | | | | | | | | | | | | | | | NB: This is experimental and a bit of a hack. There is no server-side support, so interactions with the server are coded in a "compatibility mode" using data bag items and Chef 11 style cookbook version APIs. The Policyfile PolicyBuilder uses a single document from the server as the authoritative cookbook version set and expanded run list. Enabling it will disable support for environments and roles. In addition, chef-solo and override run lists are currently unsupported, though they could be supported in the future.
| * Extract "expand node" policy builder to new filedanielsdeleo2014-01-214-471/+528
| |
| * Add indirection to allow switching policy builder implementationsdanielsdeleo2014-01-213-157/+163
|/
* Merge branch 'CHEF-4725'danielsdeleo2014-01-207-60/+59
|\
| * Fix indentation of rescue clausedanielsdeleo2014-01-201-3/+3
| |
| * CHEF-4725: Disable buffering on all log streamsNikhil Benesch2014-01-204-4/+7
| | | | | | | | | | | | Centralize handling of log buffering to ensure it's always disabled. Previously, an unconfigured 'log_location' setting would erroneously fall back to a default STDOUT stream with output buffering enabled.
| * CHEF-4725: Validate 'log_location' setting on log initializationNikhil Benesch2014-01-205-56/+52
|/ | | | | | | | | | | | | | Previously, a Mixlib::Config setter method verified that 'log_location' was set to a writeable file or a valid IO stream. Due to the way Mixlib::Config handles `default` and `merge!`, this setter method would not fire if the 'log_location' was left unconfigured or if it was set on the command line, resulting in inconsistent error handling. This commit moves the validation logic out of the configuration layer and into the log initializion layer. This ensures that error handling is consistent, regardless of where the 'log_location' setting is configured. Validation logic is also simplified, relying on the MonoLogger class to open and configure any necessary IO streams.
View Lorry Controller Status