RF: return values (#1350 continued) #1409
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…eption when asked to commit files unknown to git (Closes #1291)
…e when used with explicit paths to commit
ENH: provide 'files' parameter for commit ENH: Don't log git commit call failures at ERROR level
ENH: enable ok_clean_git's untracked_files feature for any AnnexRepo
…rectory - just as 'git status' does
BF: AnnexRepo.untracked_files does what is availabe, if annex-status-bug is preventing regular call BF: Workaround for AnnexRepo.add(git=True) in v6. Ugly but enables us to see behind test failures, caused by this one. TST: Several v6 fixes in test_annexrepo.py Additional notes and minor changes
…s' own recursive inspection of submodules
…cursion in AnnexRepo.status()
BF: Again, don't use annex-proxy for committing in direct mode, since it fails, when a passed path is a submodule (in direct mode) Additional notes and minor corrections
Disabling the lazy loading of Dataset.repo, since it conflicts with flyweight implementation. Both needs further discussion. See issue #1345
Necessary to support frequently employed calls without explicit kwarg specification.
This is a rough approximation of the cases that I can see. I am sure there are bits missing. However, the actual cases of install calls that we already had in the tests and that always wanted this feature seem to be happy with this draft.
Inline with gh-1476 the default renderer should be something that is uniform across commands and not something that is tailored.
It may be a 'clone' command that performs an actions, but a user is actually performing and expecting an 'install'
Now reports on directories and files in git too. This is a step towards only reporting results matching actual requests, and not end up with no apparent result despite a specific request.
Last mile, hopefully.
|
conflicts are because of me I guess so I will take a look now [done] |
* origin/master: (22 commits) RF: unify log line for publishing data.. commented out initial sketch to properly handle decision to try to copy data ENH: Q&D tool to transfer urls from datalad to web remote BF: try to ignore .this-is-a-test-key files annex might leave behind ruining whereis command BF(minor): early and correct report on mispecified output kward for whereis BF(TEMP): publish - sort datasets in reverse order by path (Closes #1479) BF(TEMP): publish - sort datasets in reverse order by path (Closes #1479) BF/RF: remove wrapping with '' of func node str args BF(TST): skip test if no creds for s3://openneuro BF: add archive content -- should check for existing based on full path BF: rmtree -- also use unlink when trying to remove single file RF: make unlinking in add_url_to_file optional (unlink_existing) BF: publish data only if annex BF: publish data first where needed/possible (Closes #1480) BF: do not rely on diff but comparing commits if no paths restriction. Allow to --force publish BF: (re)add (remove/overwrite) file from within archive under annex, if already exists ENH: skip_if_url_is_not_available and skip test needing NITRC (Closes #1472) BF: we should verify versioned url, not original Attempt to get PY2 test going again (fixes gh-1471) BF: s3 versioned urls assert that the latest version before filtering out delete markers ENH: func nodes should get beter description in logs ...
|
Merci! |
now page is back but not functioning properly -- requires auth etc
|
FTR: #1476 is a candidate for merge in here. I will not touch |
This was referenced May 5, 2017
|
FTR: Test failures are a timeout, and this (no idea): |
* origin/master: (27 commits) BF(TST): fixed up test for recent changes to the hook RF: turn js minimization back on BF(workaround): in recursive ls json mode do not return full nested list of node, just generate their jsons BF+TST: tried to understand + fixup js tests... more to grasp ENH: just converted tabs to spaces and some minor reformatting in test.html BF: commented out ds.submodules use with a note RF: use href= instead of .assign since some folks recommended to reflect navigation in history ENH: we need to store either we caching for parent dir fighting the windmills ENH: cache persistently within session, not only type, but also metadata path, and ds url ENH: allow to start previously stopped container for webserver BF: Fixed up for trailing / and path to the installed subdatasets to get breadcrubs coloring corrected RF: all -> all_, long_ -> long even at the level of API BF: update submodule upon publish/push within hook BF+ENH: pickup description for the datasets, enhance hook script BF: determining path to the web metadata file BF: Allow nodeJson to return complete metadata json RF: Remove scaffolding json param from main funcs ENH: Update crumb spans/colors based on their type BF: Return default type outside root ds boundary ... Conflicts: datalad/interface/ls.py -- maintained "master" way of getting relpaths, and then erecting a dataset
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Basic idea: commands
yieldreport dictionaries. And wrappers act upon this information in whatever desired way.Sits on top of #1348. Present tail of the procedure is in #1409
Change summary:
updatedid not fail on unavailable paths as argumentreturn_valuehelper module with:- function to give a default dict
- function to convert git-annex json record into our own structure
typeand bystatus(using new constraint above)filter_results)return_type,{'generator', 'list', 'item-or-list'})--on-failureswitch (fixes New general optionon_failureto indicate action when something goes wrong #819)assert_status,assert_message,assert_in_results,assert_not_in_results,assert_result_count,assert_fields_equalCommands converted:
clonecommand (fully converted); high-level condition checking ininstallitself still needs conversionget)add-siblingresults through)siblingcommand)Copied from #1468:
installis a new-style generator command nowgetfails on unavailable subdataset with incomprehensible message #1469addfor registering install submodules #1173install#1110installno longer uses any low-level helpers directly, instead it sits rather cleanly on top ofcloneandget, and performs argument inspection and switching logic only. It seems that this is now at a stage where we could think about adding more convenience functionality toinstallwithout much pain (e.g. install command should expose "*wanted*" options, get should use those as well #1347, datalad install --shared #1146)install, plus a duplicate (and buggy) implementation of aclonehelper removedinstallFate ofif_dirtyin the light of #1169 and other API RFs #1175getnow consolidated inget.pykwargsinstalladdshould use meaningful remote URL for registering submodules #1474getreport on things that annex is silent about (whole dirs, files in Git, ...)