summaryrefslogtreecommitdiffstats
path: root/guide/push_prod.sh
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@robocracy.org>2018-12-14 22:01:14 +0800
committerBryan Newbold <bnewbold@robocracy.org>2018-12-14 22:01:16 +0800
commit595a15cfd3b3bd1e06d13f6274867616bea61ef0 (patch)
treedfb10d3da8cc16d3868723cb66a1e2dc8d49ea3f /guide/push_prod.sh
parente2deb74b7839aa05c1827a830ffc5e3c68ed58b4 (diff)
downloadfatcat-595a15cfd3b3bd1e06d13f6274867616bea61ef0.tar.gz
fatcat-595a15cfd3b3bd1e06d13f6274867616bea61ef0.zip
many redirect implementations
Probably should have split this commit up, it's huge: - accept the state of "redirect to a deletion", where redirect_id is Some but rev_id is None. call this a "redirect"; previously this was an invalid state. - GET for a deleted entity returns a 200 and a stub entity, not a 404 - to PUT a redirect, or to "revert" an entity to point at a specific pre-existing revision, PUT a stub entity. things are getting messy here... to detect this state, ensure the 'state' field is blank/none (this is for API usage ergonomics, where results from a GET are often re-used in a PUT or even POST) - rustfmt - maybe even more small tweaks along the way? mystery meat! Tests are in python, not rust (and a future commit)
Diffstat (limited to 'guide/push_prod.sh')
0 files changed, 0 insertions, 0 deletions