Clarify pagination for catalog API endpoint

This change clarifies the way the catalog endpoint returns results
when pagination was not explicitly requested.

Signed-off-by: Patrick Devine <patrick.devine@docker.com>
This commit is contained in:
Patrick Devine 2015-12-02 17:26:52 -08:00
parent 329c353411
commit b084c77c70
2 changed files with 15 additions and 3 deletions

View file

@ -822,7 +822,13 @@ the presence of a repository only guarantees that it is there but not that it
is _not_ there. is _not_ there.
For registries with a large number of repositories, this response may be quite For registries with a large number of repositories, this response may be quite
large. If such a response is expected, one should use pagination. large. If such a response is expected, one should use pagination. A registry
may also limit the amount of responses returned even if pagination was not
explicitly requested. In this case the `Link` header will be returned along
with the results, and subsequent results can be obtained by following the link
as if pagination had been initially requested.
For details of the `Link` header, please see the _Pagination_ section.
#### Pagination #### Pagination

View file

@ -325,7 +325,7 @@ Above, we have bytestring `C` passed into a function, `SHA256`, that returns a
bytestring `B`, which is the hash of `C`. `D` gets the algorithm concatenated bytestring `B`, which is the hash of `C`. `D` gets the algorithm concatenated
with the hex encoding of `B`. We then define the identifier of `C` to `ID(C)` with the hex encoding of `B`. We then define the identifier of `C` to `ID(C)`
as equal to `D`. A digest can be verified by independently calculating `D` and as equal to `D`. A digest can be verified by independently calculating `D` and
comparing it with identifier `ID(C)` comparing it with identifier `ID(C)`.
#### Digest Header #### Digest Header
@ -822,7 +822,13 @@ the presence of a repository only guarantees that it is there but not that it
is _not_ there. is _not_ there.
For registries with a large number of repositories, this response may be quite For registries with a large number of repositories, this response may be quite
large. If such a response is expected, one should use pagination. large. If such a response is expected, one should use pagination. A registry
may also limit the amount of responses returned even if pagination was not
explicitly requested. In this case the `Link` header will be returned along
with the results, and subsequent results can be obtained by following the link
as if pagination had been initially requested.
For details of the `Link` header, please see the _Pagination_ section.
#### Pagination #### Pagination