Commit Graph

82 Commits

Author SHA1 Message Date
Graham Paye 6818aaeb5a
add generated files to yardopts (#859) 2020-01-08 10:50:16 -08:00
Daniel Azuma 29f9544b14
fix: Prevent OpenCensus integration from getting out of sync on upload commands 2019-11-26 11:53:36 -08:00
Daniel Azuma ddd4ecea51
fix: Remove duplicate clauses in X-Goog-Api-Client header 2019-11-11 15:03:28 -08:00
Daniel Azuma 85763ece2f
fix: Prevent upload requests from form-encoding their queries 2019-11-05 08:28:54 -08:00
Mike Moore 33a4aea60d
fix: Use apply_request_options for batch auth header
Use HttpCommand#apply_request_options to add the Authorization header
for individual batch requests, as it supports both OAuth token strings and
googleauth/signet objects.

[pr #823, refs #822, closes #817]
2019-08-19 15:16:51 -06:00
Egor Vorobiev 837cf00706 fix: Add Auth header to individual batch requests
* Fixes an issue where the authorization HTTP header wasn't
  being applied to individual batch requests.

[pr #822, refs #817]
2019-08-19 12:08:44 -06:00
Daniel Azuma 8cf4330494
feat: send x-goog-api-client header in all requests 2019-08-16 10:13:11 -07:00
Daniel Azuma 4e38887c1f
Convert query values explicitly to strings (#810) 2019-07-29 11:29:32 -07:00
Daniel Azuma 75606c6ca6
Support JSON serialization options for model objects (#811) 2019-07-29 10:40:39 -07:00
Bruno Bornsztein c1a26e0e37 Replace mime-types dependency with mini_mime (#789) 2019-05-13 12:21:53 -07:00
Sampo Verkasalo 812b374103 Deserialize type: Date values (#776)
* Be explicit about datetimes
* Parse Date objects as specified in the types
2019-05-10 11:09:40 -07:00
Mike Moore 6e90789b27
Handle bad response errors without response object (#765)
We have received word that a bad proxy setting can cause this behavior.

[fixes #764]
2019-02-14 14:14:34 -07:00
Daniel Azuma 9b43faa06e
Redact response logs for certain cases (#757) 2019-01-30 14:14:21 -08:00
Daniel Azuma 55320753c7
Change logged representation of certain classes (#752) 2019-01-03 22:48:27 -08:00
Janko Marohnić e20e6bdffd Allow download destination to only respond to #write (#638)
According to the documentation the destination IO only needs to respond
to #write. However, the downloader breaks this by calling #flush.

We fix this by calling #flush only if the destination IO responds to
that method. We also add a test to enforce that the destination IO only
needs to respond to #write.
2018-12-04 13:09:11 -08:00
Daniel Azuma 50c4897e18
OpenCensus integration for outgoing requests (#724) 2018-10-10 13:23:11 -07:00
Yoann Lecuyer 224497c4a3 Add Signet::UnexpectedStatusError (#711) 2018-09-24 11:46:31 -07:00
Daniel Azuma 406fcb0032
Add logging to gemfile to fix specs (#707)
* Add logging to gemfile to fix specs
* Get rid of huge puts in the generator test that was overwhelming travis
* Fix HTTP tests by making the new Signet::RemoteServerError retriable
2018-08-22 13:42:05 -07:00
Alberto Almagro Sotelo 9f16b63dab
Fix little typo on docs
While reading the source file documentation I spotted this little typo.
2017-12-28 20:10:23 +01:00
Nilayan Bhattacharya 1a25d1ddeb Passing gzip compression as a client option 2017-11-29 12:45:11 -08:00
Sai Cheemalapati 2a4efbf3b9 Allow retry on Signet::AuthorizationError (#611) 2017-09-18 12:40:11 -07:00
Janko Marohnić 0eeb3ee9f9 Recover from non-Ranged responses without rewinding
Not all IO objects know how to #rewind themselves. For example, Ruby
pipes (returned by `IO.pipe`) do implement #rewind, but they will throw
an error if you try to call it.

  rd, wr = IO.pipe
  wr.rewind # Errno::ESPIPE: Illegal seek

But we don't need to rewind and overwrite the IO object if we didn't get
the Ranged response we expected, we could instead wait out the content
that has already been downloaded, and start appending again once we
reached where we left off. This is what this commit does.
2017-06-23 11:22:37 +02:00
Sai Cheemalapati e3792b19d2 Add Hash support to fetch_all (#586)
This commit aligns the behavior of `fetch_all` over map responses with
that of arrays (returning a single collection with all entries).
2017-06-01 23:52:09 -04:00
Steven Bazyl 6a4e9184e7 Allow using v2 error messages + fix to_json signature to allow arbitrary args 2017-05-19 17:16:22 -06:00
sonots ed27c6ec07 remove print debug 2017-04-04 17:36:46 +09:00
Steve Bazyl 48916bfce7 Merge branch 'master' of https://github.com/sqrrrl/google-api-ruby-client into sqrrrl-master 2017-04-03 12:31:20 -07:00
Steve Bazyl 4682267161 Move timeouts from request to clientoptions, update migration/changelog 2017-04-03 11:10:54 -07:00
Steve Bazyl 97ff7f4d9a #475, #488 -- allow raw JSON if explicitly requested, fix empty body handling & test cleanup 2017-03-31 15:26:51 -07:00
Steve Bazyl 489378db58 Ensure content length when empty body, hook up httpclient to logger when debugging 2017-03-31 14:59:07 -07:00
Steve Bazyl 5447ddb7c0 Don't normalize unicode in templates (fixed #557 + GCS issues) 2017-03-30 12:33:49 -07:00
Steve Bazyl c4d0ed5d3c Merge branch 'master' into v0.10 2017-03-29 16:25:34 -07:00
Steven Bazyl 5804593f25 Tighten validation around batch operations 2017-03-13 13:40:15 -07:00
Sai Cheemalapati 4f143ae551 Bump `representable` dependency to 3.0 (#532)
Also fixes a breaking change introduced between `representable` 2.3 and
3.0. Specifically, `:skip_undefined` is now be passed via the
`:user_options` option.
2017-02-13 12:24:14 -08:00
Thomas Coffee 16f0a768f7 Allow abnormal page token cases for paging methods
Addresses https://github.com/googleapis/toolkit/issues/692.

Manual changes are in:
lib/google/apis/core/base_service.rb
spec/google/apis/core/service_spec.rb
2017-02-03 12:09:18 -08:00
Austin Wang e9d7e47c75 Update check_status to better support different types of errors (#517) 2017-02-01 23:20:19 -08:00
Steve Bazyl a68b8cb758 #467 - Restore error handling for failed downloads 2016-09-19 14:16:36 -07:00
Steve Bazyl e13da8e05e Tighten use of form-encoding to only cases where a JSON body is not ever allowed 2016-09-08 13:40:30 -07:00
Steve Bazyl 70596d26f6 #459 - Fix iteration for genomics API which only has 1 item per page 2016-08-31 13:26:24 -07:00
Steve Bazyl 39d0a164b5 Update generator to correctly handle int64/uint64 types as fixnum/bignum instead of strings 2016-08-17 15:27:23 -07:00
Steve Bazyl 2c190e9745 Update representable, mimetype dependencies + others 2016-08-17 14:38:25 -07:00
Steve Bazyl 2578736b44 #445 - Treat 429 status codes as rate limit errors 2016-08-17 13:58:07 -07:00
Steve Bazyl 2046e00f14 Remove Hurley as a dependency 2016-08-17 13:51:09 -07:00
Steve Bazyl d8286f642c Prefer mimetype columnar store if available to reduce mem footprint 2016-06-23 12:44:06 -07:00
Steve Bazyl edb841310f Gracefully handle nil collections when paging 2016-06-23 12:43:37 -07:00
Steve Bazyl 1ee88320e9 #403 - allow nil dates 2016-04-25 10:49:00 -07:00
Steve Bazyl 0155b81771 Rewind output stream if requesting partial content but get the full media body instead 2016-04-22 14:30:30 -07:00
Steve Bazyl 6e3342b671 Cache pages in fetch_all to allow multiple iterations 2016-03-11 14:24:27 -08:00
Steve Bazyl f4453f6139 Add helper method for automatic paging 2016-03-08 13:26:28 -08:00
Steve Bazyl 47a02e897e #363 - Add convenience methods for json parsing/formatting to generated objects 2016-02-25 14:37:00 -08:00
Steve Bazyl 07ca6e8ca0 Handle SocketError. #359 2016-02-25 13:30:27 -08:00