From 1bfc14202485d0496e0b4d88dc0e22c98ed3cb89 Mon Sep 17 00:00:00 2001
From: Dan Brown <jdanbrown@gmail.com>
Date: Tue, 4 Apr 2017 19:18:42 -0700
Subject: [PATCH] utils._http: Fix bytes -> str for py3

- http response content is a bytes, but downstream code expects it to be
  a str
- A previous commit (14ce30b2) had fixed the type mismatch in the
  successful code path (`json.loads`) but not in the error code path
  (`RequestException(..., content)`)

Test plan:
- Run simple query to test success path:
  - `bq.Query('select 3').to_dataframe()`
  - Before: returns dataframe with `3`
  - After: same behavior
- Run failing query that triggers the http error path:
  - e.g. `bq.Query('select * from table').to_dataframe()` where table is
    big and triggers error `Response too large to return`
  - Before:
```
Traceback (most recent call last):
  File "/Users/danb/hack/pydatalab/datalab/utils/_http.py", line 46, in __init__
    error = json.loads(content)['error']
  File "/Users/danb/miniconda3/envs/dwh/lib/python3.5/json/__init__.py", line 312, in loads
    s.__class__.__name__))
TypeError: the JSON object must be str, not 'bytes'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 322, in to_dataframe
    return self.results(use_cache=use_cache, dialect=dialect, billing_tier=billing_tier) \
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 228, in results
    self.execute(use_cache=use_cache, dialect=dialect, billing_tier=billing_tier)
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 528, in execute
    self._results = job.wait()
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query_job.py", line 84, in wait
    raise e
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query_job.py", line 82, in wait
    timeout=poll * 1000)
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_api.py", line 237, in jobs_query_results
    return datalab.utils.Http.request(url, args=args, credentials=self._credentials)
  File "/Users/danb/hack/pydatalab/datalab/utils/_http.py", line 154, in request
    raise RequestException(response.status, content)
  File "/Users/danb/hack/pydatalab/datalab/utils/_http.py", line 51, in __init__
    lines = content.split('\n') if isinstance(content, basestring) else []
TypeError: a bytes-like object is required, not 'str'
```
  - After:
```
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 322, in to_dataframe
    return self.results(use_cache=use_cache, dialect=dialect, billing_tier=billing_tier) \
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 228, in results
    self.execute(use_cache=use_cache, dialect=dialect, billing_tier=billing_tier)
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query.py", line 528, in execute
    self._results = job.wait()
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query_job.py", line 84, in wait
    raise e
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_query_job.py", line 82, in wait
    timeout=poll * 1000)
  File "/Users/danb/hack/pydatalab/datalab/bigquery/_api.py", line 237, in jobs_query_results
    return datalab.utils.Http.request(url, args=args, credentials=self._credentials)
  File "/Users/danb/hack/pydatalab/datalab/utils/_http.py", line 153, in request
    raise RequestException(response.status, content)
datalab.utils._http.RequestException: HTTP request failed: Response too large to return. Consider setting allowLargeResults to true in your job configuration. For more information, see https://cloud.google.com/bigquery/troubleshooting-errors
```
---
 datalab/utils/_http.py        | 5 ++---
 google/datalab/utils/_http.py | 5 ++---
 2 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/datalab/utils/_http.py b/datalab/utils/_http.py
index b545c9748..aa986cd97 100644
--- a/datalab/utils/_http.py
+++ b/datalab/utils/_http.py
@@ -143,13 +143,12 @@ def request(url, args=None, data=None, headers=None, method=None,
                                        method=method,
                                        body=data,
                                        headers=headers)
+      content = content.decode()
       if 200 <= response.status < 300:
         if raw_response:
           return content
-        if type(content) == str:
-          return json.loads(content)
         else:
-          return json.loads(str(content, encoding='UTF-8'))
+          return json.loads(content)
       else:
         raise RequestException(response.status, content)
     except ValueError:
diff --git a/google/datalab/utils/_http.py b/google/datalab/utils/_http.py
index 0254c96fc..115034e68 100644
--- a/google/datalab/utils/_http.py
+++ b/google/datalab/utils/_http.py
@@ -143,13 +143,12 @@ def request(url, args=None, data=None, headers=None, method=None,
                                        method=method,
                                        body=data,
                                        headers=headers)
+      content = content.decode()
       if 200 <= response.status < 300:
         if raw_response:
           return content
-        if type(content) == str:
-          return json.loads(content)
         else:
-          return json.loads(str(content, encoding='UTF-8'))
+          return json.loads(content)
       else:
         raise RequestException(response.status, content)
     except ValueError: