ARROW-201: [C++] Initial ParquetWriter implementation #78
Conversation
|
Also with this PR comes read support for |
| NodePtr node_ = | ||
| GroupNode::Make("schema", Repetition::REQUIRED, std::vector<NodePtr>({pnode})); | ||
| return std::static_pointer_cast<GroupNode>(node_); | ||
| } |
There was a problem hiding this comment.
In testing more data types, we may wish to use templates and parameterized test cases to enable more code reuse
|
Finished reviewing -- looking good. Nice to get these first end-to-end tests working, so that we can then add the rest of the data types (and other missing features, e.g. compression) and work on read/write performance |
|
Incorporated all comments, travis builds seem to be breaking in an unrelated manner. What's missing is the initial check if we really have a flat column. Therefore I would need its definition. |
|
Rebased and now also Travis is happy. |
|
A column is flat if:
does that seem right? might be worth having a helper function for this |
|
We now also have a check for a flat column. I also wanted to add some DCHECKs (not needed for a merge but I like have code with assertions) but I get the following error if I added it somewhere in
|
|
looks good, +1. let's look into the namespace issue in |
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
Fixed some bugs in the filter code path.
merge Native sql engine branch
* Initial commit * Add int64 seed support * Add unit test cases
* Initial commit * Add int64 seed support * Add unit test cases
* Initial commit * Add int64 seed support * Add unit test cases
* Initial commit * Add int64 seed support * Add unit test cases
No description provided.