Uh oh!
There was an error while loading. Please reload this page.
Splitting Bigtable Row into 3 classes based on the mutation type / RPC method used in commit() - #1557
Conversation
| for append mutations. | ||
| :rtype: :class:`.Row` | ||
| :rtype: :class:`.DirectRow` |
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
dhermes
commented
Mar 3, 2016
@tseaver just saw tests were failing on Travis. Looking into it. What do we need to do to keep this review moving? Really close to the finish line and want to get a release out |
dhermes
commented
Mar 3, 2016
OK the failure is simple to address (just a long failure) |
dhermes
commented
Mar 3, 2016
*lint failure. Phone autocorrect! |
For now, just renaming Row->DirectRow and creating 2 do-nothing subclasses. Class will be torn down in subsequent commits.
Simplifying most downstream issues by just removing any code in DirectRow that branched based on the presence of filter_.
Also simplifying DirectRow.commit() in the process.
Factoring behavior out into a non-public helper _set_cell() and using that helper in both DirectRow and ConditionalRow.
Factoring behavior out into a non-public helper _delete() and using that helper in both DirectRow and ConditionalRow. Also making sure ConditionalRow.delete() gets tested (codepath was only previously traversed via the parent class).
Factoring behavior out into a non-public helper _delete_cells() and using that helper in both DirectRow and ConditionalRow. Also making sure ConditionalRow.delete_cell(s) get tested (codepaths were only previously traversed via the parent class).
Long-term fix will be to create the `gcloud.bigtable.row_filters` module to reduce the amount of code in `gcloud.bigtable.row`. Also removing superflous assignment.
theacodes
commented
Mar 3, 2016
This mostly LGTM, pending the one open item. |
tseaver
commented
Mar 3, 2016
LGTM |
dhermes
commented
Mar 8, 2016
Merging this, will handle the open item in a small separate PR. |
Splitting Bigtable Row into 3 classes based on the mutation type / RPC method used in commit()
This is a follow-up to googleapis#1557 and prevents strange-ly true statements like `isinstance(cond_row, DirectRow)`.
This is a follow-up to googleapis#1557 and prevents strange-ly true statements like `isinstance(cond_row, DirectRow)`.
This is a follow-up to googleapis#1557 and prevents strange-ly true statements like `isinstance(cond_row, DirectRow)`.
Co-authored-by: release-please[bot] <55107282+release-please[bot]@users.noreply.github.com>
Credentials expects `aws_security_credentials_supplier`, not `aws_security_token_supplier`
This PR updates the librarian container sha for v1.0.0 support. --------- Co-authored-by: Anthonios Partheniou <partheniou@google.com>
chore: remove unwanted comments in AsyncMRD
* docs: add warning for bigframes 2.0 * style update * update content * update text * upadte based on comment * update wording * move to lower * Add llm change * Update README.rst --------- Co-authored-by: Tim Sweña (Swast) <swast@google.com>
Docs are still being updated, but I wanted to get this out there to speed up the review. Alternate implementation based on #1550.
I strongly urge reviewers to examine this commit-by-commit. The commit messages are fairly helpful (for the most part) and the diff of the PR as a whole is not so easy on the eyes.
Fixes#1548.
/cc @jonparrott@jgeewax