[BEAM-10894] Basic CSV reading and writing. #12841
Conversation
|
R: @chamikaramj |
Codecov Report
@@ Coverage Diff @@
## master #12841 +/- ##
==========================================
+ Coverage 82.31% 82.33% +0.02%
==========================================
Files 451 452 +1
Lines 53874 53985 +111
==========================================
+ Hits 44344 44449 +105
- Misses 9530 9536 +6
Continue to review full report at Codecov.
|
| from apache_beam.dataframe import frame_base | ||
|
|
||
|
|
||
| def read_csv(path, *args, **kwargs): |
There was a problem hiding this comment.
Please add pydocs for public API here (or add a TODO/JIRA for this).
There was a problem hiding this comment.
The idea here is to mirror the Pandas APIs. I suppose I should reference them at least.
|
|
||
| def read_csv(path, *args, **kwargs): | ||
| return _ReadFromPandas(pd.read_csv, path, args, kwargs) | ||
|
|
There was a problem hiding this comment.
Is there a reason to provide these as methods instead of transforms (similar to other IO connectors) ?
There was a problem hiding this comment.
Mirroring the Pandas APIs. The _ReadFromPandas will be widely shared.
|
|
||
| pcoll = ( | ||
| paths_pcoll | ||
| | beam.FlatMap(expand_pattern) |
There was a problem hiding this comment.
Probably we can replace above file-handling related transforms with transforms available in fileio.
https://github.com/apache/beam/blob/master/sdks/python/apache_beam/io/fileio.py
For example,
fileio.MatchFiles(self.path) | ParDo(_ReadFromPandasFromReadableFileDoFn())
(ReadableFile.metadata.path gives the file path).
| self.binary)) | ||
|
|
||
|
|
||
| class _WriteToPandasFileBasedSink(io.FileBasedSink): |
There was a problem hiding this comment.
Instead of extending FileBasedSink here please implement a fileio.FileSink and use fileio.WriteToFiles.
https://github.com/apache/beam/blob/master/sdks/python/apache_beam/io/fileio.py#L76
|
Thanks for the feedback, I'll revise based on ReadableFile and WriteToFiles (if they're flexible enough). |
|
Updated to use the new fileio framework. (As an aside, I noticed that it's still marked experimental, should that be dropped.) PTAL |
There was a problem hiding this comment.
Thanks LGTM.
Regarding, experiment tags, I believe we can remove those now since that fileio transforms have been available in multiple Beam releases. cc: @pabloem
|
|
||
|
|
||
| def write_csv(df, path, *args, **kwargs): | ||
| from apache_beam.dataframe import convert |
There was a problem hiding this comment.
Import at top (here and below) ?
There was a problem hiding this comment.
I was running into circular import issues. Will add a comment.
| yield line.rstrip('\n') | ||
|
|
||
| @unittest.skipIf(sys.version_info[0] < 3, 'unicode issues') | ||
| def test_write_csv(self): |
There was a problem hiding this comment.
test_read_csv (seems like this is testing read_csv) ?
There was a problem hiding this comment.
This tests tests both read and write.
| def test_write_csv(self): | ||
| input = self.temp_dir({'1.csv': 'a,b\n1,2\n', '2.csv': 'a,b\n3,4\n'}) | ||
| output = self.temp_dir() | ||
| with beam.Pipeline() as p: |
There was a problem hiding this comment.
Add a test for write_cvs as well ?
There was a problem hiding this comment.
This tests tests both read and write. Updated name.
|
Thanks. I'll merge once I verify all tests are still passing after the merge. |
Thank you for your contribution! Follow this checklist to help us incorporate your contribution quickly and easily:
R: @username).[BEAM-XXX] Fixes bug in ApproximateQuantiles, where you replaceBEAM-XXXwith the appropriate JIRA issue, if applicable. This will automatically link the pull request to the issue.CHANGES.mdwith noteworthy changes.See the Contributor Guide for more tips on how to make review process smoother.
Post-Commit Tests Status (on master branch)
Pre-Commit Tests Status (on master branch)
See .test-infra/jenkins/README for trigger phrase, status and link of all Jenkins jobs.
GitHub Actions Tests Status (on master branch)
See CI.md for more information about GitHub Actions CI.