[SPARK-50616][SQL] Add File Extension Option to CSV DataSource Writer #49233
+49
−3
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jabbaugh
force-pushed
the
jbaugh-add-csv-file-ext
branch
2 times, most recently
from
cede15e to
acd7d3e
Compare
github-actions
bot
added
STRUCTURED STREAMING
DEPLOY
BUILD
CORE
WINDOWS
INFRA
PYTHON
CONNECT
and removed
STRUCTURED STREAMING
DEPLOY
BUILD
CORE
WINDOWS
INFRA
PYTHON
CONNECT
labels
jabbaugh
force-pushed
the
jbaugh-add-csv-file-ext
branch
from
eee47c5 to
de3b891
Compare
jabbaugh
changed the title
HyukjinKwon
changed the title
MaxGekk
reviewed
Dec 19, 2024
| @@ -86,7 +86,7 @@ class CSVFileFormat extends TextBasedFileFormat with DataSourceRegister { | |||
| } | |||
|
|
|||
| override def getFileExtension(context: TaskAttemptContext): String = { | |||
| ".csv" + CodecStreams.getCompressionExtension(context) | |||
| "." + csvOptions.fileExtension + CodecStreams.getCompressionExtension(context) | |||
There was a problem hiding this comment.
In this way, the user can overwrite almost anything in the file system:
./../../../../../path/<file to overwrite>.xz
Should we do some sanity check of the option?
There was a problem hiding this comment.
That seems reasonable to me. I can add a check that limits the file extension to three characters that are limited to a-z. If other characters are found we can throw an exception stating the limitation.
There was a problem hiding this comment.
I have added a limit of 3 letters when we set the option.
See: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVOptions.scala
jabbaugh
force-pushed
the
jbaugh-add-csv-file-ext
branch
from
de3b891 to
2ff5f81
Compare
jabbaugh
force-pushed
the
jbaugh-add-csv-file-ext
branch
from
2ff5f81 to
3c63274
Compare
What changes were proposed in this pull request? The existing CSV DataSource allows one to set the delimiter/separator but does not allow the changing of the file extension. This means that a file can have values separated by tabs but me marked as a ".csv" file. This change allows one to change the file extension to match the delimiter/separator (e.g. ".tsv" for a tab separated value file). Why are the changes needed? This PR adds an additional option to set the fileExtension. The end result is that when a separator is set that is not a comma that the output file has a file extension that matches the separator (e.g. file.tsv, file.psv, etc...). Notes on Previous Pull Request apache#17973 A pull request adding this option was discussed 7 years ago. One reason it wasn't added was: "I would like to suggest to leave this out if there is no better reason for now. Downside of this is, it looks this allows arbitrary name and it does not gurantee the extention is, say, tsv when the delmiter is a tab. It is purely up to the user." I don't believe this is a good reason to not let the user set the extension. If we let them set the delimiter/separator to an arbitrary string/char then why not let the user also set the file extension to specify the separator that the file uses (e.g. tsv, psv, etc...). This addition keeps the "csv" file extension as the default and has the benefit of allowing other separators to match the file extension. Does this PR introduce any user-facing change? Yes. This PR adds one row to the options table for the CSV DataSource documentation to include the "fileExtension" option. How was this patch tested? One unit test was added to validate a file is written with the new extension. Was this patch authored or co-authored using generative AI tooling? No
jabbaugh
force-pushed
the
jbaugh-add-csv-file-ext
branch
from
3c63274 to
ec7d8cc
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
The existing CSV DataSource allows one to set the delimiter/separator but does not allow the changing of the file extension. This means that a file can have values separated by tabs but me marked as a ".csv" file. This change allows one to change the file extension to match the delimiter/separator (e.g. ".tsv" for a tab separated value file).
Why are the changes needed?
This PR adds an additional option to set the fileExtension. The end result is that when a separator is set that is not a comma that the output file has a file extension that matches the separator (e.g. file.tsv, file.psv, etc...).
Notes on Previous Pull Request #17973
A pull request adding this option was discussed 7 years ago. One reason it wasn't added was:
"I would like to suggest to leave this out if there is no better reason for now. Downside of this is, it looks this allows arbitrary name and it does not gurantee the extention is, say, tsv when the delmiter is a tab. It is purely up to the user."
I don't believe this is a good reason to not let the user set the extension. If we let them set the delimiter/separator to an arbitrary string/char then why not let the user also set the file extension to specify the separator that the file uses (e.g. tsv, psv, etc...). This addition keeps the "csv" file extension as the default and has the benefit of allowing other separators to match the file extension.
Does this PR introduce any user-facing change?
Yes. This PR adds one row to the options table for the CSV DataSource documentation to include the "fileExtension" option.
How was this patch tested?
One unit test was added to validate a file is written with the new extension.
Was this patch authored or co-authored using generative AI tooling?
No