[ https://issues.apache.org/jira/browse/SPARK-21996?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16167941#comment-16167941 ]
Xiayun Sun edited comment on SPARK-21996 at 9/15/17 2:30 PM: ------------------------------------------------------------- I can reproduce this issue for master branch, and found out it was from using {{.toString}} instead of {{.getPath}} to get the file path in string format. I pushed a PR to make it work for streaming. But there're other places in spark where we're using `.toString`. Not sure if it's a general concern. was (Author: xiayunsun): I can reproduce this issue for master branch, and found out it was from using `.toString` instead of `.getPath` to get the file path in string format. I pushed a PR to make it work for streaming. But there're other places in spark where we're using `.toString`. Not sure if it's a general concern. > Streaming ignores files with spaces in the file names > ----------------------------------------------------- > > Key: SPARK-21996 > URL: https://issues.apache.org/jira/browse/SPARK-21996 > Project: Spark > Issue Type: Bug > Components: SQL > Affects Versions: 2.2.0 > Environment: openjdk version "1.8.0_131" > OpenJDK Runtime Environment (build 1.8.0_131-8u131-b11-2ubuntu1.17.04.3-b11) > OpenJDK 64-Bit Server VM (build 25.131-b11, mixed mode) > Reporter: Ivan Sharamet > Attachments: spark-streaming.zip > > > I tried to stream text files from folder and noticed that files inside this > folder with spaces in their names ignored and there are some warnings in the > log: > {code} > 17/09/13 16:15:14 WARN InMemoryFileIndex: The directory > file:/in/two%20two.txt was not found. Was it deleted very recently? > {code} > I found that this happens due to duplicate file path URI encoding (I suppose) > and the actual URI inside path objects looks like this > {{file:/in/two%2520two.txt}}. > To reproduce this issue just place some text files with spaces in their names > and execute some simple streaming code: > {code:java} > /in > /one.txt > /two two.txt > /three.txt > {code} > {code} > sparkSession.readStream.textFile("/in") > .writeStream > .option("checkpointLocation", "/checkpoint") > .format("text") > .start("/out") > .awaitTermination() > {code} > The result will contain only content of files {{one.txt}} and {{three.txt}}. -- This message was sent by Atlassian JIRA (v6.4.14#64029) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org