Github user gengliangwang commented on a diff in the pull request:

    https://github.com/apache/spark/pull/20535#discussion_r180714763
  
    --- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DataSourceOptions.java 
---
    @@ -17,16 +17,61 @@
     
     package org.apache.spark.sql.sources.v2;
     
    +import java.io.IOException;
     import java.util.HashMap;
     import java.util.Locale;
     import java.util.Map;
     import java.util.Optional;
    +import java.util.stream.Stream;
    +
    +import com.fasterxml.jackson.databind.ObjectMapper;
     
     import org.apache.spark.annotation.InterfaceStability;
     
     /**
      * An immutable string-to-string map in which keys are case-insensitive. 
This is used to represent
      * data source options.
    + *
    + * Each data source implementation can define its own options and teach 
its users how to set them.
    + * Spark doesn't have any restrictions about what options a data source 
should or should not have.
    + * Instead Spark defines some standard options that data sources can 
optionally adopt. It's possible
    + * that some options are very common and many data sources use them. 
However different data
    + * sources may define the common options(key and meaning) differently, 
which is quite confusing to
    + * end users.
    + *
    + * The standard options defined by Spark:
    + * <table summary="standard data source options">
    + *   <tr>
    + *     <th><b>Option key</b></th>
    + *     <th><b>Option value</b></th>
    + *   </tr>
    + *   <tr>
    + *     <td>path</td>
    + *     <td>A path string of the data files/directories, like
    + *     <code>path1</code>, <code>/absolute/file2</code>, 
<code>path3/*</code>. The path can
    + *     either be relative or absolute, points to either file or directory, 
and can contain
    + *     wildcards. This option is commonly used by file-based data 
sources.</td>
    + *   </tr>
    + *   <tr>
    + *     <td>paths</td>
    + *     <td>A JSON array style paths string of the data files/directories, 
like
    + *     <code>["path1", "/absolute/file2"]</code>. The format of each path 
is same as the
    + *     <code>path</code> option, plus it should follow JSON string literal 
format, e.g. quotes
    + *     should be escaped, <code>pa\"th</code> means pa"th.
    --- End diff --
    
    pa\"th?


---

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

Reply via email to