AnsweredAssumed Answered

Why is Drill sensitive to numeric file names?

Question asked by amsimms01 on Mar 3, 2016
Latest reply on Mar 22, 2016 by aengelbrecht
We have some generated json data files that have names that look like: 9_4-2003.json If I try to do a SELECT * against this 9_4-2003.json, Drill screams bloody murder.  If I copy it to a name that begins with an alpha, i.e. A9_4-2003.json, all is well:     sqlline> !connect jdbc:drill:zk=localhost:5181
    scan complete in 451ms
    0: jdbc:drill:zk=localhost:5181> SELECT * FROM dfs.lpi.`/team1/data/CHCSAB/JSONLD/9_4-2003.json` LIMIT 2;
    Error: VALIDATION ERROR: From line 1, column 15 to line 1, column 17: Table 'dfs.lpi./team1/data/CHCSAB/JSONLD/9_4-2003.json' not found
    
    
    [Error Id: 82c70b4b-f2ea-4a66-97e8-3f6fd2afae5f on hadoop-data06.lpi.localdomain:31010] (state=,code=0)
    0: jdbc:drill:zk=localhost:5181> SELECT * FROM dfs.lpi.`/team1/data/CHCSAB/JSONLD/A9_4-2003.json` LIMIT 2;
    +----------+--------+-------------+-------+
    | @context | @graph | generatedAt | query |
    +----------+--------+-------------+-------+
    | {"rdfs":"http://www.w3.org/2000/01/rdf-schema#","owl":"http://www.w3.org/2002/07/owl#","xsd":"http://www.w3.org/2001/XMLSchema#","skos":"http://www.w3.org/2004/02/skos/core#","prov":"http://www.w3.org/ns/prov#","fms":"http://datasets.caregraf.org/fms/","fmql":"http://datasets.caregraf.org/fmql/","chcss":"http://datasets.caregraf.org/chcss/","generatedAt":{"@id":"prov:generatedAtTime","@type":"xsd#date"},"id":"@id","_id":"@id","type":"@type","value":"@value","list":"@list","label":{"@id":"rdfs:label"},"sameAs":{"@id":"owl:sameAs","@type":"@id"} Where are the allowed names for files documented for Drill?
Is there a way to quote the name, to avoid renaming thousands of files? Thanks! --Andrew

Outcomes