[ZEPPELIN-3125] Invalid UTF-8 middle byte

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Affects Version/s: 0.7.3
• Component/s: Core
• Flags: Important
• Assignee: Unassigned
• Reporter: Maziyar PANAHI
• Created: 02/Jan/18 18:23
• Updated: 03/Jan/18 13:11
• Resolved: 03/Jan/18 12:36

Description

If user writes any character that is not UTF-8 inside the notebook, it will immediately result in an error.

For instance:

wikipediaDF.filter($"article" === "série")

Error with 400 StatusCode: "Invalid UTF-8 middle byte 0x72\n at [Source: HttpInputOverHTTP@29b395f7; line: 2, column: 107]"

Log output from Zeppelin:

Job 20180102-180648_179114920 is finished, status: ERROR, exception: null, result: %text Error with 400 StatusCode: "Invalid UTF-8 middle byte 0x72\n at [Source: HttpInputOverHTTP@29b395f7; line: 2, column: 107]"

The same code works in Hue notebooks under the same environment. Looking at the sent request in Hue:
{"id":"d6a73e06-bc51-10b3-6b69-2381bb359fb2","type":"spark","status":"available","statementType":"text","statement":"wikipediaDF.filter($\"article\".rlike(\"\"\"série_.*\"\"\")).show(false)","statementPath":"","associatedDocumentUuid":null,"properties":{},"result":{"id":"68077196-0d82-a6d2-1726-11bff0a7aca5","type":"text","handle":{"has_result_set":true,"id":2}},"database":"default","wasBatchExecuted":false}