[GROOVY-7476] JsonSlurper sometimes does not parse escaped double quotes on parsing with JsonParsingUsingCharacterSource

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.3.11, 2.4.3
• Fix Version/s: 2.4.4
• Component/s: JSON
• Labels: contributers-welcome
• Assignee: Pascal Schumacher
• Reporter: Gabriel Van Eyck
• Created: 25/Jun/15 20:24
• Updated: 19/Oct/15 16:12
• Resolved: 28/Jun/15 09:03

Description

This came up when parsing a JSON string that had escaped JSON in it.
E.g.

{"string": "{\"key\": \"value\"}"}

If the escape character used for ReaderCharacterSource.findNextChar is the last character in the buffer, then "foundEscape" is ignored for the first character in the recursion call (ReaderCharacterSource:229) and the match character is not added to the result.

import groovy.json.internal.ReaderCharacterSource

def testString = '"double\\"quote"'
def expectedString = testString.substring(1, testString.length() - 1)
def escapedQuotePos = testString.indexOf('"', 1)

int quoteChar = (int)'"'.charAt(0)
int backslashChar = (int)'\\'.charAt(0)

ReaderCharacterSource rcs
char[] result

rcs = new ReaderCharacterSource(new StringReader(testString), escapedQuotePos - 1)
result = rcs.findNextChar(quoteChar, backslashChar)
assert(expectedString == new String(result))

rcs = new ReaderCharacterSource(new StringReader(testString), escapedQuotePos + 1)
result = rcs.findNextChar(quoteChar, backslashChar)
assert(expectedString == new String(result))

// This one fails due to the backslash being the last character in the buffer
rcs = new ReaderCharacterSource(new StringReader(testString), escapedQuotePos)
result = rcs.findNextChar(quoteChar, backslashChar)
assert(expectedString == new String(result))