Fixed regex so that start, end and offset are correctly parsed from maintenance command output

[dnivra]

The regex used for parsing the maintenance output fails if there are multiple colons(':') in the output line such as this line:

'[2] 0x0040026d->0x0040028e at 0x0000026d: .data: ALLOC LOAD READONLY DATA HAS_CONTENTS'

The output of current regex is

[('0x0040026d', '0x0040028e', '0x0000026d: .data', 'ALLOC', 'LOAD READONLY DATA HAS_CONTENTS')]

which results in offset(index 2) not being computed correctly and thus context registers displaying usage message. The correct output should be

[('0x0040026d', '0x0040028e', '0x0000026d' , '.data', 'ALLOC', 'LOAD READONLY DATA HAS_CONTENTS')]

The new regex in this pull request matches addresses and offset value more closely by looking for hexadecimal characters.