[TECHDEBT] Parsing of sword:// URLs is non-uniform

[jaakristioja]

Summary

Libsword and BibleTime sometimes generate links in the sword: scheme, which are in many places parsed by BibleTime using many different regular expressions, instead of there being a single function or class to parse these URLs.

Consequences

Since parsing of these URLs is done in multiple places using multiple different means, it is prone to deviate towards erronous program logic, e.g. not handling all cases of URLs.

Remediation approach

Both parsing and generation of such URLs should be refactored into a single place, perhaps using some approach based on or similar to QUrl.

Notes

See also #386.