Whenever the document is modified the main loop restarts at the top. A future optimisation could possibly move a cursor pos along so that bits which have been scanned don't get scanned again, but I haven't got that far into optimising it yet and I probably wont for some time because it seems fast enough as it is.
Since the parser starts by looking for ( ) tags it makes sense to use them first, save scanning for them and finding nothing before scanning for < >
What bug btw?
In reply to Re^17: Thanks to Ikegami, Chromatic & Corion
by Logicus
in thread Thanks to Ikegami, Chromatic & Corion
by Logicus
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |