Karsten Bräckelmann wrote:
Wonder why that is -- due to the excessively long metas? The sub-rules'
REs are quite trivial.
Would constructing it using a binary (or n-ary, with small upper bound
of n) tree speed the compilation up?
Using a slightly different method - using a maximum number of children
parser. The times were taken after deleting the ~/.spamassassin folder
before each run.
Before
real 21m24.068s
user 18m58.465s
sys 0m45.532s
After using 4 children
real 13m7.309s
user 12m1.438s
sys 0m37.143s
After using 8 children
real 12m28.601s
user 11m38.879s
sys 0m37.250s
matt
