S
Simon Fairey
Hi,
I've seen a few references to swash but am still none the wiser. I
have a large SAX based parser that slowed down drastically when I
moved to 5.8 and after running DProf notice a huge number of calls to
SWASHGET/NEW. I isolated it down to a piece of code that had read a
string which happened to contain the £ sign (\243 I think) and was
performing a regexp on said string quite a few times. By running
utf8::decode() on the string in question before it's processed I
reduced my run time by 40% but am not really that much wiser.
Any info would be greatly appreciated. Is there a way I could have
opened/read from the file in the first instance to bypass this issue?
Many thanks for any info
Si
PS: Cross posted as most references I found to swash were in porters
but there is elements of the above probably better suited to
perl.misc, hope no-one minds
I've seen a few references to swash but am still none the wiser. I
have a large SAX based parser that slowed down drastically when I
moved to 5.8 and after running DProf notice a huge number of calls to
SWASHGET/NEW. I isolated it down to a piece of code that had read a
string which happened to contain the £ sign (\243 I think) and was
performing a regexp on said string quite a few times. By running
utf8::decode() on the string in question before it's processed I
reduced my run time by 40% but am not really that much wiser.
Any info would be greatly appreciated. Is there a way I could have
opened/read from the file in the first instance to bypass this issue?
Many thanks for any info
Si
PS: Cross posted as most references I found to swash were in porters
but there is elements of the above probably better suited to
perl.misc, hope no-one minds