I modified a routine that processed xref names & paths w/in a set of folders. Thie original version would open each drawing & do its thing. THis would take about 20 minutes (aprroximately 90 files). I modified it to work w/o opening each drawing. This still takes about 20 minutes.
I would have though that not avtually opeing each drawing would dramitically decrease the overall time required to run this routine.
anyone have any insight to this?
Opening the file as a side database or with ObjectDBX is much much faster then opening it in the editor.
How are you modifying the drawings?
here is what my routine does to each DWG file in the set:
check that each XREF is on a particular layer which is locked & that the insertion point is 0,0
read the XREF status into an array
unlock that layer
reset the path of each XREF to a known existing location & reload the XREF
unload all XREFs
reload those that were loaded originally (status from the array)(most remain unloaded)
lock that layer
Any way to test your code against a file on the local machine rather than the network?
FWIW - I usually pull files from a different building and the network latencies can greatly override the benefit of doing your processing in a side database. Especially when you consider the fact that others are doing their own network stuff and filling up the pipe as it were.