DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://issues.apache.org/bugzilla/show_bug.cgi?id=27901>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://issues.apache.org/bugzilla/show_bug.cgi?id=27901 [PATCH] TextCharIterator.remove() does not work properly ------- Additional Comments From [EMAIL PROTECTED] 2004-04-04 06:41 ------- The BlackList method is not my favorite, because I suspect only a few percent of extra spaces found are between words--the extra complexity appears unwarranted. My debugging has indicated that 95%+ of extra spaces all appear at the beginning of the FOText object, so it is best to just increment a start index for those. For the few percent of errant extra spaces between words, then an arraycopy can be called, removing the spaces one-by-one. (Keep in mind, one month ago we were doing an arraycopy for *every* space, even those at the beginning of the string, so a 95% reduction in arraycopies is already a huge improvement.) Please advise if you've come to different conclusions--but the complexity doesn't appear necessary to me. HOWEVER/Note: I've only been working with fo:blocks, not with fo:inlines, so my above reasoning may not be valid when those are taken into consideration. Anyway, please check the latest solution applied. Here I brought in the old solution to solve the extra spaces problem, while keeping the new solution for the more commonly occurring spaces at the front of fo:text. The (commented-out) System.out.println's in the remove() method are useful for visualizing the space removal, and indicate the following: RemoveA: Newest method, that just increments the start index until a word is found. No array copies needed here, very speedy. RemoveB and RemoveC: The previous method, ideally just activated for the relatively rare occurrences of extra spaces between words. Unfortunately, RemoveB and RemoveC are called much more often (50% or so) than they should be (As stated above, I suspect < 10%), so more optimization may be needed here. (Perhaps an error in my logic of when RemoveA vs. RemoveB should be called.) Comments most welcome. Glen
