https://bz.apache.org/ooo/show_bug.cgi?id=107619
--- Comment #19 from mikekaganski <mikekagan...@hotmail.com> --- (In reply to Kay from comment #18) > (In reply to Francis C. Costero from comment #17) > > Find: ([^ ]*)[ ]*([^ ]*) > > Match zero or more occurences of a space following the beginning (grouping > 1), followed by zero or more occurrences of "space", followed by match > zero or more occurrences of a space following the beginning of the line > (grouping 2). Incorrect. [^ ] doesn't mean "a space following the beginning of the line", it's "Any character except space" syntax. Here "space" means character 0x20, not "any whitespace character" like [:space:], of course. If it were "a space following the beginning of the line", then in any of the following examples both groups would be empty (there's no lines starting with space). Thus, this means (taking greedy processing of "*" into account): put as much non-space characters as possible into the first grouping (possibly empty if the first char is space), THEN skip any spaces (if present), THEN take as much as possible of non-space characters LEFT into the second grouping (maybe zero if no other characters after those spaces). Anyway, first and last groupings represent DIFFERENT groups of characters, they may have identical value only if those different groups have identical characters in them. > > Replace: $1$2 > > I get the following after clicking Replace All > > 987987 > > This is correct. In this case $1 and $2 are identical because there is zero > or no occurrence of the "intervening" space. Incorrect. In this case $1 MUST be "987", and $2 MUST be "". > > ab > > This is correct. There was an intervening space. OK. > > aa > > This is correct and the same as the "987" string case. No, this is INCORRECT and the same as the "987" string case. -- You are receiving this mail because: You are on the CC list for the issue. You are the assignee for the issue.
