I get sad when I see both expert and new J programmers working hard to do simple string comparisons. J makes working with strings harder than it should to be.
I still have no idea why these comparison expressions are domain errors: 'abc' > 'def' |domain error | 'abc' >'def' 'def' < 'abc' |domain error | 'def' <'abc' What problem would occur if characters could be compared with each other? Strangely, compare equal works, but not how you think: 'abc' = 'def' 0 0 0 So, of course, strings are just arrays of characters. So, if strings are arrays of characters, and equality knows how to compare characters, why don't the other comparison operators know how to compare characters? J can manage most (any?) programming tasks but common things like comparing strings ought to be easy. Yes, I know that there are string utilities in stdlib.j, but the utilities are not well evangelized. String processing is as important a function to real-world computing as is numeric processing, and J is certainly capable of it. But, it has been treated as the "red-headed step-child" and left out of most of the documentation. If you browse the help documentation tree, the word "string" occurs once in the "Primer", three times in JforC, but does not occur in any of "Usr", "LJ", "Phr", "Dic", "Rel", or "Voc" sections. There are no mentions of what stdlib. Why aren't there standard phrases for string operations in Phrases? Most modern languages provide string functions, either natively, or as "core" functions. The "Help" index really out to have a "Lib" section, with a list of the various libraries, including "stdlib" and "regex", their functions, arguments, and descriptions. Alan ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
