Provide an interface to regular expression processing for LE conforming languages

Copied from old SHARE Requirements system SSLANG12001; Voting Priority=2.6

+5=1 +4=4 +3=10 +2=6 +1=3 abstain=2 -1=1 Priority=2.6

IBM should provide an interface to regular expression processing from LE conforming languages that does not require the relatively high level of skill necessary to call the C/C++ runtime routines directly, and also does not require writing C/C++ code on the part of the application developer.

Benefit: Application programs written in languages other than C/C++ could take advantage of regular expression pattern matching and substitution as part of their logic. Currently, in order to use regular expressions in a language other than C/C++, one must either code a C/C++ routine and call it from another LE-conforming language or one must call the C runtime routines directly. The former forces C/C++ coding (and the compiler), the latter is possible but requires a significantly higher level of skill on the part of the application developer than the solution requested by this requirement. For simple pattern matching, being able to obtain the offset and length of the matched string(s) would be ideal. It would be desirable if such an LE callable service would provide the same (or extended) functionality as the existing C "regcomp", "regexec", and "regfree" functions as described at http://publib.boulder.ibm.com/infocenter/zos/v1r13/index.jsp?topic=%2Fcom.ibm.zos.r13.bpxbd00%2Fregcopp.htm, http://publib.boulder.ibm.com/infocenter/zos/v1r13/index.jsp?topic=%2Fcom.ibm.zos.r13.bpxbd00%2Fregexep.htm, and http://publib.boulder.ibm.com/infocenter/zos/v1r13/index.jsp?topic=%2Fcom.ibm.zos.r13.bpxbd00%2Fregfrep.htm, respectively. Desirable, but not required, would be regular expression substitutions, something that provides the equivalent of the s/Kirk/Picard/g functionality of Perl would be ideal.

This was originally submitted on 2014-03-13.