A full-function sort and data manipulation utility for Linux and Windows, modeled on IBM DFSORT and Syncsort MFX, and optimized for very large volumes.
Everything a mainframe sort step does, on commodity hardware.
Order records on any combination of keys, or merge any number of pre-sorted files.
Match records across files with JOINKEYS and REFORMAT, including files in different formats.
Include, omit and summarize records with the conditional logic your jobs already contain.
Rebuild records, convert data types, translate EBCDIC and ASCII, and produce headers, trailers and sections.
The aim is simple: your existing statements run without change. Where a DFSORT option is not yet supported, we tell you up front.
| Control statements | SORT, MERGE, OPTION, RECORD, INCLUDE, OMIT, INREC, OUTREC, OUTFIL, SUM, ALTSEQ, JOINKEYS, JOIN, REFORMAT, MODS |
|---|---|
| DFSORT features | IFTHEN, BUILD, OVERLAY, PARSE, FINDREP, SEQNUM, date functions, OUTFIL splitting, HEADER, TRAILER and SECTIONS report writing |
| Utilities | DFSORT ICETOOL and SYMNAMES |
| Syncsort MFX | MFX control statement syntax, including SORTSUM and XSUM |
| User exits | DFSORT-style E15 and E35 exits for processing before and after the sort |
| Record formats | F, FB, V, VB and IBM-V, plus text and delimited files |
| Data types | EBCDIC, ASCII, packed decimal (COMP-3), zoned decimal, binary and the other IBM mainframe types, with packed decimal arithmetic and EBCDIC constants |

A sort engine sits under everything else, so it has to be predictable. AHLSORT is deterministic: the same input gives the same output every time.
One code base for every client, tested against a suite of around two thousand automated tests before each release.
Bring a representative job and its data. Our engineers will help you compare the results.