Data Flow AR500P
Here is a complete cross-reference matrix showing every program in the AR500 suite (including the ones you provided) and exactly how each program touches every physical file / table.
| Program | ARDETL | ARCUST | ARCONT | GSTABL | ARCWRK | ARCUSP | SA5SHX | ?9?ARC500 | ?9?AR500X | ?9?AR500D / ?9?AR500C | ?9?ARATBS | PRINT (spooled) | Side Effects / Notes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AR500P | – | Read (validation) | Read + Write (updates ACDATE) | Read | – | – | – | – | – | – | – | – | Updates last ageing date in every ARCONT |
| AR390 | Update (writes ADAGE + bucket moves) | Update (writes AGE(1–5) packed buckets + high-balance) | Read | Read | – | – | – | – | – | – | – | – | Only program that permanently changes live A/R data |
| AR5004 | – | Read (primary cycle) | – | – | Read | – | – | Create + Write (385-byte extract) | – | – | – | – | Builds filtered customer work file |
| AR5007 | Read (primary) | – | – | – | – | – | – | Update (writes NOD flag to pos 385) + Read | Create + Write (138-byte enriched detail) | – | – | – | Injects NOD + sort keys |
| #GSORT | – | – | – | – | – | – | – | Read | Read | Create + Write (sorted versions) | – | – | Two sorts: detail → AR500D, master → AR500C |
| AR750 | – | Read (AR500C) | Read | Read | – | Read | – | Read (AR500C) | Read (AR500D) | – | – | Write | By Customer # – current production layout |
| AR502 | – | Read (AR500C) | Read | Read | – | – | – | Read (AR500C) | Read (AR500D) | – | – | Write | By Name (alpha) |
| AR501 | – | Read (AR500C) | Read | Read | – | – | Read | Read (AR500C) | Read (AR500D) | – | Write | Write | By Salesman + optional Excel export |
Legend¶
- Read = opened for input only
- Update = opened for update (records modified)
- Create + Write = file is cleared/re-created and written to
- Write = spooled output or work file output
- – = not touched
Key Takeaways¶
- Only AR390 permanently modifies live production data (ARDETL + ARCUST)
- AR500P is the only other program that writes to a live file (ARCONT.ACDATE)
- All other programs are read-only on live files and only write to temporary work files or spooled output
- The ?9? files are all temporary and automatically deleted at job end
This matrix is extremely useful for impact analysis, modernisation planning, and regression testing.