rpg_trim.zig, rpg_eval.zig: the trim of an element, and a concatenation read as a + b #745
Loading…
Reference in a new issue
No description provided.
Delete branch "lc2e"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
rpg_trim.zig, rpg_eval.zig: the trim of an element, and a concatenation read as a + b
Two rules, both found with the
lc2-element-callsbattery (SC-ILE/EXPERIMENTS/lc2-*).s = %trim(names(i)),s = %trim(rec(i).name)andn = %len(%trim(names(i)))stopped atcontrol_flow / operation 15, the head of the trim loops. The trim unit named
three sources (a field, the pointer of a parameter, a temporary). The address
operation of the element unit (
rpg_array_element.zig), with27 K; 62afterit for a part at byte K, is now a fourth (
Source.element). The element stayson the stack, and the trim operation reads it as
S260(S257)or%SUBST(S264(S257): 5: 9).argumentUnit(a trim as the argument of a call)does not take an element yet.
the next piece from the result of the copy (
97 16 3 16 0; 27 K; 62). Thecopy answered
0, so62had nothing to add to and the statement stopped atvalue_flow / stack_value / operation 62 (h-bifs-char 40,
S = C + C, was one).A copy into a temporary now saves its piece, answers the address that it
wrote, and the load of the whole buffer joins the pieces in order as
a + b. A copy is a piece only when the next operations address the bytesafter it (
27 K; 62) or a saved piece ends where it starts, so the copy of aconstant argument into a temporary keeps its own rule. A copy also reads an
element as its source.
Results:
S265.0 = S260(S257) + 'x'(lc2-concat-arr),S265.0 = %SUBST(S264(S257): 5: 9) + 'x'(lc2-concat-ds) andS264.0 = S255 + S255(h-bifs-char 40); the store of a concatenation into an element
(lc2-store-concat-*) recovers with the typed read of the buffer on the right.
Operation 75 after a concatenation
No statement of the battery, of h-bifs-char or of the dict rows stops at
type_inference / operation 75 after a concatenation. The public statements
that stop at operation 75 are the setup of a procedure call that passes an
argument by const or by reference:
29 44 96 3; 5; 75 2; 42 252 0 Nafter thecopy of the argument (r9r-decstr 4, dict calls2 to calls5, CALLS 7, program-*).
The CCSID calls of branch
l1f-ccsid-callsare operation 99 and do not meet it.The battery
lc2-concat-forms(25 cases) asks for the concatenation with avarying piece or target,
%CHARand%EDITCpieces, and a concatenation as acall argument.
Still at operation 15 in the element battery
rec(i).name = %trim(s)(the element is the target),%SCAN(%TRIM(H) : C)(dict sv4 to sv14),
%CHARof an integer element (the parkedlc2bbranch),%CHECK,%REPLACE,%SCANRPL,%UPPER, and a call that takes an element byconst or by reference.
Gates