Repository navigation
Commit db98990
Docs: re-anchor the citations INT23 moved
One re-anchor for the integration, INT22's exact-map method
(int22_exact_reanchor.py --exclusive-first): each citation is carried from
the tree its digits were written against onto this one, through an exact
difflib line map of the cited file -- main 8cf3be5 decides every anchor its
pages carry; b629aaeb and 5518a976 would decide the content lines taken
from them (one anchor, 5518a976's native_order.hpp:367 in
pine-to-native.md; b629aaeb's lines carry none).
moved 515, kept 1106, stuck 8 (6 are the floor citations 8d0f5c0 set
against this tree, which no origin carries; 2 set by hand, below)
1629 of 1629 anchors resolve (check_doc_anchors), doc lint clean.
Set by hand, each to the span its lane chose on its own tree:
- ADR-0001 risk row, `chart_day_key` -> pine_adapter.cpp:13452, the
definition (D2-D moved it below its out-of-line halves and set it so);
- native-feature-parity.md:871, `aggregates_input_bars` ->
pine_strategy_host.cpp:138, the spec overload the view overload forwards
to (D2-C set it so);
- native-feature-parity.md MG14, `chart_day_key` -> pine_adapter.cpp:
13394-13473: the map carried main's one-function span onto D2-D's
helper namespace alone (13394-13448); D2-D's own span runs from that
namespace through the end of chart_day_key.
Of the 517 moved spans, 473 cite identical text, 39 identical endpoints
around lines a lane rewrote inside, and 5 were read one by one: MG14 and
the aggregates_input_bars anchor above; OL4 cancel_all and MG4 end on a
read D2-C converted in place; HT5 calling_bar_complete (pine_aux_security
.cpp:255-336) ends on the warm-up scope D2-C rewrote -- D2-C's own page had
left that anchor at main's digits, stale in its tree. Digits only.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>1 parent dc76926 commit db98990
4 files changed
Lines changed: 278 additions & 278 deletions
File tree
- docs
- adr
- design
- pages
0 commit comments