perf(store): write resolver evidence straight into b-tree pages

Writing evidence row by row through SQL was the largest single cost of an index: 12.9 s of writes on django, 35.6 s on kubernetes. Rows now go straight to b-tree pages and digests run in parallel, the proof pool keeps a memo of the texts of a session, and hot keys no longer go through snprintf, which takes a process-wide locale lock on macOS. Writes take under 0.4 s on django and under 0.9 s on kubernetes. Databases shrink: kubernetes 4,364 to 4,162 MB, django 932 to 869 MB. Nothing readers use is removed; the same rows are stored.

Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>

scripts/memory-core-baseline.txt

@@ -12,7 +12,7 @@ internal/cbm/lsp/c_lsp.c	4
 internal/cbm/lsp/perl_lsp.c	21
 internal/cbm/macro_table.c	1
 internal/cbm/service_patterns.c	2
-internal/cbm/sqlite_writer.c	74
+internal/cbm/sqlite_writer.c	73
 src/cli/activation_transaction.c	78
 src/cli/agent_clients.c	73
 src/cli/agent_profiles.c	7
@@ -68,7 +68,7 @@ src/pipeline/pass_importance.c	4
 src/pipeline/pass_infrascan.c	2
 src/pipeline/pass_k8s.c	6
 src/pipeline/pass_lsp_cross.c	46
-src/pipeline/pass_parallel.c	29
+src/pipeline/pass_parallel.c	28
 src/pipeline/pass_pkgmap.c	63
 src/pipeline/pass_semantic.c	3
 src/pipeline/pass_tests.c	3