fix(wiki): surface merge conflicts as textual-diff warnings (closes #104)
Textual-diff conflict detection only (semantic contradiction detection via brain-gardener Judge deferred, confirmed by Mathias, comment 12594). Includes canonicalBlock whitespace-run normalization fix for the disagreement comparison. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Co-authored-by: Mathias <mthbqv@gmail.com>
This commit was merged in pull request #129.
This commit is contained in:
@@ -37,10 +37,33 @@ type section struct {
|
||||
// Sections are merged by strategy: bullet sections union unique lines,
|
||||
// append sections concatenate, all others keep a's version. Sections in b
|
||||
// not present in a are appended.
|
||||
// Merge discards any disagreement silently; prefer MergeReport where the
|
||||
// caller can act on it.
|
||||
func Merge(a, b Page) Page {
|
||||
page, _ := MergeReport(a, b)
|
||||
return page
|
||||
}
|
||||
|
||||
// MergeReport is Merge, plus the headings of sections where b carried a
|
||||
// different version of a section a already had — and lost.
|
||||
//
|
||||
// That loss is the property which makes ingest safe to run unattended (#103):
|
||||
// a regenerating writer can add to a note but never rewrite it. But the
|
||||
// discarded text is also a signal, and until now it went in the bin: the source
|
||||
// now says something different about a section we already hold. Nobody could
|
||||
// tell "the source is unchanged" from "the source changed its mind and we
|
||||
// ignored it".
|
||||
//
|
||||
// Deliberately NOT a semantic contradiction check. This reports textual
|
||||
// disagreement, which is cheap, deterministic, and needs no model on the write
|
||||
// path. Deciding whether a disagreement is a genuine contradiction stays with
|
||||
// the Judge-gated enrich pass that already exists (brain-gardener#175), which
|
||||
// is approval-gated by design — detection here, adjudication there.
|
||||
func MergeReport(a, b Page) (Page, []string) {
|
||||
fmA, preambleA, secsA := parseSections(a.Content)
|
||||
fmB, _, secsB := parseSections(b.Content)
|
||||
|
||||
var conflicts []string
|
||||
idx := make(map[string]int, len(secsA))
|
||||
for i, s := range secsA {
|
||||
idx[s.heading] = i
|
||||
@@ -59,10 +82,15 @@ func Merge(a, b Page) Page {
|
||||
secsA[i].content = mergeBullets(sA.content, sB.content)
|
||||
case appendSections[sB.heading]:
|
||||
secsA[i].content = appendNew(sA.content, sB.content)
|
||||
default:
|
||||
// a wins — but say that b disagreed, rather than dropping it mute.
|
||||
if canonicalBlock(sA.content) != canonicalBlock(sB.content) {
|
||||
conflicts = append(conflicts, sB.heading)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return Page{Path: a.Path, Content: rebuildContent(refreshOwned(fmA, fmB), preambleA, secsA)}
|
||||
return Page{Path: a.Path, Content: rebuildContent(refreshOwned(fmA, fmB), preambleA, secsA)}, conflicts
|
||||
}
|
||||
|
||||
// generatorOwned names the frontmatter fields the ingest pipeline may refresh
|
||||
@@ -301,6 +329,21 @@ func canonicalLine(line string) string {
|
||||
return strings.Join(strings.Fields(line), " ")
|
||||
}
|
||||
|
||||
// canonicalBlock applies canonicalLine per line of a multi-line section body
|
||||
// — the same whitespace-insensitivity appendNew/mergeBullets already give a
|
||||
// single line, extended to a whole section for MergeReport's disagreement
|
||||
// check (found in review, PR #129: comparing raw strings.TrimSpace only
|
||||
// tolerated leading/trailing whitespace, not a re-extraction's mid-line
|
||||
// spacing, which every other comparison in this file already tolerates).
|
||||
func canonicalBlock(block string) string {
|
||||
lines := strings.Split(block, "\n")
|
||||
out := make([]string, len(lines))
|
||||
for i, l := range lines {
|
||||
out[i] = canonicalLine(l)
|
||||
}
|
||||
return strings.Join(out, "\n")
|
||||
}
|
||||
|
||||
// appendNew concatenates the lines of b that a does not already carry.
|
||||
//
|
||||
// Append sections used to concatenate wholesale, which is correct exactly once.
|
||||
|
||||
@@ -332,3 +332,85 @@ func TestRefreshOwned_IndentedListItemsWithColonsAreNotMisparsedAsKeys(t *testin
|
||||
got := refreshOwned(a, b)
|
||||
assert.Equal(t, "---\naliases:\n - 'Chapter 3: The Reckoning'\ntitle: 'X'\nsource_hash: bbbb\n---\n", got)
|
||||
}
|
||||
|
||||
func TestMerge_ReportsSectionsWhereTheFreshVersionWasDiscarded(t *testing.T) {
|
||||
// The asymmetry that makes ingest safe (on-disk wins) also throws away a
|
||||
// real signal: the source now says something different about a section we
|
||||
// already have. Today that disagreement is silently discarded, so nobody
|
||||
// can tell "the source is unchanged" from "the source changed its mind and
|
||||
// we ignored it".
|
||||
existing := Page{Path: "wiki/sources/x.md", Content: `---
|
||||
title: 'X'
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
Appetite is fixed, scope is variable.
|
||||
|
||||
## Method
|
||||
|
||||
Six-week cycles.
|
||||
`}
|
||||
fresh := Page{Path: "wiki/sources/x.md", Content: `---
|
||||
title: 'X'
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
Appetite is negotiable and scope is fixed.
|
||||
|
||||
## Method
|
||||
|
||||
Six-week cycles.
|
||||
`}
|
||||
|
||||
got, conflicts := MergeReport(existing, fresh)
|
||||
|
||||
assert.Contains(t, got.Content, "Appetite is fixed", "on-disk version still wins")
|
||||
assert.NotContains(t, got.Content, "negotiable")
|
||||
assert.Equal(t, []string{"Summary"}, conflicts,
|
||||
"only the section that actually disagreed is reported")
|
||||
}
|
||||
|
||||
func TestMerge_ReportsNothingWhenTheFreshVersionAgrees(t *testing.T) {
|
||||
page := Page{Path: "wiki/sources/x.md", Content: `---
|
||||
title: 'X'
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
Appetite is fixed, scope is variable.
|
||||
`}
|
||||
|
||||
_, conflicts := MergeReport(page, page)
|
||||
assert.Empty(t, conflicts, "identical content is not a disagreement")
|
||||
}
|
||||
|
||||
func TestMerge_ReportsNothingForInternalWhitespaceVariantsOnly(t *testing.T) {
|
||||
// Found in review of PR #129: the conflict comparison used raw
|
||||
// strings.TrimSpace, which only tolerates LEADING/TRAILING whitespace —
|
||||
// unlike appendNew/mergeBullets, which already canonicalize internal
|
||||
// whitespace runs too (#103: a re-extraction's spacing isn't guaranteed
|
||||
// byte-stable). A mid-line double-space from re-extraction would
|
||||
// false-positive as a "disagreement" here even though every other
|
||||
// whitespace-sensitive comparison in this file already tolerates it.
|
||||
existing := Page{Path: "wiki/sources/x.md", Content: `---
|
||||
title: 'X'
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
Appetite is fixed, scope is variable.
|
||||
`}
|
||||
fresh := Page{Path: "wiki/sources/x.md", Content: `---
|
||||
title: 'X'
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
Appetite is fixed, scope is variable.
|
||||
`}
|
||||
|
||||
_, conflicts := MergeReport(existing, fresh)
|
||||
assert.Empty(t, conflicts, "internal whitespace differences alone are not a disagreement")
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user