openclaw/notcrawl
openclaw/notcrawl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

fix(tui): collapse noisy notion link runs

Browse Source
This commit is contained in:
Vincent Koc 2026-05-03 23:29:15 -07:00
parent fcaa59ad8c
commit 449015e3db
No known key found for this signature in database
3 changed files with 28 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
cmd/notcrawl/main.go
View File

@ -860,7 +860,7 @@ func blockPreviewLines(blocks []store.Block, maxLines int) []string {
}
lines := make([]string, 0, maxLines)
for _, block := range blocks {
text := notiontext.CleanLegacyArtifacts(block.Text)
text := compactPreviewNoise(notiontext.CleanLegacyArtifacts(block.Text))
if text == "" {
continue
}
@ -891,6 +891,21 @@ func blockPreviewLines(blocks []store.Block, maxLines int) []string {
return lines
}
func compactPreviewNoise(s string) string {
s = strings.ReplaceAll(s, "linked pagess", "linked pages")
for strings.Contains(s, "linked page, linked page") ||
strings.Contains(s, "linked pages, linked page") ||
strings.Contains(s, "linked page, linked pages") ||
strings.Contains(s, "linked pages, linked pages") {
s = strings.ReplaceAll(s, "linked pages, linked page", "linked pages")
s = strings.ReplaceAll(s, "linked page, linked pages", "linked pages")
s = strings.ReplaceAll(s, "linked pages, linked pages", "linked pages")
s = strings.ReplaceAll(s, "linked page, linked page", "linked pages")
s = strings.ReplaceAll(s, "linked pagess", "linked pages")
}
return s
}
func collectionPreview(collection store.Collection, space, parent string) string {
var lines []string
if space != "" {

10
cmd/notcrawl/main_test.go
View File

@ -194,6 +194,16 @@ func TestBlockPreviewCleansLegacyNotionMarkers(t *testing.T) {
}
}
func TestBlockPreviewCompactsRepeatedLinkedPages(t *testing.T) {
got := blockPreview([]store.Block{{
Type: "paragraph",
Text: "linked page, linked page, linked page Add details",
}}, tuiPagePreviewMax)
if got != "linked pages Add details" {
t.Fatalf("got %q", got)
}
}
func TestPagePreviewIncludesComments(t *testing.T) {
got := pagePreview(
[]store.Block{{Type: "paragraph", Text: "status update"}},

3
internal/notiontext/text.go
View File

@ -18,7 +18,7 @@ var (
legacyBareMentionRE = regexp.MustCompile(`‣\s+[0-9a-fA-F]{8}-[0-9a-fA-F-]{8,}`)
spaceBeforePunctuationRE = regexp.MustCompile(`\s+([,.;:])`)
repeatedCommaRE = regexp.MustCompile(`(?:,\s*){2,}`)
repeatedLinkedPageRE = regexp.MustCompile(`(?:linked page,\s*){2,}linked page`)
repeatedLinkedPageRE = regexp.MustCompile(`linked page\b(?:,\s*linked page\b)+`)
)
func Normalize(s string) string {
@ -35,6 +35,7 @@ func CleanLegacyArtifacts(s string) string {
s = Normalize(s)
s = repeatedCommaRE.ReplaceAllString(s, ", ")
s = repeatedLinkedPageRE.ReplaceAllString(s, "linked pages")
s = strings.ReplaceAll(s, "linked pagess", "linked pages")
s = spaceBeforePunctuationRE.ReplaceAllString(s, "$1")
s = strings.ReplaceAll(s, " and, ", ", ")
return Normalize(s)