feat: parse markdown links

luissimas · May 29, 2024 · 8202f76 · 8202f76
1 parent 664bbdf
commit 8202f76
Show file tree

Hide file tree

Showing 5 changed files with 37 additions and 9 deletions.
diff --git a/README.md b/README.md
@@ -10,10 +10,10 @@
 - [X] Expose prometheus metrics endpoint
 - [X] Read config
 - [X] Find all files recursivelly
-- [ ] Configurable ignore file patterns
-- [ ] Handle uncreated links
+- [X] Parse markdown links
+- [ ] Exclude links to non existing files
 - [ ] Collect backlinks
-- [ ] Parse markdown links
+- [ ] Configurable ignore file patterns
 - [ ] Get zettelkasten from git url
 - [ ] Support private repositories (Maybe with Github's PAT?)
 

diff --git a/go.mod b/go.mod
@@ -6,6 +6,8 @@ require (
 	github.com/gookit/validate v1.5.2
 	github.com/knadh/koanf v1.5.0
 	github.com/prometheus/client_golang v1.19.1
+	github.com/yuin/goldmark v1.7.1
+	go.abhg.dev/goldmark/wikilink v0.5.0
 )
 
 require (

diff --git a/go.sum b/go.sum
@@ -257,8 +257,13 @@ github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXl
 github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e h1:JVG44RsyaB9T2KIHavMF/ppJZNG9ZpyihvCd0w101no=
 github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e/go.mod h1:RbqR21r5mrJuqunuUZ/Dhy/avygyECGrLceyNeo4LiM=
 github.com/yuin/goldmark v1.1.27/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
+github.com/yuin/goldmark v1.1.32/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 github.com/yuin/goldmark v1.3.5/go.mod h1:mwnBkeHKe2W/ZEtQ+71ViKU8L12m81fl3OWwC1Zlc8k=
+github.com/yuin/goldmark v1.7.1 h1:3bajkSilaCbjdKVsKdZjZCLBNPL9pYzrCakKaf4U49U=
+github.com/yuin/goldmark v1.7.1/go.mod h1:uzxRWxtg69N339t3louHJ7+O03ezfj6PlliRlaOzY1E=
+go.abhg.dev/goldmark/wikilink v0.5.0 h1:/Gndy7+PoXzOc3reVWtXAh7Cni7wSqSxiuXDfmoYlm4=
+go.abhg.dev/goldmark/wikilink v0.5.0/go.mod h1:W1NzvDIpo6uoayolBTCsIL6y/QRAHmLTKfUUDfR75DA=
 go.etcd.io/etcd/api/v3 v3.5.4/go.mod h1:5GB2vv4A4AOn3yk7MftYGHkUfGtDHnEraIjym4dYz5A=
 go.etcd.io/etcd/client/pkg/v3 v3.5.4/go.mod h1:IJHfcCEKxYu1Os13ZdwCwIUTUVGYTSAM3YSwc9/Ac1g=
 go.etcd.io/etcd/client/v3 v3.5.4/go.mod h1:ZaRkVgBZC+L+dLCjTcF1hRXpgZXQPOvnA/Ak/gq3kiY=

diff --git a/internal/collector/note.go b/internal/collector/note.go
@@ -1,7 +1,18 @@
 package collector
 
 import (
-	"regexp"
+	"slices"
+
+	"github.com/yuin/goldmark"
+	"github.com/yuin/goldmark/ast"
+	"github.com/yuin/goldmark/text"
+	"go.abhg.dev/goldmark/wikilink"
+)
+
+var md = goldmark.New(
+	goldmark.WithExtensions(
+		&wikilink.Extender{},
+	),
 )
 
 type NoteMetrics struct {
@@ -13,7 +24,15 @@ func CollectNoteMetrics(content []byte) NoteMetrics {
 }
 
 func collectLinkCount(content []byte) int {
-	r, _ := regexp.Compile(`\[\[[^]]+\]\]`)
-	matches := r.FindAll(content, -1)
-	return len(matches)
+	linkKinds := []ast.NodeKind{ast.KindLink, wikilink.Kind}
+	reader := text.NewReader(content)
+	root := md.Parser().Parse(reader)
+	linkCount := 0
+	ast.Walk(root, func(n ast.Node, entering bool) (ast.WalkStatus, error) {
+		if entering && slices.Contains(linkKinds, n.Kind()) {
+			linkCount += 1
+		}
+		return ast.WalkContinue, nil
+	})
+	return linkCount
 }
diff --git a/internal/collector/note_test.go b/internal/collector/note_test.go
@@ -9,9 +9,11 @@ func TestCollectNoteMetrics(t *testing.T) {
 		expected NoteMetrics
 	}{
 		{"empty file", "", NoteMetrics{LinkCount: 0}},
-		{"file with a single link", "[[Link]]", NoteMetrics{LinkCount: 1}},
-		{"file with multiple links", "[[Link]]aksdjf[[anotherlink]]\n[[link]]", NoteMetrics{LinkCount: 3}},
+		{"file with a single wiki link", "[[Link]]", NoteMetrics{LinkCount: 1}},
+		{"file with multiple wiki links", "[[Link]]aksdjf[[anotherlink]]\n[[link]]", NoteMetrics{LinkCount: 3}},
 		{"wikilink dividers", "[[something|another]]\n\n[[link]]\n[[382dlk djfs link|yeah]]", NoteMetrics{LinkCount: 3}},
+		{"file with markdown link", "[Link](target.md)", NoteMetrics{LinkCount: 1}},
+		{"file with multiple links", "okok[Link](target.md)\n**ddk**[[linked]]`test`[[anothe|link]]\n\n[test](yet-another.md)", NoteMetrics{LinkCount: 4}},
 	}
 
 	for _, d := range data {