From 972cf0a887fcafa92bd317dd990ea57f2a4d054d Mon Sep 17 00:00:00 2001
From: Jacky Zhao <j.zhao2k19@gmail.com>
Date: Sun, 17 Dec 2023 12:27:19 -0800
Subject: [PATCH] feat: support emoji tags (closes #634)

---
 quartz/plugins/transformers/ofm.ts |  7 ++++---
 quartz/util/path.ts                | 22 +++++++++++-----------
 2 files changed, 15 insertions(+), 14 deletions(-)

diff --git a/quartz/plugins/transformers/ofm.ts b/quartz/plugins/transformers/ofm.ts
index 4c6a6db..b6d158c 100644
--- a/quartz/plugins/transformers/ofm.ts
+++ b/quartz/plugins/transformers/ofm.ts
@@ -121,9 +121,9 @@ const calloutRegex = new RegExp(/^\[\!(\w+)\]([+-]?)/)
 const calloutLineRegex = new RegExp(/^> *\[\!\w+\][+-]?.*$/, "gm")
 // (?:^| )              -> non-capturing group, tag should start be separated by a space or be the start of the line
 // #(...)               -> capturing group, tag itself must start with #
-// (?:[-_\p{L}])+       -> non-capturing group, non-empty string of (Unicode-aware) alpha-numeric characters, hyphens and/or underscores
-// (?:\/[-_\p{L}]+)*)   -> non-capturing group, matches an arbitrary number of tag strings separated by "/"
-const tagRegex = new RegExp(/(?:^| )#((?:[-_\p{L}\d])+(?:\/[-_\p{L}\d]+)*)/, "gu")
+// (?:[-_\p{L}\d\p{Z}])+       -> non-capturing group, non-empty string of (Unicode-aware) alpha-numeric characters and symbols, hyphens and/or underscores
+// (?:\/[-_\p{L}\d\p{Z}]+)*)   -> non-capturing group, matches an arbitrary number of tag strings separated by "/"
+const tagRegex = new RegExp(/(?:^| )#((?:[-_\p{L}\d\p{Z}])+(?:\/[-_\p{L}\d\p{Z}]+)*)/, "gu")
 const blockReferenceRegex = new RegExp(/\^([A-Za-z0-9]+)$/, "g")
 
 export const ObsidianFlavoredMarkdown: QuartzTransformerPlugin<Partial<Options> | undefined> = (
@@ -405,6 +405,7 @@ export const ObsidianFlavoredMarkdown: QuartzTransformerPlugin<Partial<Options>
               if (/^\d+$/.test(tag)) {
                 return false
               }
+
               tag = slugTag(tag)
               if (file.data.frontmatter && !file.data.frontmatter.tags.includes(tag)) {
                 file.data.frontmatter.tags.push(tag)
diff --git a/quartz/util/path.ts b/quartz/util/path.ts
index 5cf54b8..92cfabe 100644
--- a/quartz/util/path.ts
+++ b/quartz/util/path.ts
@@ -1,4 +1,4 @@
-import { slug } from "github-slugger"
+import { slug as slugAnchor } from "github-slugger"
 import type { Element as HastElement } from "hast"
 // this file must be isomorphic so it can't use node libs (e.g. path)
 
@@ -43,6 +43,14 @@ export function getFullSlug(window: Window): FullSlug {
   return res
 }
 
+function sluggify(s: string): string {
+  return s
+    .split("/")
+    .map((segment) => segment.replace(/\s/g, "-").replace(/%/g, "-percent").replace(/\?/g, "-q")) // slugify all segments
+    .join("/") // always use / as sep
+    .replace(/\/$/, "")
+}
+
 export function slugifyFilePath(fp: FilePath, excludeExt?: boolean): FullSlug {
   fp = _stripSlashes(fp) as FilePath
   let ext = _getFileExtension(fp)
@@ -51,11 +59,7 @@ export function slugifyFilePath(fp: FilePath, excludeExt?: boolean): FullSlug {
     ext = ""
   }
 
-  let slug = withoutFileExt
-    .split("/")
-    .map((segment) => segment.replace(/\s/g, "-").replace(/%/g, "-percent").replace(/\?/g, "-q")) // slugify all segments
-    .join("/") // always use / as sep
-    .replace(/\/$/, "") // remove trailing slash
+  let slug = sluggify(withoutFileExt)
 
   // treat _index as index
   if (_endsWith(slug, "_index")) {
@@ -156,14 +160,10 @@ export function splitAnchor(link: string): [string, string] {
   return [fp, anchor]
 }
 
-export function slugAnchor(anchor: string) {
-  return slug(anchor)
-}
-
 export function slugTag(tag: string) {
   return tag
     .split("/")
-    .map((tagSegment) => slug(tagSegment))
+    .map((tagSegment) => sluggify(tagSegment))
     .join("/")
 }