nsx: fix duplicated src attribute

src/formats/synology_note_station.py

@@ -5,7 +5,6 @@
 import difflib
 import json
 from pathlib import Path
-import re
 from urllib.parse import urlparse
 
 from bs4 import BeautifulSoup
@@ -27,10 +26,8 @@ class Attachment:
 
 
 def streamline_html(content_html: str) -> str:
-    # hack: In the original data, the attachment_id is stored in the
-    # "ref" attribute. Mitigate by storing it in the "src" attribute.
-    content_html = re.sub("<img.*?ref=", "<img src=", content_html)
-
+    # TODO
+    # pylint: disable=too-many-branches
     # another hack: make the first row of a table to the header
     soup = BeautifulSoup(content_html, "html.parser")
     for table in soup.find_all("table"):
@@ -74,6 +71,12 @@ def streamline_html(content_html: str) -> str:
             iframe.string = iframe.attrs["src"]
         iframe.attrs = {"href": iframe.attrs["src"]}
 
+    # hack: In the original data, the attachment_id is stored in the
+    # "ref" attribute. Mitigate by storing it in the "src" attribute.
+    for img in soup.find_all("img"):
+        if (new_src := img.attrs.get("ref")) is not None:
+            img.attrs["src"] = new_src
+
     return str(soup)
 
 

test/test_convert.py

@@ -142,6 +142,8 @@ def compare_dirs(dir1: Path, dir2: Path):
             [["synology_note_station/test_3/notestation-test-books.nsx"]],
             [["synology_note_station/test_4/test.nsx"]],
             [["synology_note_station/test_5/20241005_184010_8701_demouser.nsx"]],
+            # [["synology_note_station/test_6/test8.nsx"]],
+            # [["synology_note_station/test_7/test10.nsx"]],
             [["textbundle/test_1/example.textpack"]],
             [["textbundle/test_2/Bug report in tables_bear.textbundle"]],
             [["textbundle/test_3/Python CHP NOTES.textbundle"]],