Merge remote-tracking branch 'origin/master' into standard-notes-supp…

…ort-super-format

.github/workflows/build.yml

@@ -5,6 +5,8 @@ on:
     paths-ignore:
       - "**.md"
   pull_request:
+    paths-ignore:
+      - "**.md"
   workflow_dispatch:
 
 jobs:
@@ -22,6 +24,13 @@ jobs:
         # "macos-latest" is ARM based
         # "macos-13" is Intel based
         os: ["windows-latest", "ubuntu-latest", "macos-latest", "macos-13"]
+        # build "macos-13" executable only at release
+        # https://github.com/orgs/community/discussions/26253?sort=top#discussioncomment-3250989
+        is_release:
+          - ${{ startsWith(github.ref, 'refs/tags/v') }}
+        exclude:
+          - is_release: false
+            os: "macos-13"
     steps:
       - uses: actions/checkout@v4
         with:

.github/workflows/lint.yml

@@ -5,6 +5,8 @@ on:
     paths-ignore:
       - "**.md"
   pull_request:
+    paths-ignore:
+      - "**.md"
   workflow_dispatch:
 
 jobs:

.github/workflows/tests.yml

@@ -5,6 +5,8 @@ on:
     paths-ignore:
       - "**.md"
   pull_request:
+    paths-ignore:
+      - "**.md"
   workflow_dispatch:
 
 jobs:

src/formats/cherrytree.py

@@ -1,7 +1,9 @@
 """Convert cherrytree notes to the intermediate format."""
 
 import base64
+import logging
 from pathlib import Path
+import re
 import xml.etree.ElementTree as ET  # noqa: N817
 
 import common
@@ -10,6 +12,9 @@
 import markdown_lib
 
 
+LOGGER = logging.getLogger("jimmy")
+
+
 def convert_table(node):
     table_md = markdown_lib.common.MarkdownTable()
     for row_index, row in enumerate(node):
@@ -37,102 +42,114 @@ def fix_inline_formatting(md_content: str) -> str:
     for checked_checkbox in ("☑", "☒"):
         md_content = md_content.replace(checked_checkbox, "- [x]")
 
-    # unnumbered list
+    # numbered list
     for number in range(10):
         for bullet in (")", "-", ">"):
             md_content = md_content.replace(f"{number}{bullet}", f"{number}.")
 
     # unnumbered list
-    for bullet in ("•", "◇", "▪", "→", "⇒"):
+    for bullet in ("•", "◇", "▪", "▸", "→", "⇒"):
         md_content = md_content.replace(bullet, "-")
     return md_content
 
 
-def convert_rich_text(rich_text, logger):
+WHITESPACE_RE = re.compile(r"^(\s*)(.+?)(\s*)$")
+
+
+def separate_whitespace(string: str) -> tuple[str, str, str]:
+    # TODO: doctest
+    match_ = WHITESPACE_RE.match(string)
+    if match_ is None:
+        return ("", string, "")
+    # TODO: check types
+    return match_.groups(default="")  # type: ignore[return-value]
+
+
+def convert_rich_text(rich_text):
+    if rich_text.text is None:
+        return "", []
+    if not rich_text.text.strip():
+        return rich_text.text, []  # keep whitespaces but don't format them
     # TODO: is this fine with mixed text and child tags?
     note_links = []
-    md_content = ""
+
+    # formatting needs to be applied directly to the string without spaces
+    leading, md_content, trailing = separate_whitespace(rich_text.text)
     for attrib, attrib_value in rich_text.attrib.items():
         match attrib:
             case "background" | "foreground" | "justification":
-                if rich_text.text is not None:
-                    md_content += rich_text.text
-                logger.debug(
+                LOGGER.debug(
                     f"ignoring {attrib}={attrib_value} "
                     "as it's not supported in markdown"
                 )
             case "family":
                 match attrib_value:
                     case "monospace":
-                        md_content = f"`{rich_text.text}`"
+                        md_content = f"`{md_content}`"
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case "link":
                 url = attrib_value
                 if url.startswith("webs "):
                     # web links
                     url = url.lstrip("webs ")
                     if rich_text.text == url:
-                        md_content = f"<{url}>"
+                        md_content = f"<{md_content}>"
                     else:
-                        md_content = f"[{rich_text.text}]({url})"
+                        md_content = f"[{md_content}]({url})"
                 elif url.startswith("node "):
                     # internal node links
                     url = url.lstrip("node ")
-                    md_content = f"[{rich_text.text}]({url})"
-                    note_links.append(imf.NoteLink(md_content, url, rich_text.text))
+                    text = md_content
+                    md_content = f"[{text}]({url})"
+                    # Split the note ID from the optional title. It can look like:
+                    # "36 h2-3" or just "36".
+                    # TODO: Anchors are not supported.
+                    original_id = url.split(" ", maxsplit=1)[0]
+                    note_links.append(imf.NoteLink(md_content, original_id, text))
                 else:
                     # ?
-                    md_content = f"[{rich_text.text}]({url})"
+                    md_content = f"[{md_content}]({url})"
             case "scale":
                 match attrib_value:
                     case "sup":
-                        md_content = f"^{rich_text.text}^"
+                        md_content = f"^{md_content}^"
                     case "sub":
-                        md_content = f"~{rich_text.text}~"
-                    case "h1":
-                        md_content = f"# {rich_text.text}"
-                    case "h2":
-                        md_content = f"## {rich_text.text}"
-                    case "h3":
-                        md_content = f"### {rich_text.text}"
-                    case "h4":
-                        md_content = f"#### {rich_text.text}"
-                    case "h5":
-                        md_content = f"##### {rich_text.text}"
-                    case "h6":
-                        md_content = f"###### {rich_text.text}"
+                        md_content = f"~{md_content}~"
+                    case "h1" | "h2" | "h3" | "h4" | "h5" | "h6":
+                        leading = f"{"#" * int(attrib_value[-1])} " + leading
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case "strikethrough":
                 match attrib_value:
                     case "true":
-                        md_content = f"~~{rich_text.text}~~"
+                        md_content = f"~~{md_content}~~"
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case "style":
                 match attrib_value:
                     case "italic":
-                        md_content = f"*{rich_text.text}*"
+                        md_content = f"*{md_content}*"
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case "underline":
                 match attrib_value:
                     case "single":
-                        md_content = f"++{rich_text.text}++"
+                        md_content = f"++{md_content}++"
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case "weight":
                 match attrib_value:
                     case "heavy":
-                        md_content = f"**{rich_text.text}**"
+                        md_content = f"**{md_content}**"
                     case _:
-                        logger.warning(f"ignoring {attrib}={attrib_value}")
+                        LOGGER.warning(f"ignoring {attrib}={attrib_value}")
             case _:
-                logger.warning(f"ignoring {attrib}={attrib_value}")
+                LOGGER.warning(f"ignoring {attrib}={attrib_value}")
+    md_content = leading + md_content + trailing
     if not md_content:
         # TODO: make this more robust
-        md_content += "" if rich_text.text is None else rich_text.text
+        md_content = rich_text.text
     if not rich_text.attrib:
         # TODO: make this more robust
         # Make sure to don't break links.
@@ -171,17 +188,17 @@ def convert_to_markdown(self, node, root_notebook):
         # TODO
         # pylint: disable=too-many-locals
         title = node.attrib.get("name", "")
+        self.logger.debug(f'Converting note "{title}", parent "{root_notebook.title}"')
+        note_imf = imf.Note(title, source_application=self.format)
 
         new_root_notebook = None  # only needed if there are sub notes
-        resources = []
-        note_links = []
         note_body = ""
         for child in node:
             match child.tag:
                 case "rich_text":
-                    content_md, note_links_imf = convert_rich_text(child, self.logger)
+                    content_md, note_links_imf = convert_rich_text(child)
                     note_body += content_md
-                    note_links.extend(note_links_imf)
+                    note_imf.note_links.extend(note_links_imf)
                 case "node":
                     # there are sub notes -> create notebook with same name as note
                     if new_root_notebook is None:
@@ -192,9 +209,6 @@ def convert_to_markdown(self, node, root_notebook):
                         f"parent: {root_notebook.title}"
                     )
                     self.convert_to_markdown(child, new_root_notebook)
-                    if new_root_notebook.is_empty():
-                        # Delete the notebook if it's empty.
-                        del root_notebook.child_notebooks[-1]
                 case "codebox":
                     language = child.attrib.get("syntax_highlighting", "")
                     note_body += f"\n```{language}\n{child.text}\n```\n"
@@ -212,31 +226,22 @@ def convert_to_markdown(self, node, root_notebook):
                     # but we do it later with the common function.
                     resource_md, resource_imf = convert_png(child, self.root_path)
                     note_body += resource_md
-                    resources.append(resource_imf)
+                    note_imf.resources.append(resource_imf)
                 case "table":
                     note_body += "\n" + convert_table(child) + "\n"
                 case _:
                     self.logger.warning(f"ignoring tag {child.tag}")
 
-        self.logger.debug(f'Converting note "{title}", parent "{root_notebook.title}"')
+        note_imf.body = note_body
 
-        tags = []
         # cherrytree bookmark -> tag
-        unique_id = node.attrib["unique_id"]
-        if unique_id in self.bookmarked_nodes:
-            tags.append("cherrytree-bookmarked")
+        note_imf.original_id = node.attrib["unique_id"]
+        if note_imf.original_id in self.bookmarked_nodes:
+            note_imf.tags.append(imf.Tag("cherrytree-bookmarked"))
         if tags_str := node.attrib.get("tags", ""):
-            tags.extend(tags_str.strip().split(" "))
-
-        note_imf = imf.Note(
-            title,
-            note_body,
-            source_application=self.format,
-            tags=[imf.Tag(tag) for tag in tags],
-            resources=resources,
-            note_links=note_links,
-            original_id=unique_id,
-        )
+            note_imf.tags.extend(
+                imf.Tag(t) for t in tags_str.strip().split(" ") if t.strip()
+            )
 
         if (created_time := node.attrib.get("ts_creation")) is not None:
             note_imf.created = common.timestamp_to_datetime(int(created_time))
@@ -261,3 +266,6 @@ def convert(self, file_or_folder: Path):
                     self.convert_to_markdown(child, self.root_notebook)
                 case _:
                     self.logger.warning(f"ignoring tag {child.tag}")
+
+        # Don't export empty notebooks
+        self.remove_empty_notebooks()