32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146 | class ObsidianReader(BaseReader):
"""
input_dir (str): Path to the Obsidian vault.
extract_tasks (bool): If True, extract tasks from the text and store them in metadata.
Default is False.
remove_tasks_from_text (bool): If True and extract_tasks is True, remove the task
lines from the main document text.
Default is False.
"""
def __init__(
self,
input_dir: str,
extract_tasks: bool = False,
remove_tasks_from_text: bool = False,
):
self.input_dir = Path(input_dir)
self.extract_tasks = extract_tasks
self.remove_tasks_from_text = remove_tasks_from_text
def load_data(self, *args: Any, **load_kwargs: Any) -> List[Document]:
"""
Walks through the vault, loads each markdown file, and adds extra metadata
(file name, folder path, wikilinks, backlinks, and optionally tasks).
"""
docs: List[Document] = []
# This map will hold: {target_note: [linking_note1, linking_note2, ...]}
backlinks_map = {}
for dirpath, dirnames, filenames in os.walk(self.input_dir):
# Skip hidden directories.
dirnames[:] = [d for d in dirnames if not d.startswith(".")]
for filename in filenames:
if filename.endswith(".md"):
filepath = os.path.join(dirpath, filename)
md_docs = MarkdownReader().load_data(Path(filepath))
for i, doc in enumerate(md_docs):
file_path_obj = Path(filepath)
note_name = file_path_obj.stem
doc.metadata["file_name"] = file_path_obj.name
doc.metadata["folder_path"] = str(file_path_obj.parent)
doc.metadata["folder_name"] = str(
file_path_obj.parent.relative_to(self.input_dir)
)
doc.metadata["note_name"] = note_name
wikilinks = self._extract_wikilinks(doc.text)
doc.metadata["wikilinks"] = wikilinks
# For each wikilink found in this document, record a backlink from this note.
for link in wikilinks:
# Each link is expected to match a note name (without .md)
backlinks_map.setdefault(link, []).append(note_name)
# Optionally, extract tasks from the text.
if self.extract_tasks:
tasks, cleaned_text = self._extract_tasks(doc.text)
doc.metadata["tasks"] = tasks
if self.remove_tasks_from_text:
md_docs[i] = Document(
text=cleaned_text, metadata=doc.metadata
)
docs.extend(md_docs)
# Now that we have processed all files, assign backlinks metadata.
for doc in docs:
note_name = doc.metadata.get("note_name")
# If no backlinks exist for this note, default to an empty list.
doc.metadata["backlinks"] = backlinks_map.get(note_name, [])
return docs
def load_langchain_documents(self, **load_kwargs: Any) -> List["LCDocument"]:
"""
Loads data in the LangChain document format.
"""
docs = self.load_data(**load_kwargs)
return [d.to_langchain_format() for d in docs]
def _extract_wikilinks(self, text: str) -> List[str]:
"""
Extracts Obsidian wikilinks from the given text.
Matches patterns like:
- [[Note Name]]
- [[Note Name|Alias]]
Returns a list of unique wikilink targets (aliases are ignored).
"""
pattern = r"\[\[([^\]]+)\]\]"
matches = re.findall(pattern, text)
links = []
for match in matches:
# If a pipe is present (e.g. [[Note|Alias]]), take only the part before it.
target = match.split("|")[0].strip()
links.append(target)
return list(set(links))
def _extract_tasks(self, text: str) -> Tuple[List[str], str]:
"""
Extracts markdown tasks from the text.
A task is expected to be a checklist item in markdown, for example:
- [ ] Do something
- [x] Completed task
Returns a tuple:
(list of task strings, text with task lines removed if removal is enabled).
"""
# This regex matches lines starting with '-' or '*' followed by a checkbox.
task_pattern = re.compile(
r"^\s*[-*]\s*\[\s*(?:x|X| )\s*\]\s*(.*)$", re.MULTILINE
)
tasks = task_pattern.findall(text)
cleaned_text = text
if self.remove_tasks_from_text:
cleaned_text = task_pattern.sub("", text)
return tasks, cleaned_text
|
