Aktualisierung der atomaren Sektions-Chunking-Strategie auf Version 3.9.5 mit Implementierung des 'Pack-and-Carry-Over' Verfahrens. Einführung neuer Konfigurationsoptionen für Smart-Edge und strikte Überschriftenteilung. Verbesserte Handhabung von leeren Überschriften und Anpassungen an der Warteschlangen-Verarbeitung zur Optimierung der Chunk-Erstellung.
This commit is contained in:
1 parent
8f65e550c8
commit
3c5c567077
2 files changed
+67
-72
No files matched your search
@@ -8,19 +8,13 @@ from typing import List, Tuple, Set
|
||||
from .chunking_models import RawBlock
|
||||
from .chunking_utils import extract_frontmatter_from_text
|
||||
|
||||
def split_sentences(text: str) -> list[str]:
|
||||
"""Teilt Text in Sätze auf unter Berücksichtigung deutscher Interpunktion."""
|
||||
text = re.sub(r'\s+', ' ', text.strip())
|
||||
if not text: return []
|
||||
# Splittet bei Satzzeichen, gefolgt von Leerzeichen und Großbuchstaben
|
||||
return [s.strip() for s in re.split(r'(?<=[.!?])\s+(?=[A-ZÄÖÜ0-9„(])', text) if s.strip()]
|
||||
|
||||
def parse_blocks(md_text: str) -> Tuple[List[RawBlock], str]:
|
||||
"""Zerlegt Text in logische Einheiten (RawBlocks), inklusive H1-H6."""
|
||||
blocks = []
|
||||
h1_title = "Dokument"; section_path = "/"; current_section_title = None
|
||||
fm, text_without_fm = extract_frontmatter_from_text(md_text)
|
||||
|
||||
# H1 für Metadaten extrahieren
|
||||
h1_match = re.search(r'^#\s+(.*)', text_without_fm, re.MULTILINE)
|
||||
if h1_match: h1_title = h1_match.group(1).strip()
|
||||
|
||||
@@ -32,20 +26,26 @@ def parse_blocks(md_text: str) -> Tuple[List[RawBlock], str]:
|
||||
heading_match = re.match(r'^(#{1,6})\s+(.*)', stripped)
|
||||
|
||||
if heading_match:
|
||||
# Vorherigen Text-Block abschließen
|
||||
if buffer:
|
||||
content = "\n".join(buffer).strip()
|
||||
if content:
|
||||
blocks.append(RawBlock("paragraph", content, None, section_path, current_section_title))
|
||||
buffer = []
|
||||
|
||||
level = len(heading_match.group(1))
|
||||
title = heading_match.group(2).strip()
|
||||
|
||||
# Pfad- und Titel-Update
|
||||
if level == 1:
|
||||
current_section_title = title; section_path = "/"
|
||||
elif level == 2:
|
||||
current_section_title = title; section_path = f"/{current_section_title}"
|
||||
|
||||
blocks.append(RawBlock("heading", stripped, level, section_path, current_section_title))
|
||||
continue
|
||||
|
||||
# Trenner oder Leerzeilen beenden Blöcke, außer innerhalb von Callouts
|
||||
if (not stripped or stripped == "---") and not line.startswith('>'):
|
||||
if buffer:
|
||||
content = "\n".join(buffer).strip()
|
||||
@@ -60,28 +60,12 @@ def parse_blocks(md_text: str) -> Tuple[List[RawBlock], str]:
|
||||
if buffer:
|
||||
content = "\n".join(buffer).strip()
|
||||
if content: blocks.append(RawBlock("paragraph", content, None, section_path, current_section_title))
|
||||
|
||||
return blocks, h1_title
|
||||
|
||||
def parse_edges_robust(text: str) -> Set[str]:
|
||||
"""Extrahiert Kanten-Kandidaten aus Wikilinks und Callouts."""
|
||||
found_edges = set()
|
||||
inlines = re.findall(r'\[\[rel:([^\|\]]+)\|?([^\]]*)\]\]', text)
|
||||
for kind, target in inlines:
|
||||
k = kind.strip().lower(); t = target.strip()
|
||||
if k and t: found_edges.add(f"{k}:{t}")
|
||||
lines = text.split('\n'); current_edge_type = None
|
||||
for line in lines:
|
||||
stripped = line.strip()
|
||||
callout_match = re.match(r'>\s*\[!edge\]\s*([^:\s]+)', stripped)
|
||||
if callout_match:
|
||||
current_edge_type = callout_match.group(1).strip().lower()
|
||||
links = re.findall(r'\[\[([^\]]+)\]\]', stripped)
|
||||
for l in links:
|
||||
if "rel:" not in l: found_edges.add(f"{current_edge_type}:{l}")
|
||||
continue
|
||||
if current_edge_type and stripped.startswith('>'):
|
||||
links = re.findall(r'\[\[([^\]]+)\]\]', stripped)
|
||||
for l in links:
|
||||
if "rel:" not in l: found_edges.add(f"{current_edge_type}:{l}")
|
||||
elif not stripped.startswith('>'): current_edge_type = None
|
||||
return found_edges
|
||||
def split_sentences(text: str) -> list[str]:
|
||||
"""Teilt Text in Sätze auf unter Berücksichtigung deutscher Interpunktion."""
|
||||
text = re.sub(r'\s+', ' ', text.strip())
|
||||
if not text: return []
|
||||
# Splittet bei Satzzeichen, gefolgt von Leerzeichen und Großbuchstaben
|
||||
return [s.strip() for s in re.split(r'(?<=[.!?])\s+(?=[A-ZÄÖÜ0-9„(])', text) if s.strip()]
|
||||
Reference in new issue
Block a user