spiderforce4ai 1.9__tar.gz → 2.1__tar.gz

{spiderforce4ai-1.9 → spiderforce4ai-2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: spiderforce4ai
-Version: 1.9
+Version: 2.1
 Summary: Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service
 Home-page: https://petertam.pro
 Author: Piotr Tamulewicz

{spiderforce4ai-1.9 → spiderforce4ai-2.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spiderforce4ai"
-version = "1.9"
+version = "2.1"
 description = "Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service"
 readme = "README.md"
 authors = [{name = "Piotr Tamulewicz", email = "pt@petertam.pro"}]

{spiderforce4ai-1.9 → spiderforce4ai-2.1}/setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 
 setup(
     name="spiderforce4ai",
-    version="1.9",
+    version="2.1",
     author="Piotr Tamulewicz",
     author_email="pt@petertam.pro",
     description="Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service",

{spiderforce4ai-1.9 → spiderforce4ai-2.1}/spiderforce4ai/__init__.py

@@ -23,7 +23,7 @@ from multiprocessing import Pool
 console = Console()
 
 def extract_metadata_headers(markdown: str, url: str = '') -> str:
-    """Extract metadata and headers from markdown content with SEO formatting."""
+    """Extract metadata and headers from markdown content with enhanced SEO formatting."""
     lines = markdown.split('\n')
     extracted = []
     in_metadata = False
@@ -33,8 +33,9 @@ def extract_metadata_headers(markdown: str, url: str = '') -> str:
         'canonical_url': '',
         'language': ''
     }
+    first_paragraph = ''
     
-    # First pass - collect metadata
+    # First pass - collect metadata and first paragraph
     for i, line in enumerate(lines):
         # Check for metadata block boundaries
         if line.strip() == '---':
@@ -48,47 +49,47 @@ def extract_metadata_headers(markdown: str, url: str = '') -> str:
         # Extract metadata within the block
         if in_metadata:
             if ':' in line:
-                key, value = line.split(':', 1)
-                key = key.strip().lower()
-                value = value.strip()
+                key, value = [part.strip() for part in line.split(':', 1)]
+                key = key.lower()
+                
+                # Handle multi-line values
+                if value.startswith('>'):
+                    value = value[1:].strip()
+                    j = i + 1
+                    while j < len(lines) and lines[j].strip() and not lines[j].strip() == '---':
+                        value += ' ' + lines[j].strip()
+                        j += 1
                 
                 if key == 'title':
                     metadata['title'] = value
-                elif key == 'description':
+                elif key in ['description', 'meta_description', 'og:description', 'meta-description']:
                     metadata['description'] = value
-                elif key == 'canonical_url':
+                elif key in ['canonical_url', 'canonical']:
                     metadata['canonical_url'] = value
-                elif key == 'language':
+                elif key in ['language', 'lang']:
                     metadata['language'] = value
+        elif not in_metadata and not first_paragraph and line.strip() and not line.startswith('#'):
+            first_paragraph = line.strip()
+    
+    # Use first paragraph as fallback description if none found
+    if not metadata['description'] and first_paragraph:
+        metadata['description'] = first_paragraph[:160] + ('...' if len(first_paragraph) > 160 else '')
     
-    # Add formatted metadata section with URL first
-    extracted.append(f"URL: {url}")  # Use the actual crawled URL
-    extracted.append(f"Title: {metadata['title'] or url.split('/')[-2].replace('-', ' ').title()}")  # Fallback to URL segment
+    # Add formatted metadata section
+    extracted.append(f"URL: {url}")
+    extracted.append(f"Title: {metadata['title'] or url.split('/')[-2].replace('-', ' ').title()}")
     extracted.append(f"Description: {metadata['description']}")
-    extracted.append(f"CanonicalUrl: {metadata['canonical_url'] or url}")  # Fallback to crawled URL
-    extracted.append(f"Language: {metadata['language'] or 'en'}")  # Default to 'en' if not specified
+    extracted.append(f"CanonicalUrl: {metadata['canonical_url'] or url}")
+    extracted.append(f"Language: {metadata['language'] or 'en'}")
     extracted.append("")  # Empty line after metadata
     
     # Second pass - process headers
     for line in lines:
         if line.strip().startswith('#'):
-            # Count the number of # symbols
             level = len(line) - len(line.lstrip('#'))
             text = line.lstrip('#').strip()
-            
-            # Format header according to level
-            if level == 1:
-                extracted.append(f"H1: {text}")
-            elif level == 2:
-                extracted.append(f"H2: {text}")
-            elif level == 3:
-                extracted.append(f"H3: {text}")
-            elif level == 4:
-                extracted.append(f"H4: {text}")
-            elif level == 5:
-                extracted.append(f"H5: {text}")
-            elif level == 6:
-                extracted.append(f"H6: {text}")
+            if 1 <= level <= 6:
+                extracted.append(f"H{level}: {text}")
     
     return '\n'.join(extracted)
 
@@ -310,7 +311,7 @@ class SpiderForce4AI:
 
         # Handle combined markdown file
         if self.config.combine_to_one_markdown:
-            content = markdown if self.config.combine_to_one_markdown == 'full' else extract_metadata_headers(markdown)
+            content = markdown if config.combine_to_one_markdown == 'full' else extract_metadata_headers(markdown, url)
             combined_content = f"\n----PAGE----\n{url}\n\n{content}\n----PAGE END----\n"
             
             async with aiofiles.open(self.config.combined_markdown_file, 'a', encoding='utf-8') as f:

{spiderforce4ai-1.9 → spiderforce4ai-2.1}/spiderforce4ai.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: spiderforce4ai
-Version: 1.9
+Version: 2.1
 Summary: Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service
 Home-page: https://petertam.pro
 Author: Piotr Tamulewicz