Spaces:
Sleeping
Sleeping
Update web_scraper_tool.py
Browse files- web_scraper_tool.py +219 -157
web_scraper_tool.py
CHANGED
|
@@ -6,6 +6,55 @@ from urllib.parse import urlparse, urlunparse, urljoin
|
|
| 6 |
import tempfile
|
| 7 |
import os
|
| 8 |
import re
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
class WebScrapperTool:
|
| 11 |
def __init__(self):
|
|
@@ -13,17 +62,34 @@ class WebScrapperTool:
|
|
| 13 |
self.session.headers.update({
|
| 14 |
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36"
|
| 15 |
})
|
| 16 |
-
self.font_path = self.
|
| 17 |
if not self.font_path:
|
| 18 |
-
print("
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
-
def _find_font(self):
|
| 21 |
-
font_name = 'DejaVuSansCondensed.ttf'
|
| 22 |
-
if os.path.exists(font_name): return font_name
|
| 23 |
-
if os.path.exists(os.path.join('fonts', font_name)): return os.path.join('fonts', font_name)
|
| 24 |
-
return None
|
| 25 |
|
| 26 |
def normalize_url(self, url: str) -> str:
|
|
|
|
| 27 |
url = url.strip()
|
| 28 |
parsed_url = urlparse(url)
|
| 29 |
scheme = parsed_url.scheme
|
|
@@ -42,43 +108,38 @@ class WebScrapperTool:
|
|
| 42 |
return urlunparse(parsed_url)
|
| 43 |
|
| 44 |
def is_image_url(self, url: str) -> bool:
|
|
|
|
| 45 |
image_extensions = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.svg', '.webp']
|
| 46 |
parsed_url = urlparse(url)
|
| 47 |
return any(parsed_url.path.lower().endswith(ext) for ext in image_extensions)
|
| 48 |
|
| 49 |
def _get_content(self, url: str, is_for_image_download=False):
|
|
|
|
| 50 |
try:
|
| 51 |
-
# Si es para descargar una imagen específica, el stream es útil.
|
| 52 |
-
# Si es para contenido general, stream=False es usualmente mejor para que response.content esté completo.
|
| 53 |
stream_setting = True if is_for_image_download or self.is_image_url(url) else False
|
| 54 |
response = self.session.get(url, timeout=20, allow_redirects=True, stream=stream_setting)
|
| 55 |
response.raise_for_status()
|
| 56 |
-
|
| 57 |
content_type_header = response.headers.get('content-type', '').lower()
|
| 58 |
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
raw_content = response.content # Leer todo
|
| 62 |
return None, raw_content, content_type_header
|
| 63 |
-
|
| 64 |
-
# Si se llamó específicamente para descargar una imagen (y no es html)
|
| 65 |
if is_for_image_download and 'image' in content_type_header:
|
| 66 |
return None, response.content, content_type_header
|
| 67 |
|
| 68 |
-
# Para contenido textual
|
| 69 |
try:
|
| 70 |
content_text = response.content.decode('utf-8')
|
| 71 |
except UnicodeDecodeError:
|
| 72 |
content_text = response.text
|
| 73 |
-
|
| 74 |
return content_text, response.content, content_type_header
|
| 75 |
except requests.exceptions.Timeout:
|
| 76 |
return None, None, f"Error: Timeout al acceder a la URL: {url}"
|
| 77 |
except requests.exceptions.RequestException as e:
|
| 78 |
return None, None, f"Error de conexión/HTTP ({url}): {str(e)}"
|
| 79 |
|
|
|
|
| 80 |
def scrape_to_text(self, url: str):
|
| 81 |
-
# ... (
|
| 82 |
text_content, _, content_type_or_error_msg = self._get_content(url)
|
| 83 |
|
| 84 |
if text_content is None and not ('image' in content_type_or_error_msg):
|
|
@@ -88,7 +149,7 @@ class WebScrapperTool:
|
|
| 88 |
final_text = ""
|
| 89 |
if 'text/html' in content_type_or_error_msg and text_content:
|
| 90 |
soup = BeautifulSoup(text_content, 'html.parser')
|
| 91 |
-
for element in soup(["script", "style", "nav", "footer", "aside", "form", "button", "input", "header", "figure", "figcaption"]):
|
| 92 |
element.decompose()
|
| 93 |
body = soup.find('body')
|
| 94 |
if body:
|
|
@@ -96,11 +157,10 @@ class WebScrapperTool:
|
|
| 96 |
final_text = "\n".join(text_items)
|
| 97 |
else:
|
| 98 |
final_text = "\n".join([s.strip() for s in soup.stripped_strings if s.strip()])
|
| 99 |
-
|
| 100 |
elif 'text/plain' in content_type_or_error_msg and text_content:
|
| 101 |
final_text = text_content
|
| 102 |
elif self.is_image_url(url) or ('image' in content_type_or_error_msg):
|
| 103 |
-
return {'status': 'error', 'message': f"La URL apunta a una imagen. El formato TXT es para contenido textual.
|
| 104 |
elif text_content:
|
| 105 |
final_text = text_content
|
| 106 |
else:
|
|
@@ -120,159 +180,161 @@ class WebScrapperTool:
|
|
| 120 |
|
| 121 |
|
| 122 |
def scrape_to_pdf(self, url: str):
|
| 123 |
-
|
|
|
|
| 124 |
|
| 125 |
-
|
| 126 |
-
|
| 127 |
|
| 128 |
-
|
| 129 |
|
| 130 |
-
|
| 131 |
-
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
|
| 136 |
-
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 140 |
|
| 141 |
|
| 142 |
-
|
| 143 |
-
|
| 144 |
-
img_suffix = '.' + content_type_or_error_msg.split('/')[-1].split(';')[0].strip()
|
| 145 |
-
if img_suffix == '.': img_suffix = '.jpg'
|
| 146 |
-
valid_img_suffixes = ['.jpeg', '.jpg', '.png']
|
| 147 |
-
if img_suffix not in valid_img_suffixes:
|
| 148 |
-
if 'png' in img_suffix: img_suffix = '.png'
|
| 149 |
-
else: img_suffix = '.jpg'
|
| 150 |
-
|
| 151 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=img_suffix) as tmp_img:
|
| 152 |
-
tmp_img.write(raw_content)
|
| 153 |
-
img_path = tmp_img.name
|
| 154 |
-
|
| 155 |
try:
|
| 156 |
-
|
| 157 |
-
|
| 158 |
-
|
| 159 |
-
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
|
| 163 |
-
|
| 164 |
-
|
| 165 |
-
|
| 166 |
-
|
| 167 |
-
|
| 168 |
-
|
| 169 |
-
|
| 170 |
-
|
| 171 |
-
|
| 172 |
-
|
| 173 |
-
|
| 174 |
-
|
| 175 |
-
|
| 176 |
-
for element in soup(["script", "style", "nav", "footer", "aside", "form", "button", "input", "header"]):
|
| 177 |
-
element.decompose()
|
| 178 |
|
| 179 |
-
|
| 180 |
-
|
| 181 |
-
|
| 182 |
-
|
| 183 |
-
|
| 184 |
-
|
| 185 |
-
|
| 186 |
-
|
| 187 |
-
|
| 188 |
-
|
| 189 |
-
|
| 190 |
-
|
| 191 |
-
|
| 192 |
-
|
| 193 |
-
|
| 194 |
-
|
| 195 |
-
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
|
| 199 |
-
|
| 200 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
try:
|
| 202 |
-
|
| 203 |
-
|
| 204 |
-
pdf.
|
| 205 |
-
|
| 206 |
-
|
| 207 |
-
print(f"Error FPDF al añadir imagen {img_url_abs}: {e_fpdf_img}")
|
| 208 |
-
pdf.set_font(current_font, 'I', 9) # Cursiva y pequeño
|
| 209 |
-
pdf.multi_cell(0,5, f"[Error al renderizar imagen: {img_url_abs} - {e_fpdf_img}]")
|
| 210 |
-
pdf.set_font(current_font, '', 11) # Volver a fuente normal
|
| 211 |
-
finally:
|
| 212 |
-
if os.path.exists(tmp_img_path): os.unlink(tmp_img_path)
|
| 213 |
else:
|
| 214 |
-
|
| 215 |
-
except Exception as e_dl_img:
|
| 216 |
-
print(f"Excepción al descargar/procesar imagen {img_url_abs}: {e_dl_img}")
|
| 217 |
-
pdf.set_font(current_font, 'I', 9)
|
| 218 |
-
pdf.multi_cell(0,5, f"[Error al descargar imagen: {img_url_abs}]")
|
| 219 |
-
pdf.set_font(current_font, '', 11)
|
| 220 |
-
pdf.ln(5) # Espacio después del intento de imagen
|
| 221 |
-
|
| 222 |
-
# Manejar texto dentro de párrafos, divs, etc.
|
| 223 |
-
# Tomar texto solo de ciertos elementos o el texto 'directo' del elemento actual.
|
| 224 |
-
# Esto evita duplicar texto si `element.stripped_strings` se usa en un nodo padre.
|
| 225 |
-
# Tomar texto que es hijo directo del elemento actual y no está dentro de otro 'img' o bloque ya procesado.
|
| 226 |
-
elif element.name in ['p', 'div', 'span', 'li', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'td', 'th', 'caption', 'article', 'section', 'blockquote']:
|
| 227 |
-
# Procesar el texto que es hijo directo (string) de este elemento
|
| 228 |
-
current_element_text = ""
|
| 229 |
-
for content_child in element.contents:
|
| 230 |
-
if isinstance(content_child, str) and content_child.strip():
|
| 231 |
-
current_element_text += content_child.strip() + " "
|
| 232 |
-
|
| 233 |
-
if current_element_text.strip():
|
| 234 |
-
clean_para = self._clean_text_for_pdf(current_element_text.strip())
|
| 235 |
-
if element.name.startswith('h'): # Estilo para encabezados
|
| 236 |
-
pdf.set_font(current_font, 'B', 14 - int(element.name[1])) # h1=13, h2=12, etc.
|
| 237 |
-
pdf.multi_cell(0, 7, clean_para)
|
| 238 |
-
pdf.set_font(current_font, '', 11) # Reset
|
| 239 |
-
else:
|
| 240 |
-
pdf.multi_cell(0, 7, clean_para)
|
| 241 |
-
pdf.ln(1) # Pequeño espacio entre párrafos de texto
|
| 242 |
-
|
| 243 |
-
# Si después de todo no se añadió contenido, error
|
| 244 |
-
if pdf.page_no() == 1 and pdf.y < 30: # Heurística: si no se ha escrito mucho en la primera página
|
| 245 |
-
return {'status': 'error', 'message': "No se encontró contenido textual o imágenes extraíbles de la página HTML.", 'url': url}
|
| 246 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 247 |
|
| 248 |
-
|
| 249 |
-
|
| 250 |
-
|
| 251 |
-
|
| 252 |
-
|
| 253 |
-
|
| 254 |
-
|
| 255 |
-
|
| 256 |
-
|
|
|
|
| 257 |
|
| 258 |
-
# Guardar el PDF
|
| 259 |
-
try:
|
| 260 |
with tempfile.NamedTemporaryFile(delete=False, mode='wb', suffix='.pdf') as tmp_file:
|
| 261 |
pdf_output_bytes = pdf.output(dest='S')
|
| 262 |
tmp_file.write(pdf_output_bytes)
|
| 263 |
filepath = tmp_file.name
|
| 264 |
return {'status': 'success', 'file': filepath, 'url': url}
|
| 265 |
-
|
| 266 |
-
|
| 267 |
tb_str = traceback.format_exc()
|
| 268 |
-
error_message = f"Error
|
| 269 |
-
if len(error_message) >
|
|
|
|
| 270 |
return {'status': 'error', 'message': error_message, 'url': url}
|
| 271 |
|
| 272 |
-
|
| 273 |
-
clean = text.replace('\u2013', '-').replace('\u2014', '--')
|
| 274 |
-
clean = clean.replace('\u2018', "'").replace('\u2019', "'")
|
| 275 |
-
clean = clean.replace('\u201c', '"').replace('\u201d', '"')
|
| 276 |
-
clean = clean.replace('\u2026', '...')
|
| 277 |
-
clean = clean.replace('\u00A0', ' ')
|
| 278 |
-
return "".join(c for c in clean if c.isprintable() or c in ('\n', '\r', '\t'))
|
|
|
|
| 6 |
import tempfile
|
| 7 |
import os
|
| 8 |
import re
|
| 9 |
+
import traceback # Para un mejor logging de errores
|
| 10 |
+
|
| 11 |
+
# Helper para limpiar texto para FPDF, especialmente con fuentes no Unicode
|
| 12 |
+
def clean_problematic_chars(text, use_unicode_font=False):
|
| 13 |
+
"""
|
| 14 |
+
Limpia o reemplaza caracteres que suelen causar problemas en FPDF,
|
| 15 |
+
especialmente si no se usa una fuente Unicode completa.
|
| 16 |
+
"""
|
| 17 |
+
if use_unicode_font:
|
| 18 |
+
# Con una fuente Unicode, menos reemplazos son necesarios, pero algunos
|
| 19 |
+
# caracteres de control o muy específicos aún pueden causar problemas.
|
| 20 |
+
# El reemplazo de espacios de no ruptura es generalmente seguro.
|
| 21 |
+
text = text.replace('\u00A0', ' ') # No-breaking space
|
| 22 |
+
# Podrías añadir más reemplazos específicos para fuentes Unicode si encuentras problemas
|
| 23 |
+
else:
|
| 24 |
+
# Para fuentes no Unicode (latin-1 like)
|
| 25 |
+
replacements = {
|
| 26 |
+
'\u20AC': 'EUR', # Euro sign
|
| 27 |
+
'\u00A3': 'GBP', # Pound sign
|
| 28 |
+
'\u00A5': 'JPY', # Yen sign
|
| 29 |
+
'\u2013': '-', # En Dash
|
| 30 |
+
'\u2014': '--', # Em Dash
|
| 31 |
+
'\u2018': "'", # Left single quotation mark
|
| 32 |
+
'\u2019': "'", # Right single quotation mark
|
| 33 |
+
'\u201C': '"', # Left double quotation mark
|
| 34 |
+
'\u201D': '"', # Right double quotation mark
|
| 35 |
+
'\u2026': '...', # Horizontal ellipsis
|
| 36 |
+
'\u00A0': ' ', # No-breaking space
|
| 37 |
+
'\u00A9': '(C)', # Copyright
|
| 38 |
+
'\u00AE': '(R)', # Registered trademark
|
| 39 |
+
# Añade más según sea necesario
|
| 40 |
+
}
|
| 41 |
+
for problematic, replacement in replacements.items():
|
| 42 |
+
text = text.replace(problematic, replacement)
|
| 43 |
+
|
| 44 |
+
# Filtrar cualquier cosa que no sea imprimible o no esté en latin-1 aproximado
|
| 45 |
+
# Esto es agresivo y puede perder caracteres.
|
| 46 |
+
text = "".join(c for c in text if c.isprintable() or c in ('\n', '\r', '\t'))
|
| 47 |
+
try:
|
| 48 |
+
# Intenta codificar a latin-1 y decodificar para eliminar caracteres no compatibles
|
| 49 |
+
text = text.encode('latin-1', 'ignore').decode('latin-1')
|
| 50 |
+
except Exception:
|
| 51 |
+
# Si falla, recurre a una limpieza aún más básica
|
| 52 |
+
text = "".join(c for c in text if ord(c) < 256 or c in ('\n', '\r', '\t'))
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
# Limpieza general de caracteres de control restantes (excepto tab, lf, cr)
|
| 56 |
+
return "".join(c for c in text if c.isprintable() or c in ('\n', '\r', '\t'))
|
| 57 |
+
|
| 58 |
|
| 59 |
class WebScrapperTool:
|
| 60 |
def __init__(self):
|
|
|
|
| 62 |
self.session.headers.update({
|
| 63 |
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36"
|
| 64 |
})
|
| 65 |
+
self.font_path, self.font_name_for_fpdf = self._find_and_setup_font()
|
| 66 |
if not self.font_path:
|
| 67 |
+
print("ADVERTENCIA: No se encontró 'DejaVuSansCondensed.ttf'. Se usará Arial para PDFs (soporte Unicode limitado).")
|
| 68 |
+
print("Para un mejor soporte de caracteres internacionales, descargue DejaVuSansCondensed.ttf y colóquelo en la raíz del proyecto o en una carpeta 'fonts'.")
|
| 69 |
+
else:
|
| 70 |
+
print(f"INFO: Usando fuente {self.font_name_for_fpdf} desde {self.font_path} para PDFs.")
|
| 71 |
+
self.using_unicode_font = bool(self.font_path)
|
| 72 |
+
|
| 73 |
+
|
| 74 |
+
def _find_and_setup_font(self):
|
| 75 |
+
# Devuelve (ruta_completa_fuente, nombre_familia_para_fpdf) o (None, 'Arial')
|
| 76 |
+
font_file_name = 'DejaVuSansCondensed.ttf'
|
| 77 |
+
font_family_name = 'DejaVu' # Nombre que usaremos en FPDF
|
| 78 |
+
|
| 79 |
+
# Buscar en el directorio actual
|
| 80 |
+
if os.path.exists(font_file_name):
|
| 81 |
+
return os.path.abspath(font_file_name), font_family_name
|
| 82 |
+
|
| 83 |
+
# Buscar en una subcarpeta 'fonts'
|
| 84 |
+
fonts_dir_path = os.path.join(os.path.dirname(__file__), 'fonts', font_file_name)
|
| 85 |
+
if os.path.exists(fonts_dir_path):
|
| 86 |
+
return os.path.abspath(fonts_dir_path), font_family_name
|
| 87 |
+
|
| 88 |
+
return None, 'Arial' # Fallback a fuente core de FPDF
|
| 89 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 90 |
|
| 91 |
def normalize_url(self, url: str) -> str:
|
| 92 |
+
# ... (sin cambios)
|
| 93 |
url = url.strip()
|
| 94 |
parsed_url = urlparse(url)
|
| 95 |
scheme = parsed_url.scheme
|
|
|
|
| 108 |
return urlunparse(parsed_url)
|
| 109 |
|
| 110 |
def is_image_url(self, url: str) -> bool:
|
| 111 |
+
# ... (sin cambios)
|
| 112 |
image_extensions = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.svg', '.webp']
|
| 113 |
parsed_url = urlparse(url)
|
| 114 |
return any(parsed_url.path.lower().endswith(ext) for ext in image_extensions)
|
| 115 |
|
| 116 |
def _get_content(self, url: str, is_for_image_download=False):
|
| 117 |
+
# ... (sin cambios significativos, quizás logging)
|
| 118 |
try:
|
|
|
|
|
|
|
| 119 |
stream_setting = True if is_for_image_download or self.is_image_url(url) else False
|
| 120 |
response = self.session.get(url, timeout=20, allow_redirects=True, stream=stream_setting)
|
| 121 |
response.raise_for_status()
|
|
|
|
| 122 |
content_type_header = response.headers.get('content-type', '').lower()
|
| 123 |
|
| 124 |
+
if 'image' in content_type_header or (self.is_image_url(url) and not is_for_image_download):
|
| 125 |
+
raw_content = response.content
|
|
|
|
| 126 |
return None, raw_content, content_type_header
|
|
|
|
|
|
|
| 127 |
if is_for_image_download and 'image' in content_type_header:
|
| 128 |
return None, response.content, content_type_header
|
| 129 |
|
|
|
|
| 130 |
try:
|
| 131 |
content_text = response.content.decode('utf-8')
|
| 132 |
except UnicodeDecodeError:
|
| 133 |
content_text = response.text
|
|
|
|
| 134 |
return content_text, response.content, content_type_header
|
| 135 |
except requests.exceptions.Timeout:
|
| 136 |
return None, None, f"Error: Timeout al acceder a la URL: {url}"
|
| 137 |
except requests.exceptions.RequestException as e:
|
| 138 |
return None, None, f"Error de conexión/HTTP ({url}): {str(e)}"
|
| 139 |
|
| 140 |
+
|
| 141 |
def scrape_to_text(self, url: str):
|
| 142 |
+
# ... (sin cambios)
|
| 143 |
text_content, _, content_type_or_error_msg = self._get_content(url)
|
| 144 |
|
| 145 |
if text_content is None and not ('image' in content_type_or_error_msg):
|
|
|
|
| 149 |
final_text = ""
|
| 150 |
if 'text/html' in content_type_or_error_msg and text_content:
|
| 151 |
soup = BeautifulSoup(text_content, 'html.parser')
|
| 152 |
+
for element in soup(["script", "style", "nav", "footer", "aside", "form", "button", "input", "header", "figure", "figcaption"]):
|
| 153 |
element.decompose()
|
| 154 |
body = soup.find('body')
|
| 155 |
if body:
|
|
|
|
| 157 |
final_text = "\n".join(text_items)
|
| 158 |
else:
|
| 159 |
final_text = "\n".join([s.strip() for s in soup.stripped_strings if s.strip()])
|
|
|
|
| 160 |
elif 'text/plain' in content_type_or_error_msg and text_content:
|
| 161 |
final_text = text_content
|
| 162 |
elif self.is_image_url(url) or ('image' in content_type_or_error_msg):
|
| 163 |
+
return {'status': 'error', 'message': f"La URL apunta a una imagen. El formato TXT es para contenido textual.", 'url': url}
|
| 164 |
elif text_content:
|
| 165 |
final_text = text_content
|
| 166 |
else:
|
|
|
|
| 180 |
|
| 181 |
|
| 182 |
def scrape_to_pdf(self, url: str):
|
| 183 |
+
try: # Envolver todo el proceso de PDF para capturar errores de forma más general
|
| 184 |
+
text_content, raw_content, content_type_or_error_msg = self._get_content(url)
|
| 185 |
|
| 186 |
+
if text_content is None and raw_content is None:
|
| 187 |
+
return {'status': 'error', 'message': content_type_or_error_msg, 'url': url}
|
| 188 |
|
| 189 |
+
is_direct_image_url = 'image' in content_type_or_error_msg or self.is_image_url(url)
|
| 190 |
|
| 191 |
+
pdf = FPDF()
|
| 192 |
+
pdf.add_page()
|
| 193 |
+
pdf.set_auto_page_break(auto=True, margin=15)
|
| 194 |
+
|
| 195 |
+
current_fpdf_font_name = self.font_name_for_fpdf # Nombre de familia para FPDF
|
| 196 |
+
|
| 197 |
+
if self.using_unicode_font:
|
| 198 |
+
try:
|
| 199 |
+
pdf.add_font(self.font_name_for_fpdf, '', self.font_path, uni=True)
|
| 200 |
+
print(f"INFO: Fuente Unicode '{self.font_name_for_fpdf}' registrada en FPDF.")
|
| 201 |
+
except Exception as e_font:
|
| 202 |
+
print(f"ERROR al registrar fuente Unicode '{self.font_name_for_fpdf}' desde '{self.font_path}': {e_font}")
|
| 203 |
+
traceback.print_exc()
|
| 204 |
+
print("ADVERTENCIA: Recurriendo a fuente Arial debido a error con fuente Unicode.")
|
| 205 |
+
current_fpdf_font_name = 'Arial' # Fallback
|
| 206 |
+
self.using_unicode_font = False # Actualizar estado
|
| 207 |
+
else: # No estamos usando fuente Unicode (no se encontró o falló al cargarla)
|
| 208 |
+
print("INFO: No se está usando una fuente Unicode. El soporte de caracteres será limitado.")
|
| 209 |
|
| 210 |
|
| 211 |
+
if is_direct_image_url and raw_content:
|
| 212 |
+
# ... (lógica de imagen directa sin cambios significativos en limpieza de texto)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 213 |
try:
|
| 214 |
+
img_suffix = '.' + content_type_or_error_msg.split('/')[-1].split(';')[0].strip()
|
| 215 |
+
if img_suffix == '.': img_suffix = '.jpg'
|
| 216 |
+
valid_img_suffixes = ['.jpeg', '.jpg', '.png']
|
| 217 |
+
if img_suffix not in valid_img_suffixes:
|
| 218 |
+
if 'png' in img_suffix: img_suffix = '.png'
|
| 219 |
+
else: img_suffix = '.jpg'
|
| 220 |
+
|
| 221 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=img_suffix) as tmp_img:
|
| 222 |
+
tmp_img.write(raw_content)
|
| 223 |
+
img_path = tmp_img.name
|
| 224 |
+
|
| 225 |
+
try:
|
| 226 |
+
page_width = pdf.w - 2 * pdf.l_margin
|
| 227 |
+
pdf.image(img_path, x=pdf.l_margin, y=pdf.t_margin, w=page_width)
|
| 228 |
+
except RuntimeError as re_img:
|
| 229 |
+
return {'status': 'error', 'message': f"Error al añadir imagen directa al PDF ({img_suffix}): {str(re_img)}", 'url': url}
|
| 230 |
+
finally:
|
| 231 |
+
if os.path.exists(img_path): os.unlink(img_path)
|
| 232 |
+
except Exception as e_img:
|
| 233 |
+
return {'status': 'error', 'message': f"Error procesando imagen directa para PDF: {str(e_img)}", 'url': url}
|
|
|
|
|
|
|
| 234 |
|
| 235 |
+
elif 'text/html' in content_type_or_error_msg and text_content:
|
| 236 |
+
soup = BeautifulSoup(text_content, 'html.parser')
|
| 237 |
+
|
| 238 |
+
pdf.set_font(current_fpdf_font_name, 'B', 12)
|
| 239 |
+
cleaned_url_title = clean_problematic_chars(f"Contenido de: {url}", self.using_unicode_font)
|
| 240 |
+
pdf.multi_cell(0, 8, cleaned_url_title)
|
| 241 |
+
pdf.ln(6)
|
| 242 |
+
pdf.set_font(current_fpdf_font_name, '', 11)
|
| 243 |
+
|
| 244 |
+
for element in soup(["script", "style", "nav", "footer", "aside", "form", "button", "input", "header"]):
|
| 245 |
+
element.decompose()
|
| 246 |
+
|
| 247 |
+
content_area = soup.find('main') or soup.find('article') or soup.find('body')
|
| 248 |
+
if not content_area:
|
| 249 |
+
return {'status': 'error', 'message': "No se encontró área de contenido principal.", 'url': url}
|
| 250 |
+
|
| 251 |
+
for element in content_area.find_all(recursive=True):
|
| 252 |
+
if isinstance(element, Tag):
|
| 253 |
+
if element.name == 'img':
|
| 254 |
+
img_src = element.get('src') or element.get('data-src')
|
| 255 |
+
if img_src:
|
| 256 |
+
img_url_abs = urljoin(url, img_src)
|
| 257 |
+
pdf.ln(5)
|
| 258 |
+
try:
|
| 259 |
+
# print(f"Intentando descargar imagen: {img_url_abs}")
|
| 260 |
+
_, img_data, img_content_type = self._get_content(img_url_abs, is_for_image_download=True)
|
| 261 |
+
if img_data and 'image' in img_content_type:
|
| 262 |
+
img_sfx = '.' + img_content_type.split('/')[-1].split(';')[0].strip()
|
| 263 |
+
if img_sfx == '.': img_sfx = '.jpg'
|
| 264 |
+
|
| 265 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=img_sfx) as tmp_img_file:
|
| 266 |
+
tmp_img_file.write(img_data)
|
| 267 |
+
tmp_img_path = tmp_img_file.name
|
| 268 |
+
|
| 269 |
+
try:
|
| 270 |
+
page_w = pdf.w - 2 * pdf.l_margin
|
| 271 |
+
pdf.image(tmp_img_path, x=None, y=None, w=page_w)
|
| 272 |
+
pdf.ln(2)
|
| 273 |
+
# print(f"Imagen {img_url_abs} añadida al PDF.")
|
| 274 |
+
except RuntimeError as e_fpdf_img:
|
| 275 |
+
print(f"Error FPDF al añadir imagen {img_url_abs}: {e_fpdf_img}")
|
| 276 |
+
pdf.set_font(current_fpdf_font_name, 'I', 9)
|
| 277 |
+
err_img_msg = clean_problematic_chars(f"[Error render img: {img_url_abs} - {e_fpdf_img}]", self.using_unicode_font)
|
| 278 |
+
pdf.multi_cell(0,5, err_img_msg)
|
| 279 |
+
pdf.set_font(current_fpdf_font_name, '', 11)
|
| 280 |
+
finally:
|
| 281 |
+
if os.path.exists(tmp_img_path): os.unlink(tmp_img_path)
|
| 282 |
+
# else: print(f"No se pudo descargar o no es una imagen: {img_url_abs}")
|
| 283 |
+
except Exception as e_dl_img:
|
| 284 |
+
print(f"Excepción al descargar/procesar imagen {img_url_abs}: {e_dl_img}")
|
| 285 |
+
pdf.set_font(current_fpdf_font_name, 'I', 9)
|
| 286 |
+
err_dl_msg = clean_problematic_chars(f"[Error download img: {img_url_abs}]", self.using_unicode_font)
|
| 287 |
+
pdf.multi_cell(0,5, err_dl_msg)
|
| 288 |
+
pdf.set_font(current_fpdf_font_name, '', 11)
|
| 289 |
+
pdf.ln(5)
|
| 290 |
+
|
| 291 |
+
elif element.name in ['p', 'div', 'span', 'li', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'td', 'th', 'caption', 'article', 'section', 'blockquote']:
|
| 292 |
+
current_element_text = ""
|
| 293 |
+
for content_child in element.contents:
|
| 294 |
+
if isinstance(content_child, str) and content_child.strip():
|
| 295 |
+
current_element_text += content_child.strip() + " "
|
| 296 |
+
|
| 297 |
+
if current_element_text.strip():
|
| 298 |
+
clean_para = clean_problematic_chars(current_element_text.strip(), self.using_unicode_font)
|
| 299 |
+
if element.name.startswith('h') and len(element.name) == 2 : # h1, h2 .. h6
|
| 300 |
try:
|
| 301 |
+
header_level = int(element.name[1])
|
| 302 |
+
font_size = max(8, 16 - header_level * 1.5) # Ajusta el tamaño base y el decremento
|
| 303 |
+
pdf.set_font(current_fpdf_font_name, 'B', font_size)
|
| 304 |
+
except ValueError: # por si acaso element.name no es h[numero]
|
| 305 |
+
pdf.set_font(current_fpdf_font_name, 'B', 11) # fallback a negrita normal
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 306 |
else:
|
| 307 |
+
pdf.set_font(current_fpdf_font_name, '', 11) # Texto normal
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 308 |
|
| 309 |
+
pdf.multi_cell(0, 7, clean_para)
|
| 310 |
+
pdf.set_font(current_fpdf_font_name, '', 11) # Reset font a normal para el siguiente elemento
|
| 311 |
+
pdf.ln(1)
|
| 312 |
+
|
| 313 |
+
if pdf.page_no() == 1 and pdf.y < 30: # (y después de la URL del título)
|
| 314 |
+
return {'status': 'error', 'message': "No se encontró contenido textual o imágenes extraíbles de la página HTML.", 'url': url}
|
| 315 |
|
| 316 |
+
elif 'text/plain' in content_type_or_error_msg and text_content:
|
| 317 |
+
pdf.set_font(current_fpdf_font_name, 'B', 12)
|
| 318 |
+
cleaned_url_title = clean_problematic_chars(f"Contenido de: {url}", self.using_unicode_font)
|
| 319 |
+
pdf.multi_cell(0, 8, cleaned_url_title)
|
| 320 |
+
pdf.ln(6)
|
| 321 |
+
pdf.set_font(current_fpdf_font_name, '', 11)
|
| 322 |
+
clean_text_content = clean_problematic_chars(text_content, self.using_unicode_font)
|
| 323 |
+
pdf.multi_cell(0, 7, clean_text_content)
|
| 324 |
+
else:
|
| 325 |
+
return {'status': 'error', 'message': f"Tipo de contenido no soportado o vacío para PDF: {content_type_or_error_msg}", 'url': url}
|
| 326 |
|
|
|
|
|
|
|
| 327 |
with tempfile.NamedTemporaryFile(delete=False, mode='wb', suffix='.pdf') as tmp_file:
|
| 328 |
pdf_output_bytes = pdf.output(dest='S')
|
| 329 |
tmp_file.write(pdf_output_bytes)
|
| 330 |
filepath = tmp_file.name
|
| 331 |
return {'status': 'success', 'file': filepath, 'url': url}
|
| 332 |
+
|
| 333 |
+
except Exception as e_pdf_gen: # Captura general para la generación de PDF
|
| 334 |
tb_str = traceback.format_exc()
|
| 335 |
+
error_message = f"Error al generar PDF: {str(e_pdf_gen)}. Detalles: {tb_str}"
|
| 336 |
+
if len(error_message) > 600: error_message = error_message[:597] + "..." # Aumentar un poco el límite del mensaje de error
|
| 337 |
+
print(f"ERROR CRÍTICO en scrape_to_pdf: {error_message}")
|
| 338 |
return {'status': 'error', 'message': error_message, 'url': url}
|
| 339 |
|
| 340 |
+
# _clean_text_for_pdf ya no es un método de clase, es la función `clean_problematic_chars` al inicio del archivo.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|