goyimnose/Medios-Macina
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

jkj

Browse Source
This commit is contained in:
nose
2025-12-16 23:23:43 -08:00
parent 9873280f0e
commit 86918f2ae2
46 changed files with 2277 additions and 1347 deletions
Download Patch File Download Diff File Expand all files Collapse all files

317
Provider/libgen.py
View File

@@ -15,11 +15,11 @@ from SYS.logger import log
from models import ProgressBar
# Optional dependencies
# Optional dependency for HTML scraping fallbacks
try:
from bs4 import BeautifulSoup
from lxml import html as lxml_html
except ImportError:
BeautifulSoup = None
lxml_html = None
class Libgen(SearchProvider):
@@ -116,7 +116,7 @@ class Libgen(SearchProvider):
return []
def validate(self) -> bool:
# JSON-based searching can work without BeautifulSoup; HTML parsing is a fallback.
# JSON-based searching can work without lxml; HTML parsing is a fallback.
return True
def download(self, result: SearchResult, output_dir: Path) -> Optional[Path]:
@@ -342,8 +342,8 @@ class LibgenSearch:
Uses a total time budget across mirrors to avoid long hangs.
"""
# Prefer JSON API (no BeautifulSoup needed); HTML scraping is a fallback.
has_bs4 = BeautifulSoup is not None
# Prefer JSON API (no lxml needed); HTML scraping is a fallback.
has_lxml = lxml_html is not None
started = time.monotonic()
@@ -372,7 +372,7 @@ class LibgenSearch:
results = []
if not results:
if not has_bs4:
if not has_lxml:
continue
if "libgen.li" in mirror or "libgen.gl" in mirror:
@@ -417,57 +417,73 @@ class LibgenSearch:
resp = self.session.get(url, params=params, timeout=timeout)
resp.raise_for_status()
if BeautifulSoup is None:
if lxml_html is None:
return []
soup = BeautifulSoup(resp.text, "html.parser")
table = soup.find("table", {"class": "c"})
if not table:
tables = soup.find_all("table")
for t in tables:
if len(t.find_all("tr")) > 5:
def _text(el: Any) -> str:
return " ".join([t.strip() for t in el.itertext() if t and str(t).strip()]).strip()
try:
doc = lxml_html.fromstring(resp.content)
except Exception:
return []
table_nodes = doc.xpath(
"//table[contains(concat(' ', normalize-space(@class), ' '), ' c ')]"
)
table = table_nodes[0] if table_nodes else None
if table is None:
for t in doc.xpath("//table"):
if len(t.xpath(".//tr")) > 5:
table = t
break
if not table:
if table is None:
return []
results: List[Dict[str, Any]] = []
rows = table.find_all("tr")[1:]
rows = table.xpath(".//tr")[1:]
for row in rows:
cols = row.find_all("td")
cols = row.xpath("./td")
if len(cols) < 9:
continue
try:
libgen_id = cols[0].get_text(strip=True)
authors = [a.get_text(strip=True) for a in cols[1].find_all("a")]
if not authors:
authors = [cols[1].get_text(strip=True)]
libgen_id = _text(cols[0])
title_tag = cols[2].find("a")
title = title_tag.get_text(strip=True) if title_tag else cols[2].get_text(strip=True)
author_links = cols[1].xpath(".//a")
authors = [_text(a) for a in author_links if _text(a)]
if not authors:
authors = [_text(cols[1])]
title_tag = None
title_links = cols[2].xpath(".//a")
if title_links:
title_tag = title_links[0]
title = _text(title_tag) if title_tag is not None else _text(cols[2])
md5 = ""
if title_tag and title_tag.has_attr("href"):
if title_tag is not None:
href = str(title_tag.get("href") or "")
match = re.search(r"md5=([a-fA-F0-9]{32})", href)
if match:
md5 = match.group(1)
publisher = cols[3].get_text(strip=True)
year = cols[4].get_text(strip=True)
pages = cols[5].get_text(strip=True)
language = cols[6].get_text(strip=True)
size = cols[7].get_text(strip=True)
extension = cols[8].get_text(strip=True)
publisher = _text(cols[3])
year = _text(cols[4])
pages = _text(cols[5])
language = _text(cols[6])
size = _text(cols[7])
extension = _text(cols[8])
mirror_links = []
mirror_links: List[str] = []
for i in range(9, len(cols)):
a = cols[i].find("a")
if a and a.has_attr("href"):
mirror_links.append(a["href"])
a_nodes = cols[i].xpath(".//a[@href]")
if a_nodes:
href = str(a_nodes[0].get("href") or "").strip()
if href:
mirror_links.append(href)
if md5:
download_link = f"http://library.lol/main/{md5}"
@@ -476,24 +492,25 @@ class LibgenSearch:
else:
download_link = ""
results.append({
"id": libgen_id,
"title": title,
"author": ", ".join(authors),
"publisher": publisher,
"year": year,
"pages": pages,
"language": language,
"filesize_str": size,
"extension": extension,
"md5": md5,
"mirror_url": download_link,
"cover": "",
})
results.append(
{
"id": libgen_id,
"title": title,
"author": ", ".join([a for a in authors if a]) or "Unknown",
"publisher": publisher,
"year": year,
"pages": pages,
"language": language,
"filesize_str": size,
"extension": extension,
"md5": md5,
"mirror_url": download_link,
"cover": "",
}
)
if len(results) >= limit:
break
except Exception as e:
logging.debug(f"Error parsing row: {e}")
continue
@@ -521,21 +538,35 @@ class LibgenSearch:
resp = self.session.get(url, params=params, timeout=timeout)
resp.raise_for_status()
if BeautifulSoup is None:
if lxml_html is None:
return []
soup = BeautifulSoup(resp.text, "html.parser")
table = soup.find("table", {"id": "tablelibgen"})
if not table:
table = soup.find("table", {"class": "table table-striped"})
if not table:
def _text(el: Any) -> str:
return " ".join([t.strip() for t in el.itertext() if t and str(t).strip()]).strip()
try:
doc = lxml_html.fromstring(resp.content)
except Exception:
return []
table_nodes = doc.xpath("//table[@id='tablelibgen']")
table = table_nodes[0] if table_nodes else None
if table is None:
# Common libgen.li/gl fallback
table_nodes = doc.xpath(
"//table[contains(concat(' ', normalize-space(@class), ' '), ' table ') and "
"contains(concat(' ', normalize-space(@class), ' '), ' table-striped ')]"
)
table = table_nodes[0] if table_nodes else None
if table is None:
return []
results: List[Dict[str, Any]] = []
rows = table.find_all("tr")[1:]
rows = table.xpath(".//tr")[1:]
for row in rows:
cols = row.find_all("td")
cols = row.xpath("./td")
if len(cols) < 9:
continue
@@ -543,26 +574,30 @@ class LibgenSearch:
# Extract md5 (libgen.gl exposes /ads.php?md5=... in mirror column)
md5 = ""
mirror_url = ""
for a in row.find_all("a"):
href = a.get("href")
for a in row.xpath(".//a[@href]"):
href = str(a.get("href") or "")
if not href:
continue
m = re.search(r"md5=([a-fA-F0-9]{32})", str(href))
m = re.search(r"md5=([a-fA-F0-9]{32})", href)
if m:
md5 = m.group(1)
if "ads.php" in str(href):
mirror_url = urljoin(mirror, str(href))
if "ads.php" in href:
mirror_url = urljoin(mirror, href)
break
if not mirror_url and md5:
mirror_url = urljoin(mirror, f"/ads.php?md5={md5}")
# Extract numeric file id from /file.php?id=...
libgen_id = ""
file_link = row.find("a", href=re.compile(r"/file\.php\?id=\d+"))
if file_link and file_link.get("href"):
m = re.search(r"id=(\d+)", str(file_link.get("href")))
if m:
libgen_id = m.group(1)
for a in row.xpath(".//a[@href]"):
href = str(a.get("href") or "")
if not href:
continue
if re.search(r"/file\.php\?id=\d+", href):
m = re.search(r"id=(\d+)", href)
if m:
libgen_id = m.group(1)
break
title = ""
authors = ""
@@ -585,7 +620,7 @@ class LibgenSearch:
if offset is not None:
meta_cell = cols[offset]
meta_text = " ".join([str(s).strip() for s in meta_cell.stripped_strings if str(s).strip()])
meta_text = _text(meta_cell)
# Extract ISBNs from meta cell (avoid using them as title)
# Matches 10 or 13-digit ISBN with optional leading 978/979.
@@ -601,11 +636,11 @@ class LibgenSearch:
# Choose a "real" title from meta cell.
# libgen.gl meta can include series/edition/isbn blobs; prefer text with letters.
raw_candidates: List[str] = []
for a in meta_cell.find_all("a"):
t = a.get_text(" ", strip=True)
for a in meta_cell.xpath(".//a"):
t = _text(a)
if t:
raw_candidates.append(t)
for s in meta_cell.stripped_strings:
for s in meta_cell.itertext():
t = str(s).strip()
if t:
raw_candidates.append(t)
@@ -645,27 +680,27 @@ class LibgenSearch:
best_score = score
best_title = cand
title = best_title or meta_cell.get_text(" ", strip=True)
title = best_title or _text(meta_cell)
authors = cols[offset + 1].get_text(" ", strip=True)
publisher = cols[offset + 2].get_text(" ", strip=True)
year = cols[offset + 3].get_text(" ", strip=True)
language = cols[offset + 4].get_text(" ", strip=True)
pages = cols[offset + 5].get_text(" ", strip=True)
size = cols[offset + 6].get_text(" ", strip=True)
extension = cols[offset + 7].get_text(" ", strip=True)
authors = _text(cols[offset + 1])
publisher = _text(cols[offset + 2])
year = _text(cols[offset + 3])
language = _text(cols[offset + 4])
pages = _text(cols[offset + 5])
size = _text(cols[offset + 6])
extension = _text(cols[offset + 7])
else:
# Older fallback structure
title_col = cols[1]
title_link = title_col.find("a")
title = title_link.get_text(" ", strip=True) if title_link else title_col.get_text(" ", strip=True)
authors = cols[2].get_text(" ", strip=True)
publisher = cols[3].get_text(" ", strip=True)
year = cols[4].get_text(" ", strip=True)
language = cols[5].get_text(" ", strip=True)
pages = cols[6].get_text(" ", strip=True)
size = cols[7].get_text(" ", strip=True)
extension = cols[8].get_text(" ", strip=True)
title_links = title_col.xpath(".//a")
title = _text(title_links[0]) if title_links else _text(title_col)
authors = _text(cols[2])
publisher = _text(cols[3])
year = _text(cols[4])
language = _text(cols[5])
pages = _text(cols[6])
size = _text(cols[7])
extension = _text(cols[8])
title = (title or "").strip() or "Unknown"
authors = (authors or "").strip() or "Unknown"
@@ -729,15 +764,49 @@ def _resolve_download_url(
current_url = url
visited = set()
if BeautifulSoup is None:
_call(log_info, "[resolve] BeautifulSoup not available; cannot resolve HTML download chain")
def _resolve_html_links_regex(base_url: str, html: str) -> Optional[str]:
"""Best-effort HTML link resolver without lxml.
This is intentionally minimal: it primarily targets LibGen landing pages like
`/ads.php?md5=...` which contain a `get.php?md5=...` link.
"""
if not html:
return None
# Prefer explicit get.php md5 links (most common successful chain).
m = re.search(r'href=["\']([^"\']*get\.php\?md5=[a-fA-F0-9]{32}[^"\']*)["\']', html, flags=re.IGNORECASE)
if m:
href = str(m.group(1) or "").strip()
if href and not href.lower().startswith("javascript:"):
return urljoin(base_url, href)
# Next: library.lol main links.
m = re.search(r'href=["\']([^"\']*library\.lol[^"\']*)["\']', html, flags=re.IGNORECASE)
if m:
href = str(m.group(1) or "").strip()
if href and not href.lower().startswith("javascript:"):
return urljoin(base_url, href)
# Finally: any direct file extension link.
m = re.search(
r'href=["\']([^"\']+\.(?:pdf|epub|mobi|djvu|azw3|cbz|cbr)(?:\?[^"\']*)?)["\']',
html,
flags=re.IGNORECASE,
)
if m:
href = str(m.group(1) or "").strip()
if href and not href.lower().startswith("javascript:"):
return urljoin(base_url, href)
return None
def _find_a_by_text(pattern: str) -> Optional[Any]:
for a in soup.find_all("a"):
t = a.get_text(" ", strip=True)
def _find_href_by_text(doc: Any, pattern: str) -> Optional[str]:
for a in doc.xpath("//a[@href]"):
t = " ".join([s.strip() for s in a.itertext() if s and str(s).strip()]).strip()
if t and re.search(pattern, t, re.IGNORECASE):
return a
href = str(a.get("href") or "").strip()
if href and not href.lower().startswith("javascript:"):
return href
return None
for _ in range(6):
@@ -763,42 +832,58 @@ def _resolve_download_url(
_call(log_info, f"[resolve] Failed to fetch {current_url}: {e}")
return None
soup = BeautifulSoup(content, "html.parser")
doc = None
if lxml_html is not None:
try:
doc = lxml_html.fromstring(content)
except Exception:
doc = None
get_link = _find_a_by_text(r"^GET$")
if get_link and get_link.has_attr("href"):
return urljoin(current_url, str(get_link.get("href") or ""))
if doc is None:
next_url = _resolve_html_links_regex(current_url, content)
if next_url:
current_url = next_url
continue
_call(log_info, "[resolve] lxml not available and regex resolver found no links")
return None
get_href = _find_href_by_text(doc, r"^GET$")
if get_href:
return urljoin(current_url, get_href)
if "series.php" in current_url:
edition_link = soup.find("a", href=re.compile(r"edition\.php"))
if edition_link:
current_url = urljoin(current_url, str(edition_link.get("href") or ""))
hrefs = doc.xpath("//a[contains(@href,'edition.php')]/@href")
if hrefs:
current_url = urljoin(current_url, str(hrefs[0] or ""))
continue
if "edition.php" in current_url:
file_link = soup.find("a", href=re.compile(r"file\.php"))
if file_link:
current_url = urljoin(current_url, str(file_link.get("href") or ""))
hrefs = doc.xpath("//a[contains(@href,'file.php')]/@href")
if hrefs:
current_url = urljoin(current_url, str(hrefs[0] or ""))
continue
if "file.php" in current_url:
libgen_link = soup.find("a", title="libgen")
if not libgen_link:
libgen_link = _find_a_by_text(r"Libgen")
if libgen_link and libgen_link.has_attr("href"):
current_url = urljoin(current_url, str(libgen_link.get("href") or ""))
libgen_href = None
for a in doc.xpath("//a[@href]"):
if str(a.get("title") or "").strip().lower() == "libgen":
libgen_href = str(a.get("href") or "").strip()
break
if not libgen_href:
libgen_href = _find_href_by_text(doc, r"Libgen")
if libgen_href:
current_url = urljoin(current_url, libgen_href)
continue
if "ads.php" in current_url:
get_php_link = soup.find("a", href=re.compile(r"get\.php"))
if get_php_link:
return urljoin(current_url, str(get_php_link.get("href") or ""))
hrefs = doc.xpath("//a[contains(@href,'get.php')]/@href")
if hrefs:
return urljoin(current_url, str(hrefs[0] or ""))
for text in ["Cloudflare", "IPFS.io", "Infura"]:
link = _find_a_by_text(re.escape(text))
if link and link.has_attr("href"):
return urljoin(current_url, str(link.get("href") or ""))
href = _find_href_by_text(doc, re.escape(text))
if href:
return urljoin(current_url, href)
break

86
Provider/openlibrary.py
View File

@@ -1,6 +1,7 @@
from __future__ import annotations
import base64
import io
from concurrent import futures
import hashlib
import json as json_module
@@ -34,6 +35,53 @@ except ImportError:
tqdm = None # type: ignore
def _image_paths_to_pdf_bytes(images: List[str]) -> Optional[bytes]:
if not images:
return None
try:
from PIL import Image # type: ignore
except Exception:
return None
pil_images: List[Any] = []
try:
for p in images:
img_path = Path(p)
if not img_path.is_file():
continue
with Image.open(img_path) as im: # type: ignore[attr-defined]
# Ensure PDF-compatible mode.
if im.mode in {"RGBA", "LA", "P"}:
im = im.convert("RGB")
else:
im = im.convert("RGB")
pil_images.append(im.copy())
except Exception:
for im in pil_images:
try:
im.close()
except Exception:
pass
return None
if not pil_images:
return None
buf = io.BytesIO()
first, rest = pil_images[0], pil_images[1:]
try:
first.save(buf, format="PDF", save_all=True, append_images=rest)
return buf.getvalue()
except Exception:
return None
finally:
for im in pil_images:
try:
im.close()
except Exception:
pass
def _looks_like_isbn(text: str) -> bool:
t = (text or "").replace("-", "").strip()
return t.isdigit() and len(t) in (10, 13)
@@ -941,32 +989,22 @@ class OpenLibrary(SearchProvider):
try:
images = self._archive_download(session=session, n_threads=10, directory=temp_dir, links=links, scale=3, book_id=archive_id)
try:
import img2pdf # type: ignore
pdf_bytes = img2pdf.convert(images) if images else None
if not pdf_bytes:
log("[openlibrary] PDF conversion failed", file=sys.stderr)
try:
shutil.rmtree(temp_dir)
except Exception:
pass
return None
pdf_path = unique_path(output_dir / f"{title}.pdf")
with open(pdf_path, "wb") as f:
f.write(pdf_bytes)
try:
shutil.rmtree(temp_dir)
except Exception:
pass
return pdf_path
except ImportError:
# Keep images folder.
pdf_bytes = _image_paths_to_pdf_bytes(images)
if not pdf_bytes:
# Keep images folder for manual conversion.
log("[openlibrary] PDF conversion failed; keeping images folder", file=sys.stderr)
return Path(temp_dir)
pdf_path = unique_path(output_dir / f"{title}.pdf")
with open(pdf_path, "wb") as f:
f.write(pdf_bytes)
try:
shutil.rmtree(temp_dir)
except Exception:
pass
return pdf_path
except Exception:
try:
shutil.rmtree(temp_dir)