Agile611/post2bsky
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

TikTok2

Browse Source
This commit is contained in:
Guillem Hernandez Sola
2026-05-19 20:25:14 +02:00
parent 08b30d73a9
commit 47dea4b110
1 changed files with 169 additions and 81 deletions
Download Patch File Download Diff File Expand all files Collapse all files

216
tiktok2bsky.py
View File

@@ -34,13 +34,21 @@ from atproto import Client
from dotenv import load_dotenv from dotenv import load_dotenv
from playwright.sync_api import sync_playwright from playwright.sync_api import sync_playwright
# playwright-stealth 1.x uses stealth_sync, 2.x uses Stealth class
# ─────────────────────────────────────────────────────────────────────────────
# playwright-stealth: support v1.x (stealth_sync) and v2.x (Stealth class)
# ─────────────────────────────────────────────────────────────────────────────
_STEALTH_V2 = None # None = not available at all
try: try:
from playwright_stealth import stealth_sync from playwright_stealth import stealth_sync
_STEALTH_V2 = False _STEALTH_V2 = False
except ImportError: except ImportError:
try:
from playwright_stealth import Stealth from playwright_stealth import Stealth
_STEALTH_V2 = True _STEALTH_V2 = True
except ImportError:
pass # stealth disabled — warning emitted at runtime
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
@@ -146,7 +154,6 @@ def load_state() -> dict:
def save_state(state: dict): def save_state(state: dict):
# Prune to last STATE_MAX_ENTRIES
posted = state.get("posted", {}) posted = state.get("posted", {})
if len(posted) > STATE_MAX_ENTRIES: if len(posted) > STATE_MAX_ENTRIES:
sorted_keys = sorted( sorted_keys = sorted(
@@ -180,7 +187,7 @@ def mark_as_posted(video_id: str, state: dict, meta: dict = None):
# Cookie helpers # Cookie helpers
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
def load_cookies_from_file(path: str) -> list: def load_cookies_from_file(path: str) -> list:
"""Load cookies from a JSON file (format produced by generate_tiktok_cookies.py).""" """Load cookies from a JSON file."""
if not os.path.exists(path): if not os.path.exists(path):
logging.warning(f"⚠️ Cookie file not found: {path}") logging.warning(f"⚠️ Cookie file not found: {path}")
return [] return []
@@ -279,7 +286,6 @@ def is_transient_error(error_obj) -> bool:
def get_rate_limit_wait_seconds(error_obj, default_delay: float) -> float: def get_rate_limit_wait_seconds(error_obj, default_delay: float) -> float:
""" """
Parse rate-limit response headers and return a bounded wait time in seconds. Parse rate-limit response headers and return a bounded wait time in seconds.
Supports retry-after, x-ratelimit-after, and ratelimit-reset (unix timestamp).
""" """
try: try:
now_ts = int(time.time()) now_ts = int(time.time())
@@ -300,7 +306,6 @@ def get_rate_limit_wait_seconds(error_obj, default_delay: float) -> float:
except Exception: except Exception:
pass pass
# repr() fallback — parse headers embedded in the exception string
text = repr(error_obj) text = repr(error_obj)
for pattern, is_timestamp in [ for pattern, is_timestamp in [
(r"'retry-after':\s*'(\d+)'", False), (r"'retry-after':\s*'(\d+)'", False),
@@ -318,6 +323,54 @@ def get_rate_limit_wait_seconds(error_obj, default_delay: float) -> float:
return default_delay return default_delay
# ─────────────────────────────────────────────────────────────────────────────
# playwright-stealth application helper
# ─────────────────────────────────────────────────────────────────────────────
def apply_stealth(page):
"""
Apply playwright-stealth to a page object.
Handles all known API variants:
v1.x → stealth_sync(page)
v2.x → Stealth().use_sync(page) returns a new wrapped page
v2.x → Stealth().use(page) alternate name
none → skip gracefully with a warning
Always returns a page object (wrapped or original).
"""
if _STEALTH_V2 is None:
logging.warning("⚠️ playwright-stealth not installed. Skipping stealth.")
return page
try:
if _STEALTH_V2:
# v2.x — probe for known method names
stealth = Stealth()
if hasattr(stealth, "use_sync"):
page = stealth.use_sync(page)
logging.info("🥷 playwright-stealth v2.x applied (use_sync).")
elif hasattr(stealth, "use"):
page = stealth.use(page)
logging.info("🥷 playwright-stealth v2.x applied (use).")
else:
logging.warning(
"⚠️ playwright-stealth v2.x: no known apply method found "
"(tried use_sync, use). Skipping stealth."
)
else:
# v1.x
stealth_sync(page)
logging.info("🥷 playwright-stealth v1.x applied (stealth_sync).")
except Exception as e:
logging.warning(
f"⚠️ playwright-stealth could not be applied: "
f"{type(e).__name__}: {e}. Continuing without stealth."
)
return page
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
# Bluesky client # Bluesky client
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
@@ -327,14 +380,17 @@ def connect_bluesky(handle: str, app_password: str, base_url: str) -> Client:
for attempt in range(1, BSKY_LOGIN_MAX_RETRIES + 1): for attempt in range(1, BSKY_LOGIN_MAX_RETRIES + 1):
try: try:
logging.info(f"🔐 Bluesky login attempt {attempt}/{BSKY_LOGIN_MAX_RETRIES} for {handle}") logging.info(
f"🔐 Bluesky login attempt {attempt}/{BSKY_LOGIN_MAX_RETRIES} for {handle}"
)
client.login(handle, app_password) client.login(handle, app_password)
client.me = client.get_profile(handle) client.me = client.get_profile(handle)
logging.info(f"✅ Bluesky login successful as {handle}") logging.info(f"✅ Bluesky login successful as {handle}")
return client return client
except Exception as e: except Exception as e:
logging.warning( logging.warning(
f"⚠️ Bluesky login {type(e).__name__}: {e} (attempt {attempt}/{BSKY_LOGIN_MAX_RETRIES})" f"⚠️ Bluesky login {type(e).__name__}: {e} "
f"(attempt {attempt}/{BSKY_LOGIN_MAX_RETRIES})"
) )
if is_rate_limited_error(e): if is_rate_limited_error(e):
delay = get_rate_limit_wait_seconds(e, BSKY_LOGIN_BASE_DELAY) delay = get_rate_limit_wait_seconds(e, BSKY_LOGIN_BASE_DELAY)
@@ -355,7 +411,9 @@ def connect_bluesky(handle: str, app_password: str, base_url: str) -> Client:
logging.warning(f"⏳ Retrying login in {wait:.1f}s.") logging.warning(f"⏳ Retrying login in {wait:.1f}s.")
time.sleep(wait) time.sleep(wait)
else: else:
logging.error(f"❌ Bluesky login failed after {BSKY_LOGIN_MAX_RETRIES} attempts.") logging.error(
f"❌ Bluesky login failed after {BSKY_LOGIN_MAX_RETRIES} attempts."
)
raise raise
raise RuntimeError("Bluesky login failed: exhausted all retries.") raise RuntimeError("Bluesky login failed: exhausted all retries.")
@@ -388,12 +446,12 @@ def compress_video(
input_path: str, input_path: str,
output_path: str, output_path: str,
max_duration: int = VIDEO_MAX_DURATION_S, max_duration: int = VIDEO_MAX_DURATION_S,
max_size_bytes: int = None, # resolved at call-time from get_video_size_limit() max_size_bytes: int = None,
) -> bool: ) -> bool:
""" """
Re-encode input_path → output_path using libx264, targeting max_size_bytes. Re-encode input_path → output_path using libx264, targeting max_size_bytes.
Key fixes applied: Fixes applied:
• pad=ceil(iw/2)*2:ceil(ih/2)*2 — ensures even dimensions (libx264 requirement) • pad=ceil(iw/2)*2:ceil(ih/2)*2 — ensures even dimensions (libx264 requirement)
• -maxrate == -b:v — hard ceiling, no burst above target • -maxrate == -b:v — hard ceiling, no burst above target
• post-encode size guard — rejects file if still over limit • post-encode size guard — rejects file if still over limit
@@ -471,12 +529,12 @@ def compress_video(
return True return True
except Exception as e: except Exception as e:
logging.error(f"❌ compress_video error: {e}") logging.error(f"❌ compress_video error: {type(e).__name__}: {e}")
return False return False
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
# yt-dlp download # yt-dlp helpers
# ───────────────────────────────────────────────────────────────────────────── # ─────────────────────────────────────────────────────────────────────────────
def get_best_impersonation_target() -> str | None: def get_best_impersonation_target() -> str | None:
""" """
@@ -491,7 +549,6 @@ def get_best_impersonation_target() -> str | None:
if target in available: if target in available:
logging.info(f"🎭 yt-dlp impersonation target: {target}") logging.info(f"🎭 yt-dlp impersonation target: {target}")
return target return target
# fallback: return first available
if available: if available:
target = sorted(available)[0] target = sorted(available)[0]
logging.info(f"🎭 yt-dlp impersonation target (fallback): {target}") logging.info(f"🎭 yt-dlp impersonation target (fallback): {target}")
@@ -544,10 +601,7 @@ def download_video_ytdlp(url: str, output_path: str, cookies_path: str = None) -
def download_video(url: str, output_path: str, cookies_path: str = None) -> bool: def download_video(url: str, output_path: str, cookies_path: str = None) -> bool:
""" """Download a TikTok video via yt-dlp with browser impersonation."""
Download a TikTok video. Routes directly to yt-dlp with browser impersonation.
"""
cookies = load_cookies_from_file(cookies_path) if cookies_path else []
logging.info(f"⬇️ Downloading: {url}") logging.info(f"⬇️ Downloading: {url}")
return download_video_ytdlp(url, output_path, cookies_path=cookies_path) return download_video_ytdlp(url, output_path, cookies_path=cookies_path)
@@ -563,8 +617,8 @@ def upload_video_to_bluesky(
""" """
Upload a video file to Bluesky as a blob. Upload a video file to Bluesky as a blob.
Fix 1 applied: exception is logged as type(e).__name__: e Fix 1: exception is always logged as type(e).__name__: e
so the actual error (413, 403, network error, etc.) is always visible. so the actual error (413, 403, network error, etc.) is visible in logs.
""" """
size_mb = os.path.getsize(video_path) / 1024 / 1024 size_mb = os.path.getsize(video_path) / 1024 / 1024
logging.info(f"⬆️ Uploading to Bluesky ({size_mb:.1f} MB)...") logging.info(f"⬆️ Uploading to Bluesky ({size_mb:.1f} MB)...")
@@ -581,12 +635,8 @@ def upload_video_to_bluesky(
return blob.blob return blob.blob
except Exception as e: except Exception as e:
# ── Fix 1: always log the full exception type and message ────── # Fix 1 — always log the full exception type and message
err_detail = f"{type(e).__name__}: {e}" err_detail = f"{type(e).__name__}: {e}"
logging.warning(
f"⚠️ Blob upload attempt {attempt}/{BSKY_UPLOAD_MAX_RETRIES} "
f"failed: {err_detail}. Retrying in {delay:.1f}s..."
)
if attempt >= BSKY_UPLOAD_MAX_RETRIES: if attempt >= BSKY_UPLOAD_MAX_RETRIES:
logging.error( logging.error(
@@ -595,6 +645,10 @@ def upload_video_to_bluesky(
) )
return None return None
logging.warning(
f"⚠️ Blob upload attempt {attempt}/{BSKY_UPLOAD_MAX_RETRIES} "
f"failed: {err_detail}. Retrying in {delay:.1f}s..."
)
time.sleep(delay + random.uniform(0, BSKY_UPLOAD_JITTER_MAX)) time.sleep(delay + random.uniform(0, BSKY_UPLOAD_JITTER_MAX))
delay = min(delay * 2, BSKY_UPLOAD_MAX_DELAY) delay = min(delay * 2, BSKY_UPLOAD_MAX_DELAY)
@@ -618,7 +672,6 @@ def post_video_to_bluesky(
video_embed = models.AppBskyEmbedVideo.Main( video_embed = models.AppBskyEmbedVideo.Main(
video=blob, video=blob,
) )
client.send_post( client.send_post(
text=caption, text=caption,
embed=video_embed, embed=video_embed,
@@ -690,14 +743,14 @@ def scrape_tiktok_profile_playwright(
page = context.new_page() page = context.new_page()
if _STEALTH_V2: # Apply stealth — gracefully handles all v1/v2/missing variants
Stealth().apply(page) page = apply_stealth(page)
else:
stealth_sync(page)
for attempt in range(1, PLAYWRIGHT_MAX_RELOADS + 1): for attempt in range(1, PLAYWRIGHT_MAX_RELOADS + 1):
try: try:
logging.info(f"🌐 Loading profile (attempt {attempt}/{PLAYWRIGHT_MAX_RELOADS})...") logging.info(
f"🌐 Loading profile (attempt {attempt}/{PLAYWRIGHT_MAX_RELOADS})..."
)
page.goto( page.goto(
profile_url, profile_url,
wait_until="domcontentloaded", wait_until="domcontentloaded",
@@ -720,7 +773,9 @@ def scrape_tiktok_profile_playwright(
logging.warning(f"⚠️ Video grid not found on attempt {attempt}.") logging.warning(f"⚠️ Video grid not found on attempt {attempt}.")
ts = int(time.time()) ts = int(time.time())
page.screenshot(path=f"screenshot_no_grid_{attempt}_{ts}.png") page.screenshot(path=f"screenshot_no_grid_{attempt}_{ts}.png")
logging.info(f"📸 Screenshot saved: screenshot_no_grid_{attempt}_{ts}.png") logging.info(
f"📸 Screenshot saved: screenshot_no_grid_{attempt}_{ts}.png"
)
time.sleep(3) time.sleep(3)
continue continue
@@ -750,7 +805,10 @@ def scrape_tiktok_profile_playwright(
break break
except Exception as e: except Exception as e:
logging.warning(f"⚠️ Playwright attempt {attempt} error: {type(e).__name__}: {e}") logging.warning(
f"⚠️ Playwright attempt {attempt} error: "
f"{type(e).__name__}: {e}"
)
ts = int(time.time()) ts = int(time.time())
try: try:
page.screenshot(path=f"screenshot_error_{attempt}_{ts}.png") page.screenshot(path=f"screenshot_error_{attempt}_{ts}.png")
@@ -759,17 +817,31 @@ def scrape_tiktok_profile_playwright(
time.sleep(3) time.sleep(3)
if not videos: if not videos:
logging.warning("⚠️ Video grid not found on attempt 3.") logging.warning(
f"⚠️ Video grid not found on attempt {PLAYWRIGHT_MAX_RELOADS}."
)
ts = int(time.time()) ts = int(time.time())
try: try:
page.screenshot(path=f"screenshot_no_grid_3_{ts}.png") page.screenshot(path=f"screenshot_no_grid_{PLAYWRIGHT_MAX_RELOADS}_{ts}.png")
logging.info(f"📸 Screenshot saved: screenshot_no_grid_3_{ts}.png") logging.info(
f"📸 Screenshot saved: "
f"screenshot_no_grid_{PLAYWRIGHT_MAX_RELOADS}_{ts}.png"
)
except Exception: except Exception:
pass pass
try:
page.close() page.close()
except Exception:
pass
try:
context.close() context.close()
except Exception:
pass
try:
browser.close() browser.close()
except Exception:
pass
return videos return videos
@@ -837,7 +909,9 @@ def scrape_tiktok_profile_ytdlp(
return videos[:limit] return videos[:limit]
except Exception as e: except Exception as e:
logging.error(f"❌ yt-dlp profile scrape failed: {type(e).__name__}: {e}") logging.error(
f"❌ yt-dlp profile scrape failed: {type(e).__name__}: {e}"
)
return [] return []
@@ -850,7 +924,6 @@ def build_caption(video_info: dict, tiktok_handle: str, max_len: int = 290) -> s
url = video_info.get("url", "") url = video_info.get("url", "")
if desc: if desc:
# Truncate description to leave room for the URL
url_len = len(url) + 1 # +1 for newline url_len = len(url) + 1 # +1 for newline
max_desc = max_len - url_len max_desc = max_len - url_len
if len(desc) > max_desc: if len(desc) > max_desc:
@@ -888,7 +961,7 @@ def process_videos(
logging.info(f"⏭️ Already posted: {video_id}") logging.info(f"⏭️ Already posted: {video_id}")
continue continue
# Age filter (only if timestamp is available) # Age filter (only when timestamp is available)
ts = video.get("timestamp") ts = video.get("timestamp")
if ts: if ts:
try: try:
@@ -896,7 +969,8 @@ def process_videos(
age_days = (now - video_time).days age_days = (now - video_time).days
if age_days > max_age_days: if age_days > max_age_days:
logging.info( logging.info(
f"⏭️ Video {video_id} too old ({age_days}d > {max_age_days}d). Skipping." f"⏭️ Video {video_id} too old "
f"({age_days}d > {max_age_days}d). Skipping."
) )
continue continue
except Exception: except Exception:
@@ -904,9 +978,6 @@ def process_videos(
logging.info(f"🎬 Processing video {video_id}: {video_url}") logging.info(f"🎬 Processing video {video_id}: {video_url}")
# Re-load cookies for each video (in case file was refreshed)
load_cookies_from_file(cookies_path)
with tempfile.TemporaryDirectory() as tmpdir: with tempfile.TemporaryDirectory() as tmpdir:
raw_path = os.path.join(tmpdir, f"{video_id}_raw.mp4") raw_path = os.path.join(tmpdir, f"{video_id}_raw.mp4")
comp_path = os.path.join(tmpdir, f"{video_id}.mp4") comp_path = os.path.join(tmpdir, f"{video_id}.mp4")
@@ -952,17 +1023,43 @@ def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser( parser = argparse.ArgumentParser(
description="Cross-post TikTok videos to Bluesky." description="Cross-post TikTok videos to Bluesky."
) )
parser.add_argument("--tiktok-handle", required=True, help="TikTok username (without @)") parser.add_argument(
parser.add_argument("--bsky-handle", required=True, help="Bluesky handle") "--tiktok-handle",
parser.add_argument("--bsky-app-password", required=True, help="Bluesky app password") required=True,
parser.add_argument("--bsky-base-url", default=DEFAULT_BSKY_BASE_URL, help="TikTok username (without @)",
help=f"Bluesky PDS base URL (default: {DEFAULT_BSKY_BASE_URL})") )
parser.add_argument("--bsky-langs", nargs="+", default=DEFAULT_BSKY_LANGS, parser.add_argument(
help="BCP-47 language tags for posts (default: es)") "--bsky-handle",
parser.add_argument("--cookies-path", default=TIKTOK_COOKIES_PATH, required=True,
help=f"Path to TikTok cookies JSON (default: {TIKTOK_COOKIES_PATH})") help="Bluesky handle",
parser.add_argument("--max-age-days", type=int, default=VIDEO_MAX_AGE_DAYS, )
help=f"Skip videos older than N days (default: {VIDEO_MAX_AGE_DAYS})") parser.add_argument(
"--bsky-app-password",
required=True,
help="Bluesky app password",
)
parser.add_argument(
"--bsky-base-url",
default=DEFAULT_BSKY_BASE_URL,
help=f"Bluesky PDS base URL (default: {DEFAULT_BSKY_BASE_URL})",
)
parser.add_argument(
"--bsky-langs",
nargs="+",
default=DEFAULT_BSKY_LANGS,
help="BCP-47 language tags for posts (default: es)",
)
parser.add_argument(
"--cookies-path",
default=TIKTOK_COOKIES_PATH,
help=f"Path to TikTok cookies JSON (default: {TIKTOK_COOKIES_PATH})",
)
parser.add_argument(
"--max-age-days",
type=int,
default=VIDEO_MAX_AGE_DAYS,
help=f"Skip videos older than N days (default: {VIDEO_MAX_AGE_DAYS})",
)
return parser.parse_args() return parser.parse_args()
@@ -970,7 +1067,7 @@ def main():
load_dotenv() load_dotenv()
args = parse_args() args = parse_args()
# ── Fix 2: resolve video size limit based on PDS ────────────────────── # Fix 2 — resolve video size limit based on PDS
video_max_size_bytes = get_video_size_limit(args.bsky_base_url) video_max_size_bytes = get_video_size_limit(args.bsky_base_url)
logging.info("=" * 60) logging.info("=" * 60)
@@ -1004,17 +1101,8 @@ def main():
) )
if not videos: if not videos:
logging.warning("⚠️ Playwright grid scraping failed. Trying API fallback...") logging.warning("⚠️ Playwright grid scraping failed. Trying yt-dlp fallback...")
ts = int(time.time()) ts = int(time.time())
# Try to save a screenshot if playwright left a page open
try:
import glob
for f in glob.glob("screenshot_no_grid_*.png"):
pass # already saved inside scrape function
except Exception:
pass
# Save a "playwright failed" screenshot placeholder in logs
logging.info(f"📸 Screenshot saved: screenshot_playwright_failed_{ts}.png") logging.info(f"📸 Screenshot saved: screenshot_playwright_failed_{ts}.png")
videos = scrape_tiktok_profile_ytdlp( videos = scrape_tiktok_profile_ytdlp(