Lxml html clean fix

python3 pipeline.py Traceback (most recent call last): File "/Users/danielkliewer/obj01/pipeline.py", line 4, in <module> from modules.scraping import fetch_articles File "/Users/danielkliewer/obj01/modules/scraping.py", line 3, in <module> from newspaper import Article File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/__i nit__.py", line 10, in <module> from .api import (build, build_article, fulltext, hot, languages, popular_urls, Configuration as Config) File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/api .py", line 14, in <module> from .article import Article File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/art icle.py", line 15, in <module> from . import network File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/net work.py", line 14, in <module> from .configuration import Configuration File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/con figuration.py", line 15, in <module> from .parsers import Parser File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/newspaper/par sers.py", line 12, in <module> import lxml.html.clean File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/lxml/html/cle an.py", line 18, in <module> raise ImportError( ...<2 lines>... ) from None ImportError: lxml.html.clean module is now a separate project lxml_html_clean. Install lxml[html_clean] or lxml_html_clean directly.

The error you're seeing is due to a **breaking change in the `lxml` library**. The `lxml.html.clean` module was **split into a separate package** called [`lxml_html_clean`](https://pypi.org/project/lxml-html-clean/) starting in recent versions. ### 🔧 Fix You have two options: --- ### ✅ Option 1: Install the extra dependency with `lxml` Run this: ```bash pip install "lxml[html_clean]" ``` This installs the `lxml_html_clean` package automatically as a dependency. --- ### ✅ Option 2: Install `lxml_html_clean` directly If that fails, you can explicitly install the new package: ```bash pip install lxml-html-clean ``` --- ### ✅ Optional: Pin compatible versions (for long-term stability) If you're deploying or working in a shared environment, consider pinning versions: ```txt lxml==4.9.3 # or whatever works with your code lxml-html-clean==0.1.1 ``` Then install with: ```bash pip install -r requirements.txt ``` --- Let me know if you want help downgrading or patching the `newspaper3k` code to not require cleaning, though it’s easiest to just install the missing dependency.

python3 pipeline.py Traceback (most recent call last): File "/Users/danielkliewer/obj01/pipeline.py", line 5, in <module> from modules.translation import translate_article File "/Users/danielkliewer/obj01/modules/translation.py", line 4, in <module> package.install_from_path('https://argosopentech.nyc3.digitaloceanspaces.com/a rgos-translate/models/translate-fr_en-1_0.argosmodel') ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/argostranslat e/package.py", line 249, in install_from_path raise Exception("Not a valid Argos Model (must be a zip archive)") Exception: Not a valid Argos Model (must be a zip archive)

The error: ``` Exception: Not a valid Argos Model (must be a zip archive) ``` means that `argos-translate` is trying to install a model from a URL, but the file it's downloading is **not a valid `.argosmodel` ZIP archive**, even though it has the `.argosmodel` extension. ### ✅ Fixes Here’s how to resolve this: --- ### 🔍 Step 1: Check if the model file URL is valid Test the URL manually: 1. Open this in a browser: ``` https://argosopentech.nyc3.digitaloceanspaces.com/argos-translate/models/translate-fr_en-1_0.argosmodel ``` 2. If it **404s or downloads a broken file**, that’s the issue. --- ### ✅ Step 2: Download manually, then install locally Instead of installing from a remote URL, try this: ```bash # Download manually curl -O https://argosopentech.nyc3.digitaloceanspaces.com/argos-translate/models/translate-fr_en-1_0.argosmodel # Then install locally python3 >>> from argostranslate import package >>> package.install_from_path('translate-fr_en-1_0.argosmodel') ``` Make sure the file is downloaded completely and is **not HTML or a redirect** (check with `file translate-fr_en_1_0.argosmodel`). --- ### 🛠 Alternative: Use `argos-translate-cli` to install models You can list and install models via the CLI: ```bash pip install argos-translate # List available models argos-translate-cli --list-models # Install the French to English model argos-translate-cli --install-model translate fr en ``` --- ### ✅ Bonus: Keep a backup of the working `.argosmodel` file If the URL goes down again in the future (which happens), it’s smart to cache the `.argosmodel` files locally in your project repo or object storage. --- Let me know if you'd like a script that verifies and installs models automatically, or checks for corrupt downloads.

python3 pipeline.py 🔧 Installing translation models... Traceback (most recent call last): File "/Users/danielkliewer/obj01/pipeline.py", line 5, in <module> from modules.translation import translate_article File "/Users/danielkliewer/obj01/modules/translation.py", line 48, in <module> initialize_translator() ~~~~~~~~~~~~~~~~~~~~~^^ File "/Users/danielkliewer/obj01/modules/translation.py", line 43, in initialize _translator package.install_from_path(str(model_file)) ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^ File "/Users/danielkliewer/obj01/venv/lib/python3.13/site-packages/argostranslat e/package.py", line 249, in install_from_path raise Exception("Not a valid Argos Model (must be a zip archive)") Exception: Not a valid Argos Model (must be a zip archive)

Thanks for the update — you're still getting the `Not a valid Argos Model (must be a zip archive)` error, but now it's happening even though you're installing from a **local file (`model_file`)**. This means the file exists, but it’s **not a valid `.argosmodel` zip archive**, possibly due to one of the following: --- ### 🚩 Common Causes 1. **Partial or corrupted download** Your code likely tried to download the `.argosmodel` file, but saved an HTML error page instead of the model. 2. **Wrong content-type or redirect URL** Some hosting services (like DigitalOcean Spaces or GitHub) serve a redirect or HTML page if headers aren’t passed properly during download. --- ### ✅ Step-by-Step Fix #### 1. Check the downloaded file Go to the directory where the `.argosmodel` file was downloaded and run: ```bash file translate-fr_en-1_0.argosmodel ``` If it says something like `HTML document` or `ASCII text`, it's not valid — it's probably an error page. --- #### 2. Redownload the model manually Run: ```bash curl -L -o translate-fr_en_1_0.argosmodel \ https://argosopentech.nyc3.digitaloceanspaces.com/argos-translate/models/translate-fr_en-1_0.argosmodel ``` Then verify: ```bash unzip -l translate-fr_en_1_0.argosmodel ``` It should show files inside the zip archive (like `metadata.argosmodel`, etc.). If it doesn’t unzip, it's corrupt. --- #### 3. Manually install the model In Python: ```python from argostranslate import package package.install_from_path('translate-fr_en_1_0.argosmodel') ``` --- ### 🧰 Optional: Safer Auto-Download Script Here’s a snippet you can drop into `translation.py` to verify the file before installing: ```python import requests from pathlib import Path from argostranslate import package def download_model_safe(url: str, dest: Path): if not dest.exists() or dest.stat().st_size < 1000: print(f"📥 Downloading model from {url}") r = requests.get(url, allow_redirects=True) if r.status_code == 200 and r.headers.get('Content-Type') == 'application/zip': dest.write_bytes(r.content) else: raise Exception(f"❌ Failed to download model: {r.status_code}, content-type: {r.headers.get('Content-Type')}") else: print("✅ Model already exists and looks valid.") # Usage model_url = "https://argosopentech.nyc3.digitaloceanspaces.com/argos-translate/models/translate-fr_en-1_0.argosmodel" model_path = Path("translate-fr_en_1_0.argosmodel") download_model_safe(model_url, model_path) package.install_from_path(str(model_path)) ``` --- Let me know if you want help writing a fallback to install models interactively via CLI if the download fails.