google-news-trends-mcp 0.1.7__tar.gz → 0.1.9__tar.gz

{google_news_trends_mcp-0.1.7/src/google_news_trends_mcp.egg-info → google_news_trends_mcp-0.1.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: google-news-trends-mcp
-Version: 0.1.7
+Version: 0.1.9
 Summary: An MCP server to access Google News and Google Trends.
 Author-email: Jesse Manek <jesse.manek@gmail.com>
 License-Expression: MIT
@@ -35,8 +35,8 @@ An MCP server to access Google News and Google Trends.  Does not rely on any pai
 
 - Search Google News articles based on keyword, location, topic
 - Get top news stories from Google News
-- Google Trends keywords base on location
-- Optional NLP to summarize articles and extract keywords
+- Google Trends keywords based on location
+- Optional LLM Sampling and NLP to summarize articles and extract keywords
 
 ## Installation
 
@@ -70,7 +70,7 @@ Add to your Claude settings:
   "mcpServers": {
     "google-news-trends": {
       "command": "uvx",
-      "args": ["google-news-trends-mcp"]
+      "args": ["google-news-trends-mcp@latest"]
     }
   }
 }
@@ -103,7 +103,7 @@ Add to your Claude settings:
     "servers": {
       "google-news-trends": {
         "command": "uvx",
-        "args": ["google-news-trends-mcp"]
+        "args": ["google-news-trends-mcp@latest"]
       }
     }
   }
@@ -141,7 +141,7 @@ The following MCP tools are available:
 | **get_top_news**         | Fetch the top news stories from Google News.                       |
 | **get_trending_keywords**| Return trending keywords from Google Trends for a specified location.|
 
-All of the news related tools have an option to summarize the text of the article using NLP (nltk)
+All of the news related tools have an option to summarize the text of the article using LLM Sampling (if supported) or NLP
 
 
 ## CLI

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9}/README.md

@@ -6,8 +6,8 @@ An MCP server to access Google News and Google Trends.  Does not rely on any pai
 
 - Search Google News articles based on keyword, location, topic
 - Get top news stories from Google News
-- Google Trends keywords base on location
-- Optional NLP to summarize articles and extract keywords
+- Google Trends keywords based on location
+- Optional LLM Sampling and NLP to summarize articles and extract keywords
 
 ## Installation
 
@@ -41,7 +41,7 @@ Add to your Claude settings:
   "mcpServers": {
     "google-news-trends": {
       "command": "uvx",
-      "args": ["google-news-trends-mcp"]
+      "args": ["google-news-trends-mcp@latest"]
     }
   }
 }
@@ -74,7 +74,7 @@ Add to your Claude settings:
     "servers": {
       "google-news-trends": {
         "command": "uvx",
-        "args": ["google-news-trends-mcp"]
+        "args": ["google-news-trends-mcp@latest"]
       }
     }
   }
@@ -112,7 +112,7 @@ The following MCP tools are available:
 | **get_top_news**         | Fetch the top news stories from Google News.                       |
 | **get_trending_keywords**| Return trending keywords from Google Trends for a specified location.|
 
-All of the news related tools have an option to summarize the text of the article using NLP (nltk)
+All of the news related tools have an option to summarize the text of the article using LLM Sampling (if supported) or NLP
 
 
 ## CLI
@@ -153,4 +153,4 @@ npx @modelcontextprotocol/inspector uv run google-news-trends-mcp
 ```bash
 cd path/to/google/news/tends/mcp
 python -m pytest
-```
+```

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "google-news-trends-mcp"
-version = "0.1.7"
+version = "0.1.9"
 description = "An MCP server to access Google News and Google Trends."
 readme = "README.md"
 requires-python = ">=3.10.18"
@@ -63,4 +63,7 @@ pythonpath = "src"
 [project.urls]
 Homepage = "https://github.com/jmanek/google-news-trends-mcp"
 Repository = "https://github.com/jmanek/google-news-trends-mcp"
-Issues = "https://github.com/jmanek/google-news-trends-mcp/issues"
+Issues = "https://github.com/jmanek/google-news-trends-mcp/issues"
+
+[tool.black]
+line-length = 120

google_news_trends_mcp-0.1.9/src/google_news_trends_mcp/__init__.py

@@ -0,0 +1,2 @@
+import logging
+logging.getLogger(__name__).addHandler(logging.NullHandler())

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9}/src/google_news_trends_mcp/cli.py

@@ -17,9 +17,7 @@ def cli():
 
 @cli.command(help=get_news_by_keyword.__doc__)
 @click.argument("keyword")
-@click.option(
-    "--period", type=int, default=7, help="Period in days to search for articles."
-)
+@click.option("--period", type=int, default=7, help="Period in days to search for articles.")
 @click.option(
     "--max-results",
     "max_results",
@@ -27,24 +25,16 @@ def cli():
     default=10,
     help="Maximum number of results to return.",
 )
-@click.option(
-    "--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles."
-)
+@click.option("--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles.")
 def keyword(keyword, period, max_results, no_nlp):
-    articles = asyncio.run(
-        get_news_by_keyword(
-            keyword, period=period, max_results=max_results, nlp=not no_nlp
-        )
-    )
+    articles = asyncio.run(get_news_by_keyword(keyword, period=period, max_results=max_results, nlp=not no_nlp))
     # asyncio.run(articles)  # Ensure the articles are fetched asynchronously
     print_articles(articles)
 
 
 @cli.command(help=get_news_by_location.__doc__)
 @click.argument("location")
-@click.option(
-    "--period", type=int, default=7, help="Period in days to search for articles."
-)
+@click.option("--period", type=int, default=7, help="Period in days to search for articles.")
 @click.option(
     "--max-results",
     "max_results",
@@ -52,23 +42,15 @@ def keyword(keyword, period, max_results, no_nlp):
     default=10,
     help="Maximum number of results to return.",
 )
-@click.option(
-    "--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles."
-)
+@click.option("--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles.")
 def location(location, period, max_results, no_nlp):
-    articles = asyncio.run(
-        get_news_by_location(
-            location, period=period, max_results=max_results, nlp=not no_nlp
-        )
-    )
+    articles = asyncio.run(get_news_by_location(location, period=period, max_results=max_results, nlp=not no_nlp))
     print_articles(articles)
 
 
 @cli.command(help=get_news_by_topic.__doc__)
 @click.argument("topic")
-@click.option(
-    "--period", type=int, default=7, help="Period in days to search for articles."
-)
+@click.option("--period", type=int, default=7, help="Period in days to search for articles.")
 @click.option(
     "--max-results",
     "max_results",
@@ -76,23 +58,15 @@ def location(location, period, max_results, no_nlp):
     default=10,
     help="Maximum number of results to return.",
 )
-@click.option(
-    "--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles."
-)
+@click.option("--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles.")
 def topic(topic, period, max_results, no_nlp):
-    articles = asyncio.run(
-        get_news_by_topic(topic, period=period, max_results=max_results, nlp=not no_nlp)
-    )
+    articles = asyncio.run(get_news_by_topic(topic, period=period, max_results=max_results, nlp=not no_nlp))
     print_articles(articles)
 
 
 @cli.command(help=get_trending_terms.__doc__)
-@click.option(
-    "--geo", type=str, default="US", help="Country code, e.g. 'US', 'GB', 'IN', etc."
-)
-@click.option(
-    "--full-data", is_flag=True, default=False, help="Return full data for each trend."
-)
+@click.option("--geo", type=str, default="US", help="Country code, e.g. 'US', 'GB', 'IN', etc.")
+@click.option("--full-data", is_flag=True, default=False, help="Return full data for each trend.")
 @click.option(
     "--max-results",
     "max_results",
@@ -101,9 +75,7 @@ def topic(topic, period, max_results, no_nlp):
     help="Maximum number of results to return.",
 )
 def trending(geo, full_data, max_results):
-    trending_terms = asyncio.run(
-        get_trending_terms(geo=geo, full_data=full_data, max_results=max_results)
-    )
+    trending_terms = asyncio.run(get_trending_terms(geo=geo, full_data=full_data, max_results=max_results))
     if trending_terms:
         print("Trending terms:")
         for term in trending_terms:
@@ -116,9 +88,7 @@ def trending(geo, full_data, max_results):
 
 
 @cli.command(help=get_top_news.__doc__)
-@click.option(
-    "--period", type=int, default=3, help="Period in days to search for top articles."
-)
+@click.option("--period", type=int, default=3, help="Period in days to search for top articles.")
 @click.option(
     "--max-results",
     "max_results",
@@ -126,13 +96,9 @@ def trending(geo, full_data, max_results):
     default=10,
     help="Maximum number of results to return.",
 )
-@click.option(
-    "--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles."
-)
+@click.option("--no-nlp", is_flag=True, default=False, help="Disable NLP processing for articles.")
 def top(period, max_results, no_nlp):
-    articles = asyncio.run(
-        get_top_news(max_results=max_results, period=period, nlp=not no_nlp)
-    )
+    articles = asyncio.run(get_top_news(max_results=max_results, period=period, nlp=not no_nlp))
     print_articles(articles)
     print(f"Found {len(articles)} top articles.")
 

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9}/src/google_news_trends_mcp/news.py

@@ -8,7 +8,6 @@ It will fallback to using Playwright for websites that are difficult to scrape w
 
 import re
 import json
-import time
 import asyncio
 from gnews import GNews
 import newspaper  # newspaper4k
@@ -16,7 +15,6 @@ from googlenewsdecoder import gnewsdecoder
 import cloudscraper
 from playwright.async_api import async_playwright, Browser, Playwright
 from trendspy import Trends, TrendKeyword
-import click
 from typing import Optional, cast, overload, Literal, Awaitable
 import atexit
 from contextlib import asynccontextmanager
@@ -97,30 +95,15 @@ async def download_article_with_playwright(url) -> newspaper.Article | None:
             await page.goto(url, wait_until="domcontentloaded")
             await asyncio.sleep(2)  # Wait for the page to load completely
             content = await page.content()
-            article = newspaper.article(url, input_html=content, language="en")
+            article = newspaper.article(url, input_html=content)
             return article
     except Exception as e:
-        logging.warning(
-            f"Error downloading article with Playwright from {url}\n {e.args}"
-        )
+        logging.warning(f"Error downloading article with Playwright from {url}\n {e.args}")
         return None
 
 
-async def download_article(url: str, nlp: bool = True) -> newspaper.Article | None:
-    """
-    Download an article from a given URL using newspaper4k and cloudscraper (async).
-    """
+def download_article_with_scraper(url) -> newspaper.Article | None:
     article = None
-    if url.startswith("https://news.google.com/rss/"):
-        try:
-            decoded_url = gnewsdecoder(url)
-            if decoded_url.get("status"):
-                url = decoded_url["decoded_url"]
-            else:
-                logging.debug("Failed to decode Google News RSS link:")
-                return None
-        except Exception as err:
-            logging.warning(f"Error while decoding url {url}\n {err.args}")
     try:
         article = newspaper.article(url)
     except Exception as e:
@@ -135,22 +118,33 @@ async def download_article(url: str, nlp: bool = True) -> newspaper.Article | No
                     f"Failed to download article with cloudscraper from {url}, status code: {response.status_code}"
                 )
         except Exception as e:
-            logging.debug(
-                f"Error downloading article with cloudscraper from {url}\n {e.args}"
-            )
+            logging.debug(f"Error downloading article with cloudscraper from {url}\n {e.args}")
+    return article
 
-    try:
-        if article is None or not article.text:
-            # If newspaper failed, try downloading with Playwright
-            logging.debug(f"Retrying with Playwright for {url}")
-            article = await download_article_with_playwright(url)
-        article = cast(newspaper.Article, article)
-        article.parse()
-        if nlp:
-            article.nlp()
-    except Exception as e:
-        logging.warning(f"Error parsing article from {url}\n {e.args}")
+
+def decode_url(url: str) -> str:
+    if url.startswith("https://news.google.com/rss/"):
+        try:
+            decoded_url = gnewsdecoder(url)
+            if decoded_url.get("status"):
+                url = decoded_url["decoded_url"]
+            else:
+                logging.debug("Failed to decode Google News RSS link:")
+                return ""
+        except Exception as err:
+            logging.warning(f"Error while decoding url {url}\n {err.args}")
+    return url
+
+
+async def download_article(url: str) -> newspaper.Article | None:
+    """
+    Download an article from a given URL using newspaper4k and cloudscraper (async).
+    """
+    if not (url := decode_url(url)):
         return None
+    article = download_article_with_scraper(url)
+    if article is None or not article.text:
+        article = await download_article_with_playwright(url)
     return article
 
 
@@ -166,12 +160,13 @@ async def process_gnews_articles(
     articles = []
     total = len(gnews_articles)
     for idx, gnews_article in enumerate(gnews_articles):
-        article = await download_article(gnews_article["url"], nlp=nlp)
+        article = await download_article(gnews_article["url"])
         if article is None or not article.text:
-            logging.debug(
-                f"Failed to download article from {gnews_article['url']}:\n{article}"
-            )
+            logging.debug(f"Failed to download article from {gnews_article['url']}:\n{article}")
             continue
+        article.parse()
+        if nlp:
+            article.nlp()
         articles.append(article)
         if report_progress:
             await report_progress(idx, total)
@@ -196,13 +191,9 @@ async def get_news_by_keyword(
     google_news.max_results = max_results
     gnews_articles = google_news.get_news(keyword)
     if not gnews_articles:
-        logging.debug(
-            f"No articles found for keyword '{keyword}' in the last {period} days."
-        )
+        logging.debug(f"No articles found for keyword '{keyword}' in the last {period} days.")
         return []
-    return await process_gnews_articles(
-        gnews_articles, nlp=nlp, report_progress=report_progress
-    )
+    return await process_gnews_articles(gnews_articles, nlp=nlp, report_progress=report_progress)
 
 
 async def get_top_news(
@@ -223,9 +214,7 @@ async def get_top_news(
     if not gnews_articles:
         logging.debug("No top news articles found.")
         return []
-    return await process_gnews_articles(
-        gnews_articles, nlp=nlp, report_progress=report_progress
-    )
+    return await process_gnews_articles(gnews_articles, nlp=nlp, report_progress=report_progress)
 
 
 async def get_news_by_location(
@@ -245,13 +234,9 @@ async def get_news_by_location(
     google_news.max_results = max_results
     gnews_articles = google_news.get_news_by_location(location)
     if not gnews_articles:
-        logging.debug(
-            f"No articles found for location '{location}' in the last {period} days."
-        )
+        logging.debug(f"No articles found for location '{location}' in the last {period} days.")
         return []
-    return await process_gnews_articles(
-        gnews_articles, nlp=nlp, report_progress=report_progress
-    )
+    return await process_gnews_articles(gnews_articles, nlp=nlp, report_progress=report_progress)
 
 
 async def get_news_by_topic(
@@ -279,13 +264,9 @@ async def get_news_by_topic(
     google_news.max_results = max_results
     gnews_articles = google_news.get_news_by_topic(topic)
     if not gnews_articles:
-        logging.debug(
-            f"No articles found for topic '{topic}' in the last {period} days."
-        )
+        logging.debug(f"No articles found for topic '{topic}' in the last {period} days.")
         return []
-    return await process_gnews_articles(
-        gnews_articles, nlp=nlp, report_progress=report_progress
-    )
+    return await process_gnews_articles(gnews_articles, nlp=nlp, report_progress=report_progress)
 
 
 @overload
@@ -314,13 +295,9 @@ async def get_trending_terms(
     """
     try:
         trends = list(tr.trending_now(geo=geo))
-        trends = list(sorted(trends, key=lambda tt: tt.volume, reverse=True))[
-            :max_results
-        ]
+        trends = list(sorted(trends, key=lambda tt: tt.volume, reverse=True))[:max_results]
         if not full_data:
-            return [
-                {"keyword": trend.keyword, "volume": trend.volume} for trend in trends
-            ]
+            return [{"keyword": trend.keyword, "volume": trend.volume} for trend in trends]
         return trends
     except Exception as e:
         logging.warning(f"Error fetching trending terms: {e}")

google_news_trends_mcp-0.1.9/src/google_news_trends_mcp/server.py

@@ -0,0 +1,341 @@
+from typing import Annotated, Optional, Any, TYPE_CHECKING
+from fastmcp import FastMCP, Context
+from fastmcp.server.middleware.timing import TimingMiddleware
+from fastmcp.server.middleware.logging import LoggingMiddleware
+from fastmcp.server.middleware.rate_limiting import RateLimitingMiddleware
+from fastmcp.server.middleware.error_handling import ErrorHandlingMiddleware
+from mcp.types import TextContent
+from pydantic import BaseModel, Field, model_serializer
+from google_news_trends_mcp import news
+from newspaper import settings as newspaper_settings
+from newspaper.article import Article
+
+
+class BaseModelClean(BaseModel):
+    @model_serializer
+    def serializer(self, **kwargs) -> dict[str, Any]:
+        return {
+            field: self.__getattribute__(field)
+            for field in self.model_fields_set
+            if self.__getattribute__(field) is not None
+        }
+
+    if TYPE_CHECKING:
+
+        def model_dump(self, **kwargs) -> dict[str, Any]: ...
+
+
+class ArticleOut(BaseModelClean):
+    title: Annotated[str, Field(description="Title of the article.")]
+    url: Annotated[str, Field(description="Original article URL.")]
+    read_more_link: Annotated[Optional[str], Field(description="Link to read more about the article.")] = None
+    language: Annotated[Optional[str], Field(description="Language code of the article.")] = None
+    meta_img: Annotated[Optional[str], Field(description="Meta image URL.")] = None
+    movies: Annotated[Optional[list[str]], Field(description="List of movie URLs or IDs.")] = None
+    meta_favicon: Annotated[Optional[str], Field(description="Favicon URL from meta data.")] = None
+    meta_site_name: Annotated[Optional[str], Field(description="Site name from meta data.")] = None
+    authors: Annotated[Optional[list[str]], Field(description="list of authors.")] = None
+    publish_date: Annotated[Optional[str], Field(description="Publish date in ISO format.")] = None
+    top_image: Annotated[Optional[str], Field(description="URL of the top image.")] = None
+    images: Annotated[Optional[list[str]], Field(description="list of image URLs.")] = None
+    text: Annotated[Optional[str], Field(description="Full text of the article.")] = None
+    summary: Annotated[Optional[str], Field(description="Summary of the article.")] = None
+    keywords: Annotated[Optional[list[str]], Field(description="Extracted keywords.")] = None
+    tags: Annotated[Optional[list[str]], Field(description="Tags for the article.")] = None
+    meta_keywords: Annotated[Optional[list[str]], Field(description="Meta keywords from the article.")] = None
+    meta_description: Annotated[Optional[str], Field(description="Meta description from the article.")] = None
+    canonical_link: Annotated[Optional[str], Field(description="Canonical link for the article.")] = None
+    meta_data: Annotated[Optional[dict[str, str | int]], Field(description="Meta data dictionary.")] = None
+    meta_lang: Annotated[Optional[str], Field(description="Language of the article.")] = None
+    source_url: Annotated[Optional[str], Field(description="Source URL if different from original.")] = None
+
+
+class TrendingTermArticleOut(BaseModelClean):
+    title: Annotated[str, Field(description="Article title.")] = ""
+    url: Annotated[str, Field(description="Article URL.")] = ""
+    source: Annotated[Optional[str], Field(description="News source name.")] = None
+    picture: Annotated[Optional[str], Field(description="URL to article image.")] = None
+    time: Annotated[Optional[str | int], Field(description="Publication time or timestamp.")] = None
+    snippet: Annotated[Optional[str], Field(description="Article preview text.")] = None
+
+
+class TrendingTermOut(BaseModelClean):
+    keyword: Annotated[str, Field(description="Trending keyword.")]
+    volume: Annotated[Optional[int], Field(description="Search volume.")] = None
+    geo: Annotated[Optional[str], Field(description="Geographic location code.")] = None
+    started_timestamp: Annotated[
+        Optional[list],
+        Field(description="When the trend started (year, month, day, hour, minute, second)."),
+    ] = None
+    ended_timestamp: Annotated[
+        Optional[list],
+        Field(description="When the trend ended (year, month, day, hour, minute, second)."),
+    ] = None
+    volume_growth_pct: Annotated[Optional[float], Field(description="Percentage growth in search volume.")] = None
+    trend_keywords: Annotated[Optional[list[str]], Field(description="Related keywords.")] = None
+    topics: Annotated[Optional[list[str | int]], Field(description="Related topics.")] = None
+    news: Annotated[
+        Optional[list[TrendingTermArticleOut]],
+        Field(description="Related news articles."),
+    ] = None
+    news_tokens: Annotated[Optional[list], Field(description="Associated news tokens.")] = None
+    normalized_keyword: Annotated[Optional[str], Field(description="Normalized form of the keyword.")] = None
+
+
+mcp = FastMCP(
+    name="google-news-trends",
+    instructions="This server provides tools to search, analyze, and summarize Google News articles and Google Trends",
+    on_duplicate_tools="replace",
+)
+
+mcp.add_middleware(ErrorHandlingMiddleware())  # Handle errors first
+mcp.add_middleware(RateLimitingMiddleware(max_requests_per_second=50))
+mcp.add_middleware(TimingMiddleware())  # Time actual execution
+mcp.add_middleware(LoggingMiddleware())  # Log everything
+
+
+def set_newspaper_article_fields(full_data: bool = False):
+    if full_data:
+        newspaper_settings.article_json_fields = [
+            "url",
+            "read_more_link",
+            "language",
+            "title",
+            "top_image",
+            "meta_img",
+            "images",
+            "movies",
+            "keywords",
+            "keyword_scores",
+            "meta_keywords",
+            "tags",
+            "authors",
+            "publish_date",
+            "summary",
+            "meta_description",
+            "meta_lang",
+            "meta_favicon",
+            "meta_site_name",
+            "canonical_link",
+            "text",
+        ]
+    else:
+        newspaper_settings.article_json_fields = [
+            "url",
+            "title",
+            "publish_date",
+            "summary",
+        ]
+
+
+async def summarize_article(article: Article, ctx: Context) -> None:
+    if article.text:
+        prompt = f"Please provide a concise summary of the following news article:\n\n{article.text}"
+        response = await ctx.sample(prompt)
+        if isinstance(response, TextContent):
+            if not response.text:
+                await ctx.warning("LLM Sampling response is empty. Unable to summarize article.")
+                article.summary = "No summary available."
+            else:
+                article.summary = response.text
+        else:
+            await ctx.warning("LLM Sampling response is not a TextContent object. Unable to summarize article.")
+            article.summary = "No summary available."
+    else:
+        article.summary = "No summary available."
+
+
+@mcp.tool(
+    description=news.get_news_by_keyword.__doc__,
+    tags={"news", "articles", "keyword"},
+)
+async def get_news_by_keyword(
+    ctx: Context,
+    keyword: Annotated[str, Field(description="Search term to find articles.")],
+    period: Annotated[int, Field(description="Number of days to look back for articles.", ge=1)] = 7,
+    max_results: Annotated[int, Field(description="Maximum number of results to return.", ge=1)] = 10,
+    full_data: Annotated[
+        bool,
+        Field(
+            description="Return full data for each article. If False a summary should be created by setting the summarize flag"
+        ),
+    ] = False,
+    summarize: Annotated[
+        bool,
+        Field(
+            description="Generate a summary of the article, will first try LLM Sampling but if unavailable will use nlp"
+        ),
+    ] = True,
+) -> list[ArticleOut]:
+    set_newspaper_article_fields(full_data)
+    articles = await news.get_news_by_keyword(
+        keyword=keyword,
+        period=period,
+        max_results=max_results,
+        nlp=False,
+        report_progress=ctx.report_progress,
+    )
+    if summarize:
+        total_articles = len(articles)
+        try:
+            for idx, article in enumerate(articles):
+                await summarize_article(article, ctx)
+                await ctx.report_progress(idx, total_articles)
+        except Exception as err:
+            await ctx.debug(f"Failed to use LLM sampling for article summary:\n{err.args}")
+            for idx, article in enumerate(articles):
+                article.nlp()
+                await ctx.report_progress(idx, total_articles)
+    await ctx.report_progress(progress=len(articles), total=len(articles))
+    return [ArticleOut(**a.to_json(False)) for a in articles]
+
+
+@mcp.tool(
+    description=news.get_news_by_location.__doc__,
+    tags={"news", "articles", "location"},
+)
+async def get_news_by_location(
+    ctx: Context,
+    location: Annotated[str, Field(description="Name of city/state/country.")],
+    period: Annotated[int, Field(description="Number of days to look back for articles.", ge=1)] = 7,
+    max_results: Annotated[int, Field(description="Maximum number of results to return.", ge=1)] = 10,
+    full_data: Annotated[
+        bool,
+        Field(
+            description="Return full data for each article. If False a summary should be created by setting the summarize flag"
+        ),
+    ] = False,
+    summarize: Annotated[
+        bool,
+        Field(
+            description="Generate a summary of the article, will first try LLM Sampling but if unavailable will use nlp"
+        ),
+    ] = True,
+) -> list[ArticleOut]:
+    set_newspaper_article_fields(full_data)
+    articles = await news.get_news_by_location(
+        location=location,
+        period=period,
+        max_results=max_results,
+        nlp=False,
+        report_progress=ctx.report_progress,
+    )
+    if summarize:
+        total_articles = len(articles)
+        try:
+            for idx, article in enumerate(articles):
+                await summarize_article(article, ctx)
+                await ctx.report_progress(idx, total_articles)
+        except Exception as err:
+            await ctx.debug(f"Failed to use LLM sampling for article summary:\n{err.args}")
+            for idx, article in enumerate(articles):
+                article.nlp()
+                await ctx.report_progress(idx, total_articles)
+    await ctx.report_progress(progress=len(articles), total=len(articles))
+    return [ArticleOut(**a.to_json(False)) for a in articles]
+
+
+@mcp.tool(description=news.get_news_by_topic.__doc__, tags={"news", "articles", "topic"})
+async def get_news_by_topic(
+    ctx: Context,
+    topic: Annotated[str, Field(description="Topic to search for articles.")],
+    period: Annotated[int, Field(description="Number of days to look back for articles.", ge=1)] = 7,
+    max_results: Annotated[int, Field(description="Maximum number of results to return.", ge=1)] = 10,
+    full_data: Annotated[
+        bool,
+        Field(
+            description="Return full data for each article. If False a summary should be created by setting the summarize flag"
+        ),
+    ] = False,
+    summarize: Annotated[
+        bool,
+        Field(
+            description="Generate a summary of the article, will first try LLM Sampling but if unavailable will use nlp"
+        ),
+    ] = True,
+) -> list[ArticleOut]:
+    set_newspaper_article_fields(full_data)
+    articles = await news.get_news_by_topic(
+        topic=topic,
+        period=period,
+        max_results=max_results,
+        nlp=False,
+        report_progress=ctx.report_progress,
+    )
+    if summarize:
+        total_articles = len(articles)
+        try:
+            for idx, article in enumerate(articles):
+                await summarize_article(article, ctx)
+                await ctx.report_progress(idx, total_articles)
+        except Exception as err:
+            await ctx.debug(f"Failed to use LLM sampling for article summary:\n{err.args}")
+            for idx, article in enumerate(articles):
+                article.nlp()
+                await ctx.report_progress(idx, total_articles)
+
+    await ctx.report_progress(progress=len(articles), total=len(articles))
+    return [ArticleOut(**a.to_json(False)) for a in articles]
+
+
+@mcp.tool(description=news.get_top_news.__doc__, tags={"news", "articles", "top"})
+async def get_top_news(
+    ctx: Context,
+    period: Annotated[int, Field(description="Number of days to look back for top articles.", ge=1)] = 3,
+    max_results: Annotated[int, Field(description="Maximum number of results to return.", ge=1)] = 10,
+    full_data: Annotated[
+        bool,
+        Field(
+            description="Return full data for each article. If False a summary should be created by setting the summarize flag"
+        ),
+    ] = False,
+    summarize: Annotated[
+        bool,
+        Field(
+            description="Generate a summary of the article, will first try LLM Sampling but if unavailable will use nlp"
+        ),
+    ] = True,
+) -> list[ArticleOut]:
+    set_newspaper_article_fields(full_data)
+    articles = await news.get_top_news(
+        period=period,
+        max_results=max_results,
+        nlp=False,
+        report_progress=ctx.report_progress,
+    )
+    if summarize:
+        total_articles = len(articles)
+        try:
+            for idx, article in enumerate(articles):
+                await summarize_article(article, ctx)
+                await ctx.report_progress(idx, total_articles)
+        except Exception as err:
+            await ctx.debug(f"Failed to use LLM sampling for article summary:\n{err.args}")
+            for idx, article in enumerate(articles):
+                article.nlp()
+                await ctx.report_progress(idx, total_articles)
+
+    await ctx.report_progress(progress=len(articles), total=len(articles))
+    return [ArticleOut(**a.to_json(False)) for a in articles]
+
+
+@mcp.tool(description=news.get_trending_terms.__doc__, tags={"trends", "google", "trending"})
+async def get_trending_terms(
+    geo: Annotated[str, Field(description="Country code, e.g. 'US', 'GB', 'IN', etc.")] = "US",
+    full_data: Annotated[
+        bool,
+        Field(description="Return full data for each trend. Should be False for most use cases."),
+    ] = False,
+    max_results: Annotated[int, Field(description="Maximum number of results to return.", ge=1)] = 100,
+) -> list[TrendingTermOut]:
+
+    if not full_data:
+        trends = await news.get_trending_terms(geo=geo, full_data=False, max_results=max_results)
+        return [TrendingTermOut(keyword=str(tt["keyword"]), volume=tt["volume"]) for tt in trends]
+
+    trends = await news.get_trending_terms(geo=geo, full_data=True, max_results=max_results)
+    return [TrendingTermOut(**tt.__dict__) for tt in trends]
+
+
+def main():
+    mcp.run()

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9/src/google_news_trends_mcp.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: google-news-trends-mcp
-Version: 0.1.7
+Version: 0.1.9
 Summary: An MCP server to access Google News and Google Trends.
 Author-email: Jesse Manek <jesse.manek@gmail.com>
 License-Expression: MIT
@@ -35,8 +35,8 @@ An MCP server to access Google News and Google Trends.  Does not rely on any pai
 
 - Search Google News articles based on keyword, location, topic
 - Get top news stories from Google News
-- Google Trends keywords base on location
-- Optional NLP to summarize articles and extract keywords
+- Google Trends keywords based on location
+- Optional LLM Sampling and NLP to summarize articles and extract keywords
 
 ## Installation
 
@@ -70,7 +70,7 @@ Add to your Claude settings:
   "mcpServers": {
     "google-news-trends": {
       "command": "uvx",
-      "args": ["google-news-trends-mcp"]
+      "args": ["google-news-trends-mcp@latest"]
     }
   }
 }
@@ -103,7 +103,7 @@ Add to your Claude settings:
     "servers": {
       "google-news-trends": {
         "command": "uvx",
-        "args": ["google-news-trends-mcp"]
+        "args": ["google-news-trends-mcp@latest"]
       }
     }
   }
@@ -141,7 +141,7 @@ The following MCP tools are available:
 | **get_top_news**         | Fetch the top news stories from Google News.                       |
 | **get_trending_keywords**| Return trending keywords from Google Trends for a specified location.|
 
-All of the news related tools have an option to summarize the text of the article using NLP (nltk)
+All of the news related tools have an option to summarize the text of the article using LLM Sampling (if supported) or NLP
 
 
 ## CLI

{google_news_trends_mcp-0.1.7 → google_news_trends_mcp-0.1.9}/tests/test_server.py

@@ -11,50 +11,56 @@ def mcp_server():
 
 async def test_get_news_by_keyword(mcp_server):
     async with Client(mcp_server) as client:
-        params = {"keyword": "AI", "period": 3, "max_results": 2, "nlp": True}
+        params = {"keyword": "AI", "period": 3, "max_results": 2}
         result = await client.call_tool("get_news_by_keyword", params)
         assert isinstance(result, list)
         assert len(result) <= 2
         for article in result:
-            article = json.loads(article.text)[
-                0
-            ]  # Assuming articles are returned as JSON strings
+            article = json.loads(article.text)
+            if isinstance(article, list):
+                article = article[0]  # Assuming articles are returned as JSON strings
             assert "title" in article
             assert "url" in article
 
 
 async def test_get_news_by_location(mcp_server):
     async with Client(mcp_server) as client:
-        params = {"location": "California", "period": 3, "max_results": 2, "nlp": False}
+        params = {"location": "California", "period": 3, "max_results": 2}
         result = await client.call_tool("get_news_by_location", params)
         assert isinstance(result, list)
         assert len(result) <= 2
         for article in result:
-            article = json.loads(article.text)[0]
+            article = json.loads(article.text)
+            if isinstance(article, list):
+                article = article[0]
             assert "title" in article
             assert "url" in article
 
 
 async def test_get_news_by_topic(mcp_server):
     async with Client(mcp_server) as client:
-        params = {"topic": "TECHNOLOGY", "period": 3, "max_results": 2, "nlp": True}
+        params = {"topic": "TECHNOLOGY", "period": 3, "max_results": 2}
         result = await client.call_tool("get_news_by_topic", params)
         assert isinstance(result, list)
         assert len(result) <= 2
         for article in result:
-            article = json.loads(article.text)[0]
+            article = json.loads(article.text)
+            if isinstance(article, list):
+                article = article[0]
             assert "title" in article
             assert "url" in article
 
 
 async def test_get_top_news(mcp_server):
     async with Client(mcp_server) as client:
-        params = {"period": 2, "max_results": 2, "nlp": False}
+        params = {"period": 2, "max_results": 2}
         result = await client.call_tool("get_top_news", params)
         assert isinstance(result, list)
         assert len(result) <= 2
         for article in result:
-            article = json.loads(article.text)[0]
+            article = json.loads(article.text)
+            if isinstance(article, list):
+                article = article[0]
             assert "title" in article
             assert "url" in article
 

google_news_trends_mcp-0.1.7/src/google_news_trends_mcp/__init__.py

@@ -1,2 +0,0 @@
-import logging
-logging.getLogger(__name__).addHandler(logging.NullHandler())

google_news_trends_mcp-0.1.7/src/google_news_trends_mcp/server.py

@@ -1,329 +0,0 @@
-from fastmcp import FastMCP, Context
-from fastmcp.exceptions import ToolError
-from fastmcp.server.dependencies import get_context
-from pydantic import BaseModel, Field
-from typing import Optional
-from google_news_trends_mcp import news
-from typing import Annotated
-from newspaper import settings as newspaper_settings
-from fastmcp.server.middleware.timing import TimingMiddleware
-from fastmcp.server.middleware.logging import LoggingMiddleware
-from fastmcp.server.middleware.rate_limiting import RateLimitingMiddleware
-from fastmcp.server.middleware.error_handling import ErrorHandlingMiddleware
-
-
-class ArticleOut(BaseModel):
-    read_more_link: Annotated[
-        Optional[str], Field(description="Link to read more about the article.")
-    ] = None
-    language: Annotated[
-        Optional[str], Field(description="Language code of the article.")
-    ] = None
-    meta_img: Annotated[Optional[str], Field(description="Meta image URL.")] = None
-    movies: Annotated[
-        Optional[list[str]], Field(description="List of movie URLs or IDs.")
-    ] = None
-    meta_favicon: Annotated[
-        Optional[str], Field(description="Favicon URL from meta data.")
-    ] = None
-    meta_site_name: Annotated[
-        Optional[str], Field(description="Site name from meta data.")
-    ] = None
-    title: Annotated[str, Field(description="Title of the article.")]
-    authors: Annotated[Optional[list[str]], Field(description="list of authors.")] = (
-        None
-    )
-    publish_date: Annotated[
-        Optional[str], Field(description="Publish date in ISO format.")
-    ] = None
-    top_image: Annotated[Optional[str], Field(description="URL of the top image.")] = (
-        None
-    )
-    images: Annotated[Optional[list[str]], Field(description="list of image URLs.")] = (
-        None
-    )
-    text: Annotated[str, Field(description="Full text of the article.")]
-    url: Annotated[str, Field(description="Original article URL.")]
-    summary: Annotated[Optional[str], Field(description="Summary of the article.")] = (
-        None
-    )
-    keywords: Annotated[
-        Optional[list[str]], Field(description="Extracted keywords.")
-    ] = None
-    tags: Annotated[Optional[list[str]], Field(description="Tags for the article.")] = (
-        None
-    )
-    meta_keywords: Annotated[
-        Optional[list[str]], Field(description="Meta keywords from the article.")
-    ] = None
-    meta_description: Annotated[
-        Optional[str], Field(description="Meta description from the article.")
-    ] = None
-    canonical_link: Annotated[
-        Optional[str], Field(description="Canonical link for the article.")
-    ] = None
-    meta_data: Annotated[
-        Optional[dict[str, str | int]], Field(description="Meta data dictionary.")
-    ] = None
-    meta_lang: Annotated[
-        Optional[str], Field(description="Language of the article.")
-    ] = None
-    source_url: Annotated[
-        Optional[str], Field(description="Source URL if different from original.")
-    ] = None
-
-
-class TrendingTermArticleOut(BaseModel):
-    title: Annotated[str, Field(description="Article title.")] = ""
-    url: Annotated[str, Field(description="Article URL.")] = ""
-    source: Annotated[Optional[str], Field(description="News source name.")] = None
-    picture: Annotated[Optional[str], Field(description="URL to article image.")] = None
-    time: Annotated[
-        Optional[str | int], Field(description="Publication time or timestamp.")
-    ] = None
-    snippet: Annotated[Optional[str], Field(description="Article preview text.")] = None
-
-
-class TrendingTermOut(BaseModel):
-    keyword: Annotated[str, Field(description="Trending keyword.")]
-    volume: Annotated[Optional[int], Field(description="Search volume.")] = None
-    geo: Annotated[Optional[str], Field(description="Geographic location code.")] = None
-    started_timestamp: Annotated[
-        Optional[list],
-        Field(
-            description="When the trend started (year, month, day, hour, minute, second)."
-        ),
-    ] = None
-    ended_timestamp: Annotated[
-        Optional[tuple[int, int]],
-        Field(
-            description="When the trend ended (year, month, day, hour, minute, second)."
-        ),
-    ] = None
-    volume_growth_pct: Annotated[
-        Optional[float], Field(description="Percentage growth in search volume.")
-    ] = None
-    trend_keywords: Annotated[
-        Optional[list[str]], Field(description="Related keywords.")
-    ] = None
-    topics: Annotated[
-        Optional[list[str | int]], Field(description="Related topics.")
-    ] = None
-    news: Annotated[
-        Optional[list[TrendingTermArticleOut]],
-        Field(description="Related news articles."),
-    ] = None
-    news_tokens: Annotated[
-        Optional[list], Field(description="Associated news tokens.")
-    ] = None
-    normalized_keyword: Annotated[
-        Optional[str], Field(description="Normalized form of the keyword.")
-    ] = None
-
-
-mcp = FastMCP(
-    name="google-news-trends",
-    instructions="This server provides tools to search, analyze, and summarize Google News articles and Google Trends",
-    on_duplicate_tools="replace",
-)
-
-mcp.add_middleware(ErrorHandlingMiddleware())  # Handle errors first
-mcp.add_middleware(RateLimitingMiddleware(max_requests_per_second=50))
-mcp.add_middleware(TimingMiddleware())  # Time actual execution
-mcp.add_middleware(LoggingMiddleware())  # Log everything
-
-
-# Configure newspaper settings for article extraction
-def set_newspaper_article_fields(full_data: bool = False):
-    if full_data:
-        newspaper_settings.article_json_fields = [
-            "url",
-            "read_more_link",
-            "language",
-            "title",
-            "top_image",
-            "meta_img",
-            "images",
-            "movies",
-            "keywords",
-            "keyword_scores",
-            "meta_keywords",
-            "tags",
-            "authors",
-            "publish_date",
-            "summary",
-            "meta_description",
-            "meta_lang",
-            "meta_favicon",
-            "meta_site_name",
-            "canonical_link",
-            "text",
-        ]
-    else:
-        newspaper_settings.article_json_fields = [
-            "url",
-            "title",
-            "text",
-            "publish_date",
-            "summary",
-            "keywords",
-        ]
-
-
-@mcp.tool(
-    description=news.get_news_by_keyword.__doc__,
-    tags={"news", "articles", "keyword"},
-)
-async def get_news_by_keyword(
-    ctx: Context,
-    keyword: Annotated[str, Field(description="Search term to find articles.")],
-    period: Annotated[
-        int, Field(description="Number of days to look back for articles.", ge=1)
-    ] = 7,
-    max_results: Annotated[
-        int, Field(description="Maximum number of results to return.", ge=1)
-    ] = 10,
-    nlp: Annotated[
-        bool, Field(description="Whether to perform NLP on the articles.")
-    ] = False,
-    full_data: Annotated[
-        bool, Field(description="Return full data for each article.")
-    ] = False,
-) -> list[ArticleOut]:
-    set_newspaper_article_fields(full_data)
-    articles = await news.get_news_by_keyword(
-        keyword=keyword,
-        period=period,
-        max_results=max_results,
-        nlp=nlp,
-        report_progress=ctx.report_progress,
-    )
-    await ctx.report_progress(progress=len(articles), total=len(articles))
-    return [ArticleOut(**a.to_json(False)) for a in articles]
-
-
-@mcp.tool(
-    description=news.get_news_by_location.__doc__,
-    tags={"news", "articles", "location"},
-)
-async def get_news_by_location(
-    ctx: Context,
-    location: Annotated[str, Field(description="Name of city/state/country.")],
-    period: Annotated[
-        int, Field(description="Number of days to look back for articles.", ge=1)
-    ] = 7,
-    max_results: Annotated[
-        int, Field(description="Maximum number of results to return.", ge=1)
-    ] = 10,
-    nlp: Annotated[
-        bool, Field(description="Whether to perform NLP on the articles.")
-    ] = False,
-    full_data: Annotated[
-        bool, Field(description="Return full data for each article.")
-    ] = False,
-) -> list[ArticleOut]:
-    set_newspaper_article_fields(full_data)
-    articles = await news.get_news_by_location(
-        location=location,
-        period=period,
-        max_results=max_results,
-        nlp=nlp,
-        report_progress=ctx.report_progress,
-    )
-    await ctx.report_progress(progress=len(articles), total=len(articles))
-    return [ArticleOut(**a.to_json(False)) for a in articles]
-
-
-@mcp.tool(
-    description=news.get_news_by_topic.__doc__, tags={"news", "articles", "topic"}
-)
-async def get_news_by_topic(
-    ctx: Context,
-    topic: Annotated[str, Field(description="Topic to search for articles.")],
-    period: Annotated[
-        int, Field(description="Number of days to look back for articles.", ge=1)
-    ] = 7,
-    max_results: Annotated[
-        int, Field(description="Maximum number of results to return.", ge=1)
-    ] = 10,
-    nlp: Annotated[
-        bool, Field(description="Whether to perform NLP on the articles.")
-    ] = False,
-    full_data: Annotated[
-        bool, Field(description="Return full data for each article.")
-    ] = False,
-) -> list[ArticleOut]:
-    set_newspaper_article_fields(full_data)
-    articles = await news.get_news_by_topic(
-        topic=topic,
-        period=period,
-        max_results=max_results,
-        nlp=nlp,
-        report_progress=ctx.report_progress,
-    )
-    await ctx.report_progress(progress=len(articles), total=len(articles))
-    return [ArticleOut(**a.to_json(False)) for a in articles]
-
-
-@mcp.tool(description=news.get_top_news.__doc__, tags={"news", "articles", "top"})
-async def get_top_news(
-    ctx: Context,
-    period: Annotated[
-        int, Field(description="Number of days to look back for top articles.", ge=1)
-    ] = 3,
-    max_results: Annotated[
-        int, Field(description="Maximum number of results to return.", ge=1)
-    ] = 10,
-    nlp: Annotated[
-        bool, Field(description="Whether to perform NLP on the articles.")
-    ] = False,
-    full_data: Annotated[
-        bool, Field(description="Return full data for each article.")
-    ] = False,
-) -> list[ArticleOut]:
-    set_newspaper_article_fields(full_data)
-    articles = await news.get_top_news(
-        period=period,
-        max_results=max_results,
-        nlp=nlp,
-        report_progress=ctx.report_progress,
-    )
-    await ctx.report_progress(progress=len(articles), total=len(articles))
-    return [ArticleOut(**a.to_json(False)) for a in articles]
-
-
-@mcp.tool(
-    description=news.get_trending_terms.__doc__, tags={"trends", "google", "trending"}
-)
-async def get_trending_terms(
-    geo: Annotated[
-        str, Field(description="Country code, e.g. 'US', 'GB', 'IN', etc.")
-    ] = "US",
-    full_data: Annotated[
-        bool,
-        Field(
-            description="Return full data for each trend. Should be False for most use cases."
-        ),
-    ] = False,
-    max_results: Annotated[
-        int, Field(description="Maximum number of results to return.", ge=1)
-    ] = 100,
-) -> list[TrendingTermOut]:
-
-    if not full_data:
-        trends = await news.get_trending_terms(
-            geo=geo, full_data=False, max_results=max_results
-        )
-        return [
-            TrendingTermOut(keyword=str(tt["keyword"]), volume=tt["volume"])
-            for tt in trends
-        ]
-
-    trends = await news.get_trending_terms(
-        geo=geo, full_data=True, max_results=max_results
-    )
-    return [TrendingTermOut(**tt.__dict__) for tt in trends]
-
-
-def main():
-    mcp.run()