Spaces:
Running
Running
michaelkri commited on
Commit ·
74e7eda
1
Parent(s): 6d34b7a
Article photos
Browse files- app/database.py +1 -0
- app/news_fetcher.py +10 -8
- app/scraper.py +15 -3
- app/static/style.css +78 -1
- app/templates/index.html +4 -1
app/database.py
CHANGED
|
@@ -19,6 +19,7 @@ class Article(Base):
|
|
| 19 |
content: Mapped[str] = mapped_column(UnicodeText)
|
| 20 |
date: Mapped[datetime.datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
|
| 21 |
category: Mapped[str] = mapped_column(UnicodeText, server_default='news')
|
|
|
|
| 22 |
|
| 23 |
sources = relationship('Source', backref='article')
|
| 24 |
|
|
|
|
| 19 |
content: Mapped[str] = mapped_column(UnicodeText)
|
| 20 |
date: Mapped[datetime.datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
|
| 21 |
category: Mapped[str] = mapped_column(UnicodeText, server_default='news')
|
| 22 |
+
image_url: Mapped[str] = mapped_column(UnicodeText, nullable=True, server_default=None)
|
| 23 |
|
| 24 |
sources = relationship('Source', backref='article')
|
| 25 |
|
app/news_fetcher.py
CHANGED
|
@@ -8,15 +8,16 @@ from datetime import datetime
|
|
| 8 |
|
| 9 |
def topic_summary(summarizer: Summarizer, query: str, max_results: int = 5, min_cluster_size: int = 2) -> str:
|
| 10 |
logging.debug(f'Beginning search for \'{query}\'...')
|
| 11 |
-
|
| 12 |
-
logging.debug(f'Retrieved {len(
|
| 13 |
-
|
| 14 |
-
articles = [item[0] for item in articles_and_urls]
|
| 15 |
-
urls = [item[1] for item in articles_and_urls]
|
| 16 |
|
|
|
|
| 17 |
summary = summarizer.multisource_summary(articles, min_cluster_size=min_cluster_size)
|
| 18 |
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
| 20 |
|
| 21 |
|
| 22 |
def fetch_headlines(summarizer: Summarizer, rss_feed_urls: list[str], max_articles=50) -> list[str]:
|
|
@@ -49,7 +50,7 @@ def news_summary(summarizer: Summarizer, rss_feed_urls: list[str]):
|
|
| 49 |
|
| 50 |
# summarize each topic and yield articles
|
| 51 |
for headline in headlines:
|
| 52 |
-
summary, source_urls = topic_summary(summarizer, headline)
|
| 53 |
|
| 54 |
# no articles were found
|
| 55 |
if not summary:
|
|
@@ -62,7 +63,8 @@ def news_summary(summarizer: Summarizer, rss_feed_urls: list[str]):
|
|
| 62 |
title=headline,
|
| 63 |
content=summary,
|
| 64 |
date=datetime.now(),
|
| 65 |
-
category=category
|
|
|
|
| 66 |
)
|
| 67 |
|
| 68 |
sources = [Source(url=s_url, article_id=article.id) for s_url in source_urls]
|
|
|
|
| 8 |
|
| 9 |
def topic_summary(summarizer: Summarizer, query: str, max_results: int = 5, min_cluster_size: int = 2) -> str:
|
| 10 |
logging.debug(f'Beginning search for \'{query}\'...')
|
| 11 |
+
articles, urls, images = get_articles(query, max_results=max_results)
|
| 12 |
+
logging.debug(f'Retrieved {len(urls)} articles')
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
+
# create summary of all articles
|
| 15 |
summary = summarizer.multisource_summary(articles, min_cluster_size=min_cluster_size)
|
| 16 |
|
| 17 |
+
# grab an arbitrary image for article
|
| 18 |
+
image = next((image for image in images if image is not None), None)
|
| 19 |
+
|
| 20 |
+
return summary, urls, image
|
| 21 |
|
| 22 |
|
| 23 |
def fetch_headlines(summarizer: Summarizer, rss_feed_urls: list[str], max_articles=50) -> list[str]:
|
|
|
|
| 50 |
|
| 51 |
# summarize each topic and yield articles
|
| 52 |
for headline in headlines:
|
| 53 |
+
summary, source_urls, article_image = topic_summary(summarizer, headline)
|
| 54 |
|
| 55 |
# no articles were found
|
| 56 |
if not summary:
|
|
|
|
| 63 |
title=headline,
|
| 64 |
content=summary,
|
| 65 |
date=datetime.now(),
|
| 66 |
+
category=category,
|
| 67 |
+
image_url=article_image
|
| 68 |
)
|
| 69 |
|
| 70 |
sources = [Source(url=s_url, article_id=article.id) for s_url in source_urls]
|
app/scraper.py
CHANGED
|
@@ -25,7 +25,19 @@ def get_articles(query, max_results=5, min_article_length=100) -> list[tuple[str
|
|
| 25 |
# get article urls
|
| 26 |
urls = set([r['url'] for r in search_results])
|
| 27 |
|
| 28 |
-
# try to retrieve articles
|
| 29 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
# get article urls
|
| 26 |
urls = set([r['url'] for r in search_results])
|
| 27 |
|
| 28 |
+
# try to retrieve articles
|
| 29 |
+
texts = []
|
| 30 |
+
urls = []
|
| 31 |
+
images = []
|
| 32 |
+
for result in search_results:
|
| 33 |
+
article_url = result['url']
|
| 34 |
+
article_image = result['image']
|
| 35 |
+
article_text = retrieve_article(article_url)
|
| 36 |
|
| 37 |
+
# filter short and inaccessible articles
|
| 38 |
+
if article_text and len(article_text) > min_article_length:
|
| 39 |
+
texts.append(article_text)
|
| 40 |
+
urls.append(article_url)
|
| 41 |
+
images.append(article_image)
|
| 42 |
+
|
| 43 |
+
return texts, urls, images
|
app/static/style.css
CHANGED
|
@@ -1,4 +1,3 @@
|
|
| 1 |
-
/* Focal Stylesheet */
|
| 2 |
/*! tailwindcss v4.1.13 | MIT License | https://tailwindcss.com */
|
| 3 |
@layer properties;
|
| 4 |
@layer theme, base, components, utilities;
|
|
@@ -317,6 +316,9 @@
|
|
| 317 |
.block {
|
| 318 |
display: block;
|
| 319 |
}
|
|
|
|
|
|
|
|
|
|
| 320 |
.flex {
|
| 321 |
display: flex;
|
| 322 |
}
|
|
@@ -344,6 +346,9 @@
|
|
| 344 |
.h-8 {
|
| 345 |
height: calc(var(--spacing) * 8);
|
| 346 |
}
|
|
|
|
|
|
|
|
|
|
| 347 |
.h-full {
|
| 348 |
height: 100%;
|
| 349 |
}
|
|
@@ -535,6 +540,9 @@
|
|
| 535 |
.bg-white {
|
| 536 |
background-color: var(--color-white);
|
| 537 |
}
|
|
|
|
|
|
|
|
|
|
| 538 |
.p-2 {
|
| 539 |
padding: calc(var(--spacing) * 2);
|
| 540 |
}
|
|
@@ -691,6 +699,9 @@
|
|
| 691 |
--tw-shadow: 0 1px 3px 0 var(--tw-shadow-color, rgb(0 0 0 / 0.1)), 0 1px 2px -1px var(--tw-shadow-color, rgb(0 0 0 / 0.1));
|
| 692 |
box-shadow: var(--tw-inset-shadow), var(--tw-inset-ring-shadow), var(--tw-ring-offset-shadow), var(--tw-ring-shadow), var(--tw-shadow);
|
| 693 |
}
|
|
|
|
|
|
|
|
|
|
| 694 |
.transition {
|
| 695 |
transition-property: color, background-color, border-color, outline-color, text-decoration-color, fill, stroke, --tw-gradient-from, --tw-gradient-via, --tw-gradient-to, opacity, box-shadow, transform, translate, scale, rotate, filter, -webkit-backdrop-filter, backdrop-filter, display, content-visibility, overlay, pointer-events;
|
| 696 |
transition-timing-function: var(--tw-ease, var(--default-transition-timing-function));
|
|
@@ -970,6 +981,59 @@
|
|
| 970 |
inherits: false;
|
| 971 |
initial-value: 0 0 #0000;
|
| 972 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 973 |
@property --tw-duration {
|
| 974 |
syntax: "*";
|
| 975 |
inherits: false;
|
|
@@ -1011,6 +1075,19 @@
|
|
| 1011 |
--tw-ring-offset-width: 0px;
|
| 1012 |
--tw-ring-offset-color: #fff;
|
| 1013 |
--tw-ring-offset-shadow: 0 0 #0000;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1014 |
--tw-duration: initial;
|
| 1015 |
}
|
| 1016 |
}
|
|
|
|
|
|
|
| 1 |
/*! tailwindcss v4.1.13 | MIT License | https://tailwindcss.com */
|
| 2 |
@layer properties;
|
| 3 |
@layer theme, base, components, utilities;
|
|
|
|
| 316 |
.block {
|
| 317 |
display: block;
|
| 318 |
}
|
| 319 |
+
.contents {
|
| 320 |
+
display: contents;
|
| 321 |
+
}
|
| 322 |
.flex {
|
| 323 |
display: flex;
|
| 324 |
}
|
|
|
|
| 346 |
.h-8 {
|
| 347 |
height: calc(var(--spacing) * 8);
|
| 348 |
}
|
| 349 |
+
.h-48 {
|
| 350 |
+
height: calc(var(--spacing) * 48);
|
| 351 |
+
}
|
| 352 |
.h-full {
|
| 353 |
height: 100%;
|
| 354 |
}
|
|
|
|
| 540 |
.bg-white {
|
| 541 |
background-color: var(--color-white);
|
| 542 |
}
|
| 543 |
+
.object-cover {
|
| 544 |
+
object-fit: cover;
|
| 545 |
+
}
|
| 546 |
.p-2 {
|
| 547 |
padding: calc(var(--spacing) * 2);
|
| 548 |
}
|
|
|
|
| 699 |
--tw-shadow: 0 1px 3px 0 var(--tw-shadow-color, rgb(0 0 0 / 0.1)), 0 1px 2px -1px var(--tw-shadow-color, rgb(0 0 0 / 0.1));
|
| 700 |
box-shadow: var(--tw-inset-shadow), var(--tw-inset-ring-shadow), var(--tw-ring-offset-shadow), var(--tw-ring-shadow), var(--tw-shadow);
|
| 701 |
}
|
| 702 |
+
.filter {
|
| 703 |
+
filter: var(--tw-blur,) var(--tw-brightness,) var(--tw-contrast,) var(--tw-grayscale,) var(--tw-hue-rotate,) var(--tw-invert,) var(--tw-saturate,) var(--tw-sepia,) var(--tw-drop-shadow,);
|
| 704 |
+
}
|
| 705 |
.transition {
|
| 706 |
transition-property: color, background-color, border-color, outline-color, text-decoration-color, fill, stroke, --tw-gradient-from, --tw-gradient-via, --tw-gradient-to, opacity, box-shadow, transform, translate, scale, rotate, filter, -webkit-backdrop-filter, backdrop-filter, display, content-visibility, overlay, pointer-events;
|
| 707 |
transition-timing-function: var(--tw-ease, var(--default-transition-timing-function));
|
|
|
|
| 981 |
inherits: false;
|
| 982 |
initial-value: 0 0 #0000;
|
| 983 |
}
|
| 984 |
+
@property --tw-blur {
|
| 985 |
+
syntax: "*";
|
| 986 |
+
inherits: false;
|
| 987 |
+
}
|
| 988 |
+
@property --tw-brightness {
|
| 989 |
+
syntax: "*";
|
| 990 |
+
inherits: false;
|
| 991 |
+
}
|
| 992 |
+
@property --tw-contrast {
|
| 993 |
+
syntax: "*";
|
| 994 |
+
inherits: false;
|
| 995 |
+
}
|
| 996 |
+
@property --tw-grayscale {
|
| 997 |
+
syntax: "*";
|
| 998 |
+
inherits: false;
|
| 999 |
+
}
|
| 1000 |
+
@property --tw-hue-rotate {
|
| 1001 |
+
syntax: "*";
|
| 1002 |
+
inherits: false;
|
| 1003 |
+
}
|
| 1004 |
+
@property --tw-invert {
|
| 1005 |
+
syntax: "*";
|
| 1006 |
+
inherits: false;
|
| 1007 |
+
}
|
| 1008 |
+
@property --tw-opacity {
|
| 1009 |
+
syntax: "*";
|
| 1010 |
+
inherits: false;
|
| 1011 |
+
}
|
| 1012 |
+
@property --tw-saturate {
|
| 1013 |
+
syntax: "*";
|
| 1014 |
+
inherits: false;
|
| 1015 |
+
}
|
| 1016 |
+
@property --tw-sepia {
|
| 1017 |
+
syntax: "*";
|
| 1018 |
+
inherits: false;
|
| 1019 |
+
}
|
| 1020 |
+
@property --tw-drop-shadow {
|
| 1021 |
+
syntax: "*";
|
| 1022 |
+
inherits: false;
|
| 1023 |
+
}
|
| 1024 |
+
@property --tw-drop-shadow-color {
|
| 1025 |
+
syntax: "*";
|
| 1026 |
+
inherits: false;
|
| 1027 |
+
}
|
| 1028 |
+
@property --tw-drop-shadow-alpha {
|
| 1029 |
+
syntax: "<percentage>";
|
| 1030 |
+
inherits: false;
|
| 1031 |
+
initial-value: 100%;
|
| 1032 |
+
}
|
| 1033 |
+
@property --tw-drop-shadow-size {
|
| 1034 |
+
syntax: "*";
|
| 1035 |
+
inherits: false;
|
| 1036 |
+
}
|
| 1037 |
@property --tw-duration {
|
| 1038 |
syntax: "*";
|
| 1039 |
inherits: false;
|
|
|
|
| 1075 |
--tw-ring-offset-width: 0px;
|
| 1076 |
--tw-ring-offset-color: #fff;
|
| 1077 |
--tw-ring-offset-shadow: 0 0 #0000;
|
| 1078 |
+
--tw-blur: initial;
|
| 1079 |
+
--tw-brightness: initial;
|
| 1080 |
+
--tw-contrast: initial;
|
| 1081 |
+
--tw-grayscale: initial;
|
| 1082 |
+
--tw-hue-rotate: initial;
|
| 1083 |
+
--tw-invert: initial;
|
| 1084 |
+
--tw-opacity: initial;
|
| 1085 |
+
--tw-saturate: initial;
|
| 1086 |
+
--tw-sepia: initial;
|
| 1087 |
+
--tw-drop-shadow: initial;
|
| 1088 |
+
--tw-drop-shadow-color: initial;
|
| 1089 |
+
--tw-drop-shadow-alpha: 100%;
|
| 1090 |
+
--tw-drop-shadow-size: initial;
|
| 1091 |
--tw-duration: initial;
|
| 1092 |
}
|
| 1093 |
}
|
app/templates/index.html
CHANGED
|
@@ -5,7 +5,7 @@
|
|
| 5 |
<meta charset="UTF-8">
|
| 6 |
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
| 7 |
<title>Focal</title>
|
| 8 |
-
<link rel="stylesheet" type="text/css" href="static/style.css">
|
| 9 |
<style>
|
| 10 |
.collapsible-content {
|
| 11 |
max-height: 0;
|
|
@@ -123,6 +123,9 @@
|
|
| 123 |
<hr class="h-px my-8 bg-gray-200 border-0 dark:bg-gray-700">
|
| 124 |
<span class="text-sm text-gray-500">{{ article.date }}</span>
|
| 125 |
<h2 class="mb-4 mt-4 text-xl font-semibold tracking-tight text-gray-900">{{ article.title }}</h2>
|
|
|
|
|
|
|
|
|
|
| 126 |
<p class="text-base mb-4">
|
| 127 |
{{ article.content }}
|
| 128 |
</p>
|
|
|
|
| 5 |
<meta charset="UTF-8">
|
| 6 |
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
| 7 |
<title>Focal</title>
|
| 8 |
+
<link rel="stylesheet" type="text/css" href="../static/style.css">
|
| 9 |
<style>
|
| 10 |
.collapsible-content {
|
| 11 |
max-height: 0;
|
|
|
|
| 123 |
<hr class="h-px my-8 bg-gray-200 border-0 dark:bg-gray-700">
|
| 124 |
<span class="text-sm text-gray-500">{{ article.date }}</span>
|
| 125 |
<h2 class="mb-4 mt-4 text-xl font-semibold tracking-tight text-gray-900">{{ article.title }}</h2>
|
| 126 |
+
{% if article.image_url %}
|
| 127 |
+
<img class="w-full h-48 object-cover rounded-lg mb-4 mt-4" src="{{ article.image_url }}" />
|
| 128 |
+
{% endif %}
|
| 129 |
<p class="text-base mb-4">
|
| 130 |
{{ article.content }}
|
| 131 |
</p>
|