File size: 5,081 Bytes
f0c7f30
 
4d2d919
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4d2d919
 
 
 
 
 
 
f0c7f30
 
 
 
 
3aba82e
 
f0c7f30
a67a3ad
 
f0c7f30
3aba82e
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3aba82e
 
 
 
f0c7f30
 
3aba82e
 
 
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a67a3ad
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
3aba82e
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a67a3ad
f0c7f30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d2071ed
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
import base64
import os
import shutil
from collections import defaultdict
from datetime import date, datetime, timedelta
from io import BytesIO

import dotenv
from datasets import load_dataset
from dateutil.parser import parse
from dateutil.tz import tzutc
from fasthtml.common import *
from huggingface_hub import login, whoami

dotenv.load_dotenv()

style = Style("""
                .grid { margin-bottom: 1rem; }
                .card { display: flex; flex-direction: column; }
                .card img { margin-bottom: 0.5rem; }
                .card h5 { margin: 0; font-size: 0.9rem; line-height: 1.2; }
                .card a { color: inherit; text-decoration: none; }
                .card a:hover { text-decoration: underline; }
            """)

# delete data folder
if os.path.exists("data"):
    try:
        shutil.rmtree("data")
    except OSError as e:
        print("Error: %s : %s" % ("data", e.strerror))

app, rt = fast_app(html_style=(style,))

login(token=os.environ.get("HF_TOKEN"))

hf_user = whoami(os.environ.get("HF_TOKEN"))["name"]
HF_REPO_ID_TXT = f"{hf_user}/zotero-answer-ai-texts"
HF_REPO_ID_IMG = f"{hf_user}/zotero-answer-ai-images"

abstract_ds = load_dataset(HF_REPO_ID_TXT, "abstracts", split="train")
article_ds = load_dataset(HF_REPO_ID_TXT, "articles", split="train")

image_ds = load_dataset(HF_REPO_ID_IMG, "images_first_page", split="train")


def parse_date(date_string):
    try:
        return parse(date_string).astimezone(tzutc()).date()
    except ValueError:
        return date.today()


def get_week_start(date_obj):
    return date_obj - timedelta(days=date_obj.weekday())


week2articles = defaultdict(list)
for article in article_ds:
    date_added = parse_date(article["date_added"])
    week_start = get_week_start(date_added)
    week2articles[week_start].append(article["arxiv_id"])

weeks = sorted(week2articles.keys(), reverse=True)

arxiv2article = {article["arxiv_id"]: article for article in article_ds}
arxiv2abstract = {abstract["arxiv_id"]: abstract for abstract in abstract_ds}
arxiv2image = {image["arxiv_id"]: image for image in image_ds}


def get_article_details(arxiv_id):
    article = arxiv2article.get(arxiv_id, {})
    abstract = arxiv2abstract.get(arxiv_id, {})
    image = arxiv2image.get(arxiv_id, {})
    return article, abstract, image


def generate_week_content(current_week):
    week_index = weeks.index(current_week)
    prev_week = weeks[week_index + 1] if week_index < len(weeks) - 1 else None
    next_week = weeks[week_index - 1] if week_index > 0 else None

    nav_buttons = Group(
        Button(
            "← Previous Week",
            hx_get=f"/week/{prev_week}" if prev_week else "#",
            hx_target="#content",
            hx_swap="innerHTML",
            disabled=not prev_week,
        ),
        Button(
            "Next Week β†’",
            hx_get=f"/week/{next_week}" if next_week else "#",
            hx_target="#content",
            hx_swap="innerHTML",
            disabled=not next_week,
        ),
    )

    articles = week2articles[current_week]
    article_cards = []
    for arxiv_id in articles:
        article, abstract, image = get_article_details(arxiv_id)
        article_title = article["contents"][0].get("paper_title", "article") if article["contents"] else "article"

        card_content = [
            H5(
                A(
                    article_title,
                    href=f"https://arxiv.org/abs/{arxiv_id}",
                    target="_blank",
                )
            )
        ]

        if image:
            pil_image = image["image"]  # image[0]["image"]
            img_byte_arr = BytesIO()
            pil_image.save(img_byte_arr, format="JPEG")
            img_byte_arr = img_byte_arr.getvalue()
            image_url = f"data:image/jpeg;base64,{base64.b64encode(img_byte_arr).decode('utf-8')}"
            card_content.insert(
                0,
                Img(
                    src=image_url,
                    alt="Article image",
                    style="max-width: 100%; height: auto; margin-bottom: 15px;",
                ),
            )

        article_cards.append(Card(*card_content, cls="mb-4"))

    grid = Grid(
        *article_cards,
        style="display: grid; grid-template-columns: repeat(3, 1fr); gap: 1rem;",
    )

    week_end = current_week + timedelta(days=6)
    return Div(
        nav_buttons,
        H3(f"Week of {current_week.strftime('%B %d')} - {week_end.strftime('%B %d, %Y')} ({len(articles)} articles)"),
        grid,
        nav_buttons,
        id="content",
    )


@rt("/")
def get():
    return Titled("AnswerAI Zotero Weekly", generate_week_content(weeks[0]))


@rt("/week/{date}")
def get(date: str):
    try:
        current_week = datetime.strptime(date, "%Y-%m-%d").date()
        return generate_week_content(current_week)
    except Exception as e:
        return Div(f"Error displaying articles: {str(e)}")


# serve()

if __name__ == "__main__":
    import uvicorn

    uvicorn.run(app, host="0.0.0.0", port=int(os.environ.get("PORT", 7860)))