File size: 7,034 Bytes
5ad8cbc
 
 
 
 
 
acfde36
5ad8cbc
 
 
 
 
 
 
 
 
 
 
 
 
acfde36
5ad8cbc
 
 
 
 
 
 
e52f41a
 
5ad8cbc
 
1f871fc
5ad8cbc
 
 
 
 
3cc8ac8
 
5ad8cbc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f1ef237
5ad8cbc
 
 
 
 
 
 
 
 
b005dc6
5ad8cbc
b005dc6
52fdc4e
 
5ad8cbc
 
acfde36
5ad8cbc
 
 
 
 
 
 
 
 
 
 
 
a7c0e1b
5ad8cbc
 
 
956ed72
 
c8821e9
5ad8cbc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d6e3fde
 
 
 
 
 
5ad8cbc
 
 
 
 
 
 
 
 
 
 
f602886
 
2cddf47
5cbff0e
2cddf47
 
 
96e61f3
8c3c798
5ad8cbc
 
 
2fa4815
5ad8cbc
2fa4815
5ad8cbc
 
 
 
 
be6e01f
5ad8cbc
 
 
 
 
 
 
 
be6e01f
5ad8cbc
 
 
2cddf47
 
 
 
 
220e98c
 
2cddf47
5cbff0e
 
be6e01f
5ad8cbc
 
 
 
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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
import huggingface_hub as hf
import gradio as gr 
import os, datetime
fs = hf.HfFileSystem(token=os.environ["HF_TOKEN"])
datasetdir = "datasets/yoinked/blue-arxiv-papers/"
basecss = """
.caaard-container {
  width: 250px;
  padding: 20px;
  border: 3px solid black;
  border-radius: 15px;
  text-align: left;
}

.title {
  font-size: 24px;
  margin-bottom: 10px;
  text-align: center;
}

.caaard-containers {
  display: flex; gap: 20px; flex-wrap: wrap;
}
.extra-info {
  font-size: 14px;
  line-height: 1.5;
}
.extra-info-paperid {
  font-size: 18px;
  line-height: 1.75;
}"""
def get_papers():
    return reversed(fs.glob(datasetdir+"*.md"))
def get_papers_metadata(papiers=None):
    metadatas = []
    if papiers is None:
        papiers = get_papers()
    for paper in papiers:
        with fs.open(paper, "r") as f:
            papertxt = f.read()
        metadata = papertxt.split("---")[1]
        try:
            author = metadata.split("author: ")[1].split("\n")[0]
        except:
            author = "unknown"
        try:
            title = metadata.split("title: ")[1].split("\n")[0]
        except:
            title = "unknown"
        try:
            tags = metadata.split("tags: ")[1].split("\n")[0].split(", ")
        except:
            tags = []
        try:
            abstract = metadata.split("abstract: ")[1].split("\n")[0]
        except:
            abstract = "unknown"
        try:
            date_published = metadata.split("date_published: ")[1].split("\n")[0]
        except:
            date_published = "unknown"
        try:
            paperid = metadata.split("paperid: ")[1].split("\n")[0] #if this fails then no reason to display
            md = {"fname": paper, "metadata": metadata, "author": author, "title": title, "tags": tags, "abstract": abstract, "date_published": date_published, "paperid": paperid}
            metadatas.append(md)
        except:
            pass
    return metadatas

def make_paper_card(md):
    html = f"""
<div class="caaard-container" title="{md["abstract"]}">
  <div class="title">{md["title"]}</div>
  <br><br>
  <div class="extra-info">author: {md["author"]}</div>
  <div class="extra-info">published: {md["date_published"]}</div>
  <div class="extra-info-paperid">id: {md["paperid"]}</div>
</div>
    """
    return html

def make_paper_cards(tags=""):
    mds = get_papers_metadata()
    if tags != "":
        tags = tags.split(",")
        tags = [tag.strip() for tag in tags]
        mds = [md for md in mds if any(tag in md["tags"] for tag in tags)]
    htmls = [make_paper_card(md) for md in mds]
    fin = "<div class='caaard-containers'>" 
    for html in htmls:
        fin += html + "<br>"
    fin += "</div>"
    return fin

def get_paper_markdown(paperid):
    allpapers = get_papers_metadata()
    fname = None
    for paper in allpapers:
        if paper["paperid"] == paperid:
            fname = paper["fname"]
            break
    print(fname, paperid)
    if fname is None:
        return "## paper not found"
    else:
        with fs.open(fname, "r") as f:
            papertxt = f.read()
        return papertxt.split("---")[2]

def publish_paper(title, authors, tags, abst, data):
    paperid = ""
    year = datetime.datetime.now().year
    month = datetime.datetime.now().month
    if month < 10:
        month = "0"+str(month)
    day = datetime.datetime.now().day
    if day < 10:
        day = "0"+str(day)
    idx = 1
    while True:
        paperid = f"{year}-{month}{day}.{idx}"
        if not fs.exists(datasetdir+paperid+".md"):
            break
        idx += 1
        if idx > 100:
            return "could not generate paperid, try again tomorrow"
    bad_chars = "<>:|\\" # primitive anti-xss sanitization
    for c in bad_chars:
        title = title.replace(c, "")
        authors = authors.replace(c, "")
        tags = tags.replace(c, "")
        abst = abst.replace(c, "")
    metadata = f"""---
title: {title}
author: {authors}
tags: {tags}
abstract: {abst}
date_published: {year}-{month}-{day}
paperid: {paperid}
---\n"""
    with fs.open(datasetdir+paperid+".md", "w") as f:
        raw = metadata + data
        f.write(raw)
def makepreview(x):
    return x
def upload(prefix, fname, ext, file):
    fname = prefix+"-"+fname
    with fs.open(datasetdir+"uploads/"+fname+"."+ext, "wb") as f:
        f.write(file)
    return f"uploaded, use https://huggingface.co/datasets/yoinked/blue-arxiv-papers/resolve/main/uploads/{fname}.{ext} to include in your paper (so like ![image](https://huggingface.co/datasets/yoinked/blue-arxiv-papers/resolve/main/uploads/{fname}.{ext})) for inline img"
with gr.Blocks(css=basecss, theme='NoCrypt/miku') as demo:
    gr.Image("./blue-arxiv.png", container=False, label=None, interactive=False, show_fullscreen_button=False, show_share_button=False, show_download_button=False)
    with gr.Tab("search"):
        with gr.Row():
            query = gr.Textbox(label="tags (optional, comma seperated)", lines=1, interactive=True)
            searchbutton = gr.Button("πŸ”Ž")
        with gr.Row():
            papercards = gr.HTML("Click the πŸ”Ž to load all papers!")
    with gr.Tab("read"):
        with gr.Row():
            paperid = gr.Textbox(label="paper id", lines=1, interactive=True)
            readbutton = gr.Button("read")
        with gr.Row():
            paper = gr.Markdown()
    with gr.Tab("publish"):
        with gr.Row():
            title = gr.Textbox(label="title", lines=1, interactive=True)
            authors = gr.Textbox(label="author(s)", lines=1, interactive=True)
        with gr.Row():
            tags = gr.Textbox(label="tags (optional, comma seperated)", lines=1, interactive=True)
            abst = gr.Textbox(label="abriged abstract (aka tooltip)", lines=2, interactive=True)
        markd = gr.Textbox(label="markdown", lines=10, interactive=True, max_lines=1e3)
        preview = gr.Markdown()
        with gr.Row():
            status = gr.Textbox(label="status", lines=1, interactive=False)
            publishbutton = gr.Button("publish")
    with gr.Tab("files"):
        with gr.Row():
            prefix = gr.Textbox(label="prefix", lines=1, interactive=True)
            file_name = gr.Textbox(label="file name", lines=1, interactive=True)
        with gr.Row():
            file = gr.File(label="file", file_types=[".png", ".gif", ".webp", ".jpg", ".wav", ".mp3"], type="binary")
            fileext = gr.Dropdown(label="filetype", choices=["png", "gif", "webp", "jpg", "wav", "mp3"])
            uploadbutton = gr.Button("upload")
        statii = gr.Textbox(label="status", interactive=False)
    uploadbutton.click(fn=upload, inputs=[prefix, file_name, fileext, file], outputs=statii)
    markd.change(fn=makepreview, inputs=markd, outputs=preview)
    publishbutton.click(fn=publish_paper, inputs=[title, authors, tags, abst, markd], outputs=status)
    searchbutton.click(fn=make_paper_cards, inputs=query, outputs=papercards)
    readbutton.click(fn=get_paper_markdown, inputs=paperid, outputs=paper)
demo.launch()