Comments (8)
liquidcarbon
commented on June 9, 2024
Code:
@app.get("/sql/{base64query}", response_class=HTMLResponse)
def query_db(base64query: str|None = None):
decoded = Q.from_base64(base64query)
q = Q(decoded.format(**TABLES)) # this is to query tables defined elsewhere using string formatting
log.debug(q)
try:
if not q.is_safe:
raise Exception("naughty!")
df = q.df()
except Exception as e:
df = pd.DataFrame({"error": [e.__str__()]})
with gr.Blocks() as gradio_sql_interface:
sql_code = gr.Code(value=decoded, language="sql", label="SQL Query", interactive=True)
button = gr.Button("run") #, link=f"/sql/{Q(sql_code.value).base64}")
button.click(None, [], [], js=f"window.open('/sql/{Q(sql_code.value).base64}', '_top')")
gr.mount_gradio_app(app, gradio_sql_interface, path="/sqlgradio")
query_result_html = render(...) # jinja template here with all the jQuery stuff
html = f"""
<head>
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" integrity="sha384-4bw+/aepP/YC94hEpVNVgiZdgIC5+VKNBQNGCHeKRQN+PtmoHDEXuppvnDJzQIu9" crossorigin="anonymous">
</head>
<div class="container-fluid py-1 bg-light-subtle">
<h3>QPRST!</h3>
<div class="row m-3 bg-light-subtle">
<div class="col-4">
<iframe id="sql_iframe" src="/sqlgradio" width="100%" height="100%"></iframe>
</div>
<div class="col-8">
{query_result_html}
</div>
</div>
</div>
"""
return html
where Q is just a string with extra attributes:
import base64
import duckdb
class Q(str):
UNSAFE = ["CREATE", "DELETE", "DROP", "INSERT", "UPDATE"]
def __init__(self, query: str):
self.is_safe = not any([cmd in query.upper() for cmd in self.UNSAFE])
def run(self, sql_engine=None):
try:
if sql_engine is None:
return self.run_duckdb()
else:
return self.run_sql(sql_engine)
except Exception as e:
pass
def run_duckdb(self):
return duckdb.sql(self)
def df(self, sql_engine=None):
result = self.run(sql_engine=sql_engine)
if result is None: return
result_df = result.df()
return result_df
@property
def base64(self):
return base64.b64encode(self.encode()).decode()
@classmethod
def from_base64(cls, b64):
"""Initializing from base64-encoded URL paths."""
return cls(base64.b64decode(b64).decode())
from gradio.
abidlabs
commented on June 9, 2024
Hi @liquidcarbon, for general questions like this (that are not feature requests or bug reports), please ask in our Discord server where we have a larger community that can help answer these questions. (I'll close this issue)
from gradio.
liquidcarbon
commented on June 9, 2024
@abidlabs thanks for your message. I saw that for general questions you're invited to go to Discord but I thought this could belong under new guide or feature request.
There is not a lot of examples in the docs on solutions related to FastAPI and gradio. I'd be happy to contribute one if you find this use case interesting, but I need some help :)
Personally, I'm excited about gradio because it becomes much easier to integrate nascent ML projects into existing FastAPI infra. It is my impression that though gradio mainly serves ML UI use cases, it is underappreciated as a general purpose tool for building interesting apps with (almost) pure python. I think FastAPI backend is a huge advantage of gradio compared to, say, Streamlit (Tornado and no plans for supporting others per streamlit/streamlit#4567).
from gradio.
abidlabs
commented on June 9, 2024
For sure, thanks @liquidcarbon! I think this is possible with Gradio currently, so it doesn't need to be feature request
from gradio.
liquidcarbon
commented on June 9, 2024
Perhaps a guide request then?
from gradio.
abidlabs
commented on June 9, 2024
Its a bit niche as a Guide, instead I would publish your app as a public Space on Hugging Face (which supports FastAPI apps) and then it would get good visibility and serve as an informal guide to other users.
from gradio.
liquidcarbon
commented on June 9, 2024
ok I'm mostly there https://huggingface.co/spaces/liquidcarbon/duckdb-fastapi-gradio
I almost gave up, finally ran into #7464 (comment) that shows how the same function can process both inputs from blocks and from requests
import gradio as gr
import pandas as pd
from fastapi import FastAPI
from fastapi.responses import HTMLResponse, RedirectResponse
from itables import to_html_datatable
from sql import Q
app = FastAPI()
def query_from_request(query, request: gr.Request):
"""Process query from input block or from initial request.
https://github.com/gradio-app/gradio/issues/7464#issuecomment-1960161591
"""
if not query:
query_params = request.query_params
base64query = dict(query_params).get("q")
else:
base64query = Q(query).base64
if base64query in (None, "example"):
decoded = Q("""SELECT 42 AS answer, 'LU & E' AS question""")
base64query = decoded.base64
else:
decoded = Q.from_base64(base64query)
href = format_href(base64query)
result = f"""<iframe src="/q/{base64query}" width="90%" height="90%"></iframe>"""
return (decoded, href, result)
def format_href(url: str):
href=f"localhost:7860/sql/{url}"
return f"""<a href="{href}">{href}</a>"""
@app.get("/q/{base64query}", response_class=HTMLResponse)
def query_db(base64query: str|None = None):
decoded = Q.from_base64(base64query)
df = decoded.df()
html = to_html_datatable(df)
return f"""
<h3>{decoded}</h3>
<div>{html}</div>
"""
with gr.Blocks() as gradio_sql_interface:
with gr.Row():
with gr.Column():
header = gr.Markdown("# SQL Editor")
sql_code = gr.Code(language="sql", label="SQL Query", interactive=True)
copy_button = gr.HTML()
button = gr.Button("run")
with gr.Column():
markdown = gr.Markdown("# RESULTS")
results = gr.HTML()
button.click(query_from_request, [sql_code], [sql_code, copy_button, results])
gradio_sql_interface.load(query_from_request, [sql_code], [sql_code, copy_button, results], queue=True)
app = gr.mount_gradio_app(app, gradio_sql_interface, path="/sql")
@app.get("/")
@app.get("/sql")
@app.get("/sql/")
def redirect_to_example():
return RedirectResponse("/sql/?q=example")
from gradio.
abidlabs
commented on June 9, 2024
Awesome so is everything working now or did you run into any bugs / issues?
from gradio.
Related Issues (20)
- Z-index issue with loading bar
- AttributeError: 'Image' object has no attribute 'value' HOT 2
- Streaming complex object fast in custom component makes bug HOT 5
- Custom values are not supported when 'multiselect' is True ? HOT 1
- Speech to text output text box issue HOT 2
- Unable to run custom component dev server based on DataFrame template HOT 3
- `gr.load()` does not work inside `gr.Blocks` HOT 2
- Return of older bug. Cannot update Dataset
- ConnectionResetError: [WinError 10054] when click radio to switch HOT 3
- Drag-and-Drop Image doesn't work inside of a `gr.Tab` HOT 1
- Gradio has a conflict with Python file cp1252 HOT 3
- Gradio-Lite : zero-shot-classification pipeline returns only 1 score HOT 2
- Be able to generate random samples in `gr.Examples` or `examples` in `gr.ChatInterface` HOT 2
- `Edge` runtime compatable javascript client or REST API docs HOT 1
- cannot import name 'JsonValue' from 'pydantic' HOT 5
- Clarity needed for cli tool \ custom component development HOT 6
- how can i make button upper than the component it controls. HOT 3
- How to access local media (image/audio/video) within gradio-lite? HOT 5
- How to support carriage return sending in the textbox component
- How to support carriage return sending in the textbox component
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from gradio.