Spaces:
Sleeping
Sleeping
agharsallah commited on
Commit ·
f928c5e
1
Parent(s): f86da75
addding simple llama parse tool
Browse files- .env.example +1 -0
- app.py +31 -17
- requirements.txt +1 -0
- tools.py +51 -0
.env.example
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
LLAMA_CLOUD_API_KEY= your_api_key_here
|
app.py
CHANGED
|
@@ -1,26 +1,40 @@
|
|
| 1 |
import gradio as gr
|
|
|
|
|
|
|
| 2 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
|
| 4 |
-
def letter_counter(word, letter):
|
| 5 |
-
"""
|
| 6 |
-
Count the occurrences of a specific letter in a word.
|
| 7 |
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
-
Returns:
|
| 13 |
-
The number of times the letter appears in the word
|
| 14 |
-
"""
|
| 15 |
-
return word.lower().count(letter.lower())
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
outputs="number",
|
| 22 |
-
title="Letter Counter",
|
| 23 |
-
description="Count how many times a letter appears in a word",
|
| 24 |
)
|
| 25 |
|
| 26 |
-
|
|
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
+
from tools import letter_counter, pdf_to_text
|
| 3 |
+
import os
|
| 4 |
|
| 5 |
+
# --- Gradio Tool: Letter Counter ---
|
| 6 |
+
letter_counter_tool = gr.Interface(
|
| 7 |
+
fn=letter_counter,
|
| 8 |
+
inputs=[gr.Textbox(label="Word or Phrase"), gr.Textbox(label="Letter")],
|
| 9 |
+
outputs=gr.Number(label="Count"),
|
| 10 |
+
title="Letter Counter",
|
| 11 |
+
description="Count how many times a letter appears in a word or phrase.",
|
| 12 |
+
)
|
| 13 |
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
+
# --- Gradio Tool: PDF to Text ---
|
| 16 |
+
def gradio_pdf_to_text(pdf_file, api_key=None):
|
| 17 |
+
if api_key is None:
|
| 18 |
+
api_key = os.getenv("LLAMA_CLOUD_API_KEY")
|
| 19 |
+
if pdf_file is None:
|
| 20 |
+
return "No PDF file provided."
|
| 21 |
+
return pdf_to_text(pdf_file.name, api_key=api_key)
|
| 22 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
+
pdf_to_text_tool = gr.Interface(
|
| 25 |
+
fn=gradio_pdf_to_text,
|
| 26 |
+
inputs=[
|
| 27 |
+
gr.File(label="PDF File"),
|
| 28 |
+
],
|
| 29 |
+
outputs=gr.Textbox(label="Extracted Text"),
|
| 30 |
+
title="PDF to Text Extractor",
|
| 31 |
+
description="Extracts all text from a PDF using LlamaParse. Optionally provide your Llama Cloud API key.",
|
| 32 |
+
)
|
| 33 |
|
| 34 |
+
# --- Combine Tools in Gradio Tabbed UI ---
|
| 35 |
+
demo = gr.TabbedInterface(
|
| 36 |
+
[letter_counter_tool, pdf_to_text_tool], ["Letter Counter", "PDF to Text"]
|
|
|
|
|
|
|
|
|
|
| 37 |
)
|
| 38 |
|
| 39 |
+
if __name__ == "__main__":
|
| 40 |
+
demo.launch(mcp_server=True)
|
requirements.txt
CHANGED
|
@@ -1 +1,2 @@
|
|
| 1 |
gradio[mcp]
|
|
|
|
|
|
| 1 |
gradio[mcp]
|
| 2 |
+
llama-cloud-services
|
tools.py
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
"""
|
| 2 |
+
tools.py
|
| 3 |
+
|
| 4 |
+
This module contains modular tools for the Gradio app, including:
|
| 5 |
+
- Letter Counter Tool
|
| 6 |
+
- PDF to Text Tool (using LlamaParse)
|
| 7 |
+
"""
|
| 8 |
+
|
| 9 |
+
from llama_cloud_services import LlamaParse
|
| 10 |
+
import os
|
| 11 |
+
|
| 12 |
+
|
| 13 |
+
# --- Tool: Letter Counter ---
|
| 14 |
+
def letter_counter(word: str, letter: str) -> int:
|
| 15 |
+
"""
|
| 16 |
+
Count the occurrences of a specific letter in a word.
|
| 17 |
+
|
| 18 |
+
Args:
|
| 19 |
+
word (str): The word or phrase to analyze.
|
| 20 |
+
letter (str): The letter to count occurrences of.
|
| 21 |
+
|
| 22 |
+
Returns:
|
| 23 |
+
int: The number of times the letter appears in the word.
|
| 24 |
+
"""
|
| 25 |
+
return word.lower().count(letter.lower())
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
# --- Tool: PDF to Text ---
|
| 29 |
+
def pdf_to_text(pdf_path: str) -> str | None:
|
| 30 |
+
"""
|
| 31 |
+
Extracts text from a PDF file using LlamaParse.
|
| 32 |
+
|
| 33 |
+
Args:
|
| 34 |
+
pdf_path (str): Path to the PDF file.
|
| 35 |
+
|
| 36 |
+
Returns:
|
| 37 |
+
str: The extracted text from the PDF.
|
| 38 |
+
None: If no text could be extracted.
|
| 39 |
+
"""
|
| 40 |
+
parser = LlamaParse(
|
| 41 |
+
api_key=os.getenv("LLAMA_CLOUD_API_KEY"),
|
| 42 |
+
num_workers=1,
|
| 43 |
+
verbose=True,
|
| 44 |
+
language="en",
|
| 45 |
+
)
|
| 46 |
+
result = parser.parse(pdf_path)
|
| 47 |
+
# Get all text as a single string
|
| 48 |
+
text_documents = result.get_text_documents(split_by_page=False)
|
| 49 |
+
if text_documents:
|
| 50 |
+
return "\n".join([doc.text for doc in text_documents])
|
| 51 |
+
return None
|