Spaces:

DarylT01
/

dfa-ui

Runtime error

dfa-ui / app.py

feat(type): identify word type

15e4f2e over 2 years ago

7.81 kB

# Import libraries and modules
# Run pip install gradio and essential_generators in terminal
import gradio as gr
import pandas as pd
from essential_generators import DocumentGenerator
from main import generate_dfa

# DataFrame
conjunctions = []
with open("conjunctions.txt") as file:
    conjunctions = file.read().splitlines()
    conjunctions = [word.strip() for word in conjunctions]
with open("adverbs.txt") as file:
    adverbs = file.read().splitlines()
    adverbs = [word.strip() for word in adverbs]
with open("adjectives.txt") as file:
    adjectives = file.read().splitlines()
    adjectives = [word.strip() for word in adjectives]

df = pd.DataFrame(columns=["Words", "Type"])
for i in conjunctions:
    df.loc[len(df)] = [i, "Conjunction"]
for i in adverbs:
    df.loc[len(df)] = [i, "Adverb"]
for i in adjectives:
    df.loc[len(df)] = [i, "Adjective"]

# DFA function call
dfa_conjunctions = generate_dfa(conjunctions)
dfa_adverbs = generate_dfa(adverbs)
dfa_adjectives = generate_dfa(adjectives)


# Generate examples
def generateExamples():
    gen = DocumentGenerator()
    examples = []
    for i in range(3):
        examples.append(gen.paragraph())
    return examples


# Color match function
def color_match(text: gr.Textbox):
    colored_text = []
    pointer = 0

    # Get the result of the DFA check on the input text
    match_dict_conj = dfa_conjunctions.check(text)
    match_dict_adv = dfa_adverbs.check(text)
    match_dict_adj = dfa_adjectives.check(text)

    if not match_dict_conj and not match_dict_adv and not match_dict_adj:
        return (
            '<div style="background-color: #dc2626; color: #fff; text-align: center; width: 100%; padding: 10px; font-weight:800; font-size:1.5rem">Rejected</div>',
            None,
            None,
        )

    # Flatten the match_dict into a list of tuples and sort by the start index
    matches = sorted(
        [
            (start, end, word)
            for word, indices in match_dict_conj.items()
            for start, end in indices
        ]
        + [
            (start, end, word)
            for word, indices in match_dict_adv.items()
            for start, end in indices
        ]
        + [
            (start, end, word)
            for word, indices in match_dict_adj.items()
            for start, end in indices
        ]
    )

    for start, end, word in matches:
        colored_text.append(text[pointer:start])
        # End need to be incremented by 1 to include the last character
        colored_text.append(f"<span style='color:#4ade80'>{text[start:end + 1]}</span>")
        # Move the pointer to the end of the match
        pointer = end + 1

    # Add remaining text
    colored_text.append(text[pointer:])
    # Combine the strings
    colored_text = "".join(colored_text)

    # Create the DataFrame
    positions_df = pd.DataFrame(columns=["Words", "Type", "Positions", "Occurences"])
    for word, positions in match_dict_conj.items():
        # Convert the list of tuples to a string
        positions_str = ", ".join(map(str, positions))
        # Store the word and the positions string in the wordPositions dictionary
        positions_df.loc[len(positions_df)] = [
            word,
            "Conjunction",
            positions_str,
            len(positions),
        ]
    for word, positions in match_dict_adv.items():
        # Convert the list of tuples to a string
        positions_str = ", ".join(map(str, positions))
        # Store the word and the positions string in the wordPositions dictionary
        positions_df.loc[len(positions_df)] = [
            word,
            "Adverb",
            positions_str,
            len(positions),
        ]
    for word, positions in match_dict_adj.items():
        # Convert the list of tuples to a string
        positions_str = ", ".join(map(str, positions))
        # Store the word and the positions string in the wordPositions dictionary
        positions_df.loc[len(positions_df)] = [
            word,
            "Adjective",
            positions_str,
            len(positions),
        ]

    return colored_text, positions_df


# Search and display function
def search_and_display(search_query):
    # Filter the DataFrame based on the search query
    filtered_df = df[df["Words"].str.contains(search_query)]
    return filtered_df


def text_change_search(text: gr.Textbox):
    if text == "":
        return gr.update(interactive=False)
    else:
        return gr.update(interactive=True)


def text_change_test(text: gr.Textbox):
    if text == "":
        return gr.update(interactive=False), gr.update(interactive=False)
    else:
        return gr.update(interactive=True), gr.update(interactive=True)


def remove_output(result, position):
    return None, None


# CSS styling
# css = """
# warning {background-color: #FFCCCB}
# .feedback textarea {font-size: 24px !important}
# """

# Example to apply CSS styling
# with gr.Blocks(css=css) as demo:
#     box1 = gr.Textbox(value="Good Job", elem_classes="feedback")
#     box2 = gr.Textbox(value="Failure", elem_id="warning", elem_classes="feedback")

# Gradio UI
with gr.Blocks() as demo:
    # Title block
    # Apply CSS styling to the title
    title = gr.HTML(
        "<h1 style='color: #2563eb; font-weight:bold'>English Conjunctions/Adverb/Adjectives Finder</h1>"
    )
    with gr.Accordion("Accepted Words", open=False):
        # Search block
        search = gr.Textbox(
            label="Search",
            placeholder="Search accepted words here",
            lines=1,
            info="List of acceptable words in DFA",
            show_copy_button=True,
        )
        with gr.Row():
            cancel_btn = gr.ClearButton(search, variant="stop", interactive=False)
            search_btn = gr.Button(value="Search", variant="primary")
        resultSearch = gr.Dataframe(
            df, height=300, col_count=2, headers=["Words", "Type"]
        )

        search.change(
            text_change_search,
            inputs=[search],
            outputs=[cancel_btn],
        )

        search_btn.click(
            search_and_display, inputs=[search], outputs=[resultSearch], api_name=False
        )

    # Text block for DFA and color match
    textTitle = gr.HTML("<h2 style='color: #2563eb;'>Try it here!</h2>")
    description = gr.HTML(
        "<p style='color: #a78bfa;'>Enter a text and see the words that are accepted by the DFA highlighted in <span style='color:#4ade80'>green</span>.</p>"
    )
    text = gr.Textbox(
        autofocus=True,
        label="Text",
        placeholder="Enter text here",
        info="Enter text to check for DFA match",
        show_copy_button=True,
    )

    # Examples block
    examples_data = generateExamples()
    examples = gr.Examples(
        examples=examples_data,
        inputs=[text],
    )
    with gr.Row():
        cancel_btn = gr.ClearButton([text], variant="stop", interactive=False)
        submit_btn = gr.Button(value="Submit", variant="primary", interactive=False)

    text.change(
        text_change_test,
        inputs=[text],
        outputs=[cancel_btn, submit_btn],
    )

    # Result block
    resultTitle = gr.HTML("<h2 style='color: #2563eb;'>Result</h2>")
    result = gr.HTML("<p></p>")

    # Position block
    # positionTitle = gr.HTML("<h2 style='color: gold;'>Position</h2>")
    position = gr.Dataframe(
        show_label=True,
        col_count=4,
        headers=["Words", "Type", "Positions", "Occurences"],
        interactive=False,
    )

    submit_btn.click(
        color_match,
        inputs=[text],
        outputs=[result, position],
        api_name=False,
    )
    cancel_btn.click(
        remove_output,
        inputs=[result, position],
        outputs=[result, position],
    )

# Launch the app
demo.launch()

	# Import libraries and modules
	# Run pip install gradio and essential_generators in terminal
	import gradio as gr
	import pandas as pd
	from essential_generators import DocumentGenerator
	from main import generate_dfa

	# DataFrame
	conjunctions = []
	with open("conjunctions.txt") as file:
	conjunctions = file.read().splitlines()
	conjunctions = [word.strip() for word in conjunctions]
	with open("adverbs.txt") as file:
	adverbs = file.read().splitlines()
	adverbs = [word.strip() for word in adverbs]
	with open("adjectives.txt") as file:
	adjectives = file.read().splitlines()
	adjectives = [word.strip() for word in adjectives]

	df = pd.DataFrame(columns=["Words", "Type"])
	for i in conjunctions:
	df.loc[len(df)] = [i, "Conjunction"]
	for i in adverbs:
	df.loc[len(df)] = [i, "Adverb"]
	for i in adjectives:
	df.loc[len(df)] = [i, "Adjective"]

	# DFA function call
	dfa_conjunctions = generate_dfa(conjunctions)
	dfa_adverbs = generate_dfa(adverbs)
	dfa_adjectives = generate_dfa(adjectives)


	# Generate examples
	def generateExamples():
	gen = DocumentGenerator()
	examples = []
	for i in range(3):
	examples.append(gen.paragraph())
	return examples


	# Color match function
	def color_match(text: gr.Textbox):
	colored_text = []
	pointer = 0

	# Get the result of the DFA check on the input text
	match_dict_conj = dfa_conjunctions.check(text)
	match_dict_adv = dfa_adverbs.check(text)
	match_dict_adj = dfa_adjectives.check(text)

	if not match_dict_conj and not match_dict_adv and not match_dict_adj:
	return (
	'<div style="background-color: #dc2626; color: #fff; text-align: center; width: 100%; padding: 10px; font-weight:800; font-size:1.5rem">Rejected</div>',
	None,
	None,
	)

	# Flatten the match_dict into a list of tuples and sort by the start index
	matches = sorted(
	[
	(start, end, word)
	for word, indices in match_dict_conj.items()
	for start, end in indices
	]
	+ [
	(start, end, word)
	for word, indices in match_dict_adv.items()
	for start, end in indices
	]
	+ [
	(start, end, word)
	for word, indices in match_dict_adj.items()
	for start, end in indices
	]
	)

	for start, end, word in matches:
	colored_text.append(text[pointer:start])
	# End need to be incremented by 1 to include the last character
	colored_text.append(f"<span style='color:#4ade80'>{text[start:end + 1]}</span>")
	# Move the pointer to the end of the match
	pointer = end + 1

	# Add remaining text
	colored_text.append(text[pointer:])
	# Combine the strings
	colored_text = "".join(colored_text)

	# Create the DataFrame
	positions_df = pd.DataFrame(columns=["Words", "Type", "Positions", "Occurences"])
	for word, positions in match_dict_conj.items():
	# Convert the list of tuples to a string
	positions_str = ", ".join(map(str, positions))
	# Store the word and the positions string in the wordPositions dictionary
	positions_df.loc[len(positions_df)] = [
	word,
	"Conjunction",
	positions_str,
	len(positions),
	]
	for word, positions in match_dict_adv.items():
	# Convert the list of tuples to a string
	positions_str = ", ".join(map(str, positions))
	# Store the word and the positions string in the wordPositions dictionary
	positions_df.loc[len(positions_df)] = [
	word,
	"Adverb",
	positions_str,
	len(positions),
	]
	for word, positions in match_dict_adj.items():
	# Convert the list of tuples to a string
	positions_str = ", ".join(map(str, positions))
	# Store the word and the positions string in the wordPositions dictionary
	positions_df.loc[len(positions_df)] = [
	word,
	"Adjective",
	positions_str,
	len(positions),
	]

	return colored_text, positions_df


	# Search and display function
	def search_and_display(search_query):
	# Filter the DataFrame based on the search query
	filtered_df = df[df["Words"].str.contains(search_query)]
	return filtered_df


	def text_change_search(text: gr.Textbox):
	if text == "":
	return gr.update(interactive=False)
	else:
	return gr.update(interactive=True)


	def text_change_test(text: gr.Textbox):
	if text == "":
	return gr.update(interactive=False), gr.update(interactive=False)
	else:
	return gr.update(interactive=True), gr.update(interactive=True)


	def remove_output(result, position):
	return None, None


	# CSS styling
	# css = """
	# warning {background-color: #FFCCCB}
	# .feedback textarea {font-size: 24px !important}
	# """

	# Example to apply CSS styling
	# with gr.Blocks(css=css) as demo:
	# box1 = gr.Textbox(value="Good Job", elem_classes="feedback")
	# box2 = gr.Textbox(value="Failure", elem_id="warning", elem_classes="feedback")

	# Gradio UI
	with gr.Blocks() as demo:
	# Title block
	# Apply CSS styling to the title
	title = gr.HTML(
	"<h1 style='color: #2563eb; font-weight:bold'>English Conjunctions/Adverb/Adjectives Finder</h1>"
	)
	with gr.Accordion("Accepted Words", open=False):
	# Search block
	search = gr.Textbox(
	label="Search",
	placeholder="Search accepted words here",
	lines=1,
	info="List of acceptable words in DFA",
	show_copy_button=True,
	)
	with gr.Row():
	cancel_btn = gr.ClearButton(search, variant="stop", interactive=False)
	search_btn = gr.Button(value="Search", variant="primary")
	resultSearch = gr.Dataframe(
	df, height=300, col_count=2, headers=["Words", "Type"]
	)

	search.change(
	text_change_search,
	inputs=[search],
	outputs=[cancel_btn],
	)

	search_btn.click(
	search_and_display, inputs=[search], outputs=[resultSearch], api_name=False
	)

	# Text block for DFA and color match
	textTitle = gr.HTML("<h2 style='color: #2563eb;'>Try it here!</h2>")
	description = gr.HTML(
	"<p style='color: #a78bfa;'>Enter a text and see the words that are accepted by the DFA highlighted in <span style='color:#4ade80'>green</span>.</p>"
	)
	text = gr.Textbox(
	autofocus=True,
	label="Text",
	placeholder="Enter text here",
	info="Enter text to check for DFA match",
	show_copy_button=True,
	)

	# Examples block
	examples_data = generateExamples()
	examples = gr.Examples(
	examples=examples_data,
	inputs=[text],
	)
	with gr.Row():
	cancel_btn = gr.ClearButton([text], variant="stop", interactive=False)
	submit_btn = gr.Button(value="Submit", variant="primary", interactive=False)

	text.change(
	text_change_test,
	inputs=[text],
	outputs=[cancel_btn, submit_btn],
	)

	# Result block
	resultTitle = gr.HTML("<h2 style='color: #2563eb;'>Result</h2>")
	result = gr.HTML("<p></p>")

	# Position block
	# positionTitle = gr.HTML("<h2 style='color: gold;'>Position</h2>")
	position = gr.Dataframe(
	show_label=True,
	col_count=4,
	headers=["Words", "Type", "Positions", "Occurences"],
	interactive=False,
	)

	submit_btn.click(
	color_match,
	inputs=[text],
	outputs=[result, position],
	api_name=False,
	)
	cancel_btn.click(
	remove_output,
	inputs=[result, position],
	outputs=[result, position],
	)

	# Launch the app
	demo.launch()