add changes
Browse files- app.py +131 -0
- data/Dr-En-space-test.csv +0 -0
app.py
ADDED
|
@@ -0,0 +1,131 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import pandas as pd
|
| 2 |
+
import streamlit as st
|
| 3 |
+
import time
|
| 4 |
+
import random
|
| 5 |
+
import os
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
# File Path
|
| 9 |
+
DATA_FILENAME = "Dr-En-space-test.csv"
|
| 10 |
+
DATA_PATH = os.path.join("data", DATA_FILENAME)
|
| 11 |
+
|
| 12 |
+
st.set_page_config(layout="wide")
|
| 13 |
+
|
| 14 |
+
# Load data
|
| 15 |
+
def load_data():
|
| 16 |
+
return pd.read_csv(DATA_PATH)
|
| 17 |
+
|
| 18 |
+
def save_data(data):
|
| 19 |
+
data.to_csv(DATA_PATH, index=False)
|
| 20 |
+
|
| 21 |
+
def skip_correction():
|
| 22 |
+
noncorrected_sentences = st.session_state.data[(st.session_state.data.translated == True) & (st.session_state.data.corrected == False)]['sentence'].tolist()
|
| 23 |
+
if noncorrected_sentences:
|
| 24 |
+
st.session_state.orig_sentence = random.choice(noncorrected_sentences)
|
| 25 |
+
st.session_state.orig_translation = st.session_state.data[st.session_state.data.sentence == st.session_state.orig_sentence]['translation']
|
| 26 |
+
else:
|
| 27 |
+
st.session_state.orig_sentence = "No more sentences to be corrected"
|
| 28 |
+
st.session_state.orig_translation = "No more sentences to be corrected"
|
| 29 |
+
|
| 30 |
+
st.title("Darija Translation Corpus Collection")
|
| 31 |
+
|
| 32 |
+
if "data" not in st.session_state:
|
| 33 |
+
st.session_state.data = load_data()
|
| 34 |
+
|
| 35 |
+
if "sentence" not in st.session_state:
|
| 36 |
+
untranslated_sentences = st.session_state.data[st.session_state.data['translated'] == False]['sentence'].tolist()
|
| 37 |
+
if untranslated_sentences:
|
| 38 |
+
st.session_state.sentence = random.choice(untranslated_sentences)
|
| 39 |
+
else:
|
| 40 |
+
st.session_state.sentence = "No more sentences to translate"
|
| 41 |
+
|
| 42 |
+
if "orig_translation" not in st.session_state:
|
| 43 |
+
noncorrected_sentences = st.session_state.data[(st.session_state.data.translated == True) & (st.session_state.data.corrected == False)]['sentence'].tolist()
|
| 44 |
+
noncorrected_translations = st.session_state.data[(st.session_state.data.translated == True) & (st.session_state.data.corrected == False)]['translation'].tolist()
|
| 45 |
+
|
| 46 |
+
if noncorrected_sentences:
|
| 47 |
+
st.session_state.orig_sentence = random.choice(noncorrected_sentences)
|
| 48 |
+
st.session_state.orig_translation = st.session_state.data.loc[st.session_state.data.sentence == st.session_state.orig_sentence]['translation'].values[0]
|
| 49 |
+
else:
|
| 50 |
+
st.session_state.orig_sentence = "No more sentences to be corrected"
|
| 51 |
+
st.session_state.orig_translation = "No more sentences to be corrected"
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
if "user_translation" not in st.session_state:
|
| 56 |
+
st.session_state.user_translation = ""
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
with st.sidebar:
|
| 60 |
+
st.subheader("About")
|
| 61 |
+
st.markdown("""This is app is designed to collect Darija translation corpus.""")
|
| 62 |
+
|
| 63 |
+
tab1, tab2 = st.tabs(["Translation", "Correction"])
|
| 64 |
+
|
| 65 |
+
with tab1:
|
| 66 |
+
with st.container():
|
| 67 |
+
st.subheader("Original Text:")
|
| 68 |
+
|
| 69 |
+
st.write('<div style="height: 150px; overflow: auto; border: 2px solid #ddd; padding: 10px; border-radius: 5px;">{}</div>'.format(st.session_state.sentence), unsafe_allow_html=True)
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
st.subheader("Translation:")
|
| 73 |
+
st.session_state.user_translation = st.text_area("Enter your translation here:", value=st.session_state.user_translation)
|
| 74 |
+
|
| 75 |
+
if st.button("💾 Save"):
|
| 76 |
+
if st.session_state.user_translation:
|
| 77 |
+
st.session_state.data.loc[st.session_state.data['sentence'] == st.session_state.sentence, 'translation'] = st.session_state.user_translation
|
| 78 |
+
st.session_state.data.loc[st.session_state.data['sentence'] == st.session_state.sentence, 'translated'] = True
|
| 79 |
+
save_data(st.session_state.data)
|
| 80 |
+
|
| 81 |
+
st.session_state.user_translation = "" # Reset the input value after saving
|
| 82 |
+
|
| 83 |
+
# st.toast("Saved!", icon="👏")
|
| 84 |
+
st.success("Saved!")
|
| 85 |
+
|
| 86 |
+
# Update the sentence for the next iteration.
|
| 87 |
+
untranslated_sentences = st.session_state.data[st.session_state.data['translated'] == False]['sentence'].tolist()
|
| 88 |
+
if untranslated_sentences:
|
| 89 |
+
st.session_state.sentence = random.choice(untranslated_sentences)
|
| 90 |
+
|
| 91 |
+
else:
|
| 92 |
+
st.session_state.sentence = "No more sentences to translate"
|
| 93 |
+
|
| 94 |
+
time.sleep(0.5)
|
| 95 |
+
# Rerun the app
|
| 96 |
+
st.rerun()
|
| 97 |
+
|
| 98 |
+
with tab2:
|
| 99 |
+
with st.container():
|
| 100 |
+
st.subheader("Original Darija Text:")
|
| 101 |
+
st.write('<div style="height: 150px; overflow: auto; border: 2px solid #ddd; padding: 10px; border-radius: 5px;">{}</div>'.format(st.session_state.orig_sentence), unsafe_allow_html=True)
|
| 102 |
+
|
| 103 |
+
with st.container():
|
| 104 |
+
st.subheader("Original English Translation:")
|
| 105 |
+
st.write('<div style="height: 150px; overflow: auto; border: 2px solid #ddd; padding: 10px; border-radius: 5px;">{}</div>'.format(st.session_state.orig_translation), unsafe_allow_html=True)
|
| 106 |
+
|
| 107 |
+
st.subheader("Corrected Darija Translation:")
|
| 108 |
+
corrected_translation = st.text_area("Enter the corrected Darija translation here:")
|
| 109 |
+
|
| 110 |
+
if st.button("💾 Save Translation"):
|
| 111 |
+
if corrected_translation:
|
| 112 |
+
st.session_state.data.loc[st.session_state.data['sentence'] == st.session_state.orig_sentence, 'translation'] = corrected_translation
|
| 113 |
+
st.session_state.data.loc[st.session_state.data['sentence'] == st.session_state.orig_sentence, 'correction'] = corrected_translation
|
| 114 |
+
st.session_state.data.loc[st.session_state.data['sentence'] == st.session_state.orig_sentence, 'corrected'] = True
|
| 115 |
+
save_data(st.session_state.data)
|
| 116 |
+
|
| 117 |
+
st.success("Saved!")
|
| 118 |
+
|
| 119 |
+
# Update the sentence for the next iteration.
|
| 120 |
+
noncorrected_sentences = st.session_state.data[(st.session_state.data.translated == True) & (st.session_state.data.corrected == False)]['sentence'].tolist()
|
| 121 |
+
# noncorrected_sentences = st.session_state.data[st.session_state.data['corrected'] == False]['sentence'].tolist()
|
| 122 |
+
if noncorrected_sentences:
|
| 123 |
+
st.session_state.orig_sentence = random.choice(noncorrected_sentences)
|
| 124 |
+
st.session_state.orig_translation = st.session_state.data[st.session_state.data.sentence == st.session_state.orig_sentence]['translation']
|
| 125 |
+
|
| 126 |
+
else:
|
| 127 |
+
st.session_state.orig_translation = "No more sentences to be corrected"
|
| 128 |
+
|
| 129 |
+
corrected_translation = "" # Reset the input value after saving
|
| 130 |
+
|
| 131 |
+
st.button("⏩ Skip to the Next Pair", key="skip_button", on_click=skip_correction)
|
data/Dr-En-space-test.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|