Spaces:
Running
Running
Merge pull request #2 from CintraAI/enhancement/updated-header
Browse files
app.py
CHANGED
@@ -3,6 +3,9 @@ import json
|
|
3 |
import os
|
4 |
from Chunker import CodeChunker
|
5 |
|
|
|
|
|
|
|
6 |
# Function to load JSON data
|
7 |
def load_json_file(file_path):
|
8 |
with open(file_path, 'r') as file:
|
@@ -12,29 +15,24 @@ def load_json_file(file_path):
|
|
12 |
def read_code_from_file(uploaded_file):
|
13 |
return uploaded_file.getvalue().decode("utf-8")
|
14 |
|
15 |
-
|
16 |
-
st.set_page_config(page_title="Cintra Code Chunker", layout="wide")
|
17 |
|
18 |
-
# Assuming app.py and mock_codefiles.json are in the same directory
|
19 |
json_file_path = os.path.join(os.path.dirname(__file__), 'mock_codefiles.json')
|
20 |
code_files_data = load_json_file(json_file_path)
|
21 |
|
22 |
# Extract filenames and contents
|
23 |
code_files = list(code_files_data.keys())
|
24 |
|
25 |
-
# UI Elements
|
26 |
st.title('Cintra Code Chunker')
|
27 |
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
with col1:
|
32 |
# File selection dropdown
|
33 |
selected_file_name = st.selectbox("Select an example code file", code_files)
|
34 |
|
35 |
-
with
|
36 |
# File upload
|
37 |
-
uploaded_file = st.file_uploader("Or upload your code file", type=['py', 'js', 'css'])
|
38 |
|
39 |
# Determine the content and file extension based on selection or upload
|
40 |
if uploaded_file is not None:
|
@@ -53,16 +51,15 @@ def get_language_by_extension(file_extension):
|
|
53 |
elif file_extension == 'css':
|
54 |
return 'css'
|
55 |
else:
|
56 |
-
return None
|
57 |
|
58 |
language = get_language_by_extension(file_extension)
|
59 |
|
60 |
-
|
61 |
-
token_chunk_size = st.number_input('Token Chunk Size Target', min_value=5, max_value=1000, value=25)
|
62 |
|
63 |
-
|
64 |
|
65 |
-
with
|
66 |
st.subheader('Original File')
|
67 |
st.code(code_content, language=language)
|
68 |
|
@@ -72,8 +69,7 @@ code_chunker = CodeChunker(file_extension=file_extension)
|
|
72 |
# Chunk the code content
|
73 |
chunked_code_dict = code_chunker.chunk(code_content, token_chunk_size)
|
74 |
|
75 |
-
|
76 |
-
with col2:
|
77 |
st.subheader('Chunked Code')
|
78 |
for chunk_key, chunk_code in chunked_code_dict.items():
|
79 |
st.code(chunk_code, language=language)
|
|
|
3 |
import os
|
4 |
from Chunker import CodeChunker
|
5 |
|
6 |
+
# Set Streamlit page config at the very beginning
|
7 |
+
st.set_page_config(page_title="Cintra Code Chunker", layout="wide")
|
8 |
+
|
9 |
# Function to load JSON data
|
10 |
def load_json_file(file_path):
|
11 |
with open(file_path, 'r') as file:
|
|
|
15 |
def read_code_from_file(uploaded_file):
|
16 |
return uploaded_file.getvalue().decode("utf-8")
|
17 |
|
18 |
+
st.link_button('Contribute on GitHub', 'https://github.com/CintraAI/code-chunker', help=None, type="secondary", disabled=False, use_container_width=False)
|
|
|
19 |
|
|
|
20 |
json_file_path = os.path.join(os.path.dirname(__file__), 'mock_codefiles.json')
|
21 |
code_files_data = load_json_file(json_file_path)
|
22 |
|
23 |
# Extract filenames and contents
|
24 |
code_files = list(code_files_data.keys())
|
25 |
|
|
|
26 |
st.title('Cintra Code Chunker')
|
27 |
|
28 |
+
selection_col, upload_col = st.columns(2)
|
29 |
+
with selection_col:
|
|
|
|
|
30 |
# File selection dropdown
|
31 |
selected_file_name = st.selectbox("Select an example code file", code_files)
|
32 |
|
33 |
+
with upload_col:
|
34 |
# File upload
|
35 |
+
uploaded_file = st.file_uploader("Or upload your code file", type=['py', 'js', 'css', 'jsx'])
|
36 |
|
37 |
# Determine the content and file extension based on selection or upload
|
38 |
if uploaded_file is not None:
|
|
|
51 |
elif file_extension == 'css':
|
52 |
return 'css'
|
53 |
else:
|
54 |
+
return None
|
55 |
|
56 |
language = get_language_by_extension(file_extension)
|
57 |
|
58 |
+
token_chunk_size = st.number_input('Chunk Size Target Measured in Tokens (tiktoken, gpt-4)', min_value=5, max_value=1000, value=25)
|
|
|
59 |
|
60 |
+
original_col, chunked_col = st.columns(2)
|
61 |
|
62 |
+
with original_col:
|
63 |
st.subheader('Original File')
|
64 |
st.code(code_content, language=language)
|
65 |
|
|
|
69 |
# Chunk the code content
|
70 |
chunked_code_dict = code_chunker.chunk(code_content, token_chunk_size)
|
71 |
|
72 |
+
with chunked_col:
|
|
|
73 |
st.subheader('Chunked Code')
|
74 |
for chunk_key, chunk_code in chunked_code_dict.items():
|
75 |
st.code(chunk_code, language=language)
|