Spaces:
Sleeping
Sleeping
Upload app.py
Browse files
app.py
CHANGED
|
@@ -4,17 +4,43 @@ import dash
|
|
| 4 |
from dash import html, dcc, dash_table
|
| 5 |
import dash_cytoscape as cyto
|
| 6 |
from dash.dependencies import Input, Output, State
|
| 7 |
-
import io
|
| 8 |
import json
|
| 9 |
import urllib.parse
|
| 10 |
|
| 11 |
app = dash.Dash(__name__, suppress_callback_exceptions=True)
|
| 12 |
server = app.server
|
| 13 |
|
|
|
|
|
|
|
| 14 |
def parse_functions_from_files(file_dict):
|
| 15 |
functions = {}
|
| 16 |
defined_funcs = set()
|
| 17 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
for fname, code in file_dict.items():
|
| 19 |
tree = ast.parse(code)
|
| 20 |
for node in ast.walk(tree):
|
|
@@ -26,44 +52,56 @@ def parse_functions_from_files(file_dict):
|
|
| 26 |
for node in ast.walk(tree):
|
| 27 |
if isinstance(node, ast.FunctionDef):
|
| 28 |
func_name = node.name
|
| 29 |
-
args = [
|
|
|
|
|
|
|
| 30 |
returns = []
|
| 31 |
calls = []
|
| 32 |
reads_state = set()
|
| 33 |
writes_state = set()
|
| 34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
for sub in ast.walk(node):
|
| 36 |
-
# Function calls
|
| 37 |
if isinstance(sub, ast.Call) and isinstance(sub.func, ast.Name):
|
| 38 |
if sub.func.id in defined_funcs:
|
| 39 |
calls.append(sub.func.id)
|
| 40 |
-
|
| 41 |
-
|
|
|
|
|
|
|
| 42 |
elif isinstance(sub, ast.Return):
|
| 43 |
if sub.value is None:
|
| 44 |
continue
|
| 45 |
elif isinstance(sub.value, ast.Tuple):
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
| 47 |
else:
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
elif isinstance(sub, ast.Subscript):
|
| 52 |
-
if (
|
| 53 |
-
isinstance(sub.value, ast.Attribute)
|
| 54 |
and isinstance(sub.value.value, ast.Name)
|
| 55 |
and sub.value.value.id == "st"
|
| 56 |
-
and sub.value.attr == "session_state"
|
| 57 |
-
):
|
| 58 |
key = None
|
| 59 |
try:
|
| 60 |
-
if isinstance(sub.slice, ast.Constant):
|
| 61 |
key = sub.slice.value
|
| 62 |
-
elif isinstance(sub.slice, ast.Index) and isinstance(sub.slice.value, ast.Constant):
|
| 63 |
key = sub.slice.value.value
|
| 64 |
except Exception:
|
| 65 |
pass
|
| 66 |
-
|
| 67 |
if key:
|
| 68 |
if isinstance(sub.ctx, ast.Store):
|
| 69 |
writes_state.add(str(key))
|
|
@@ -76,7 +114,7 @@ def parse_functions_from_files(file_dict):
|
|
| 76 |
"calls": calls,
|
| 77 |
"filename": fname,
|
| 78 |
"reads_state": sorted(reads_state),
|
| 79 |
-
"writes_state": sorted(writes_state)
|
| 80 |
}
|
| 81 |
|
| 82 |
return functions
|
|
@@ -92,56 +130,113 @@ def get_reachable_functions(start, graph):
|
|
| 92 |
return visited
|
| 93 |
|
| 94 |
app.layout = html.Div([
|
| 95 |
-
html.H2("
|
| 96 |
-
|
| 97 |
-
dcc.
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
html.Div(id="file-name", style={"marginTop": "10px"}),
|
| 102 |
-
cyto.Cytoscape(
|
| 103 |
-
id="cytoscape-graph",
|
| 104 |
-
layout={"name": "breadthfirst", "directed": True, "padding": 30, "spacingFactor": 1.5},
|
| 105 |
-
style={"width": "100%", "height": "600px"},
|
| 106 |
-
elements=[],
|
| 107 |
-
stylesheet=[
|
| 108 |
-
{"selector": "node", "style": {
|
| 109 |
-
"label": "data(label)",
|
| 110 |
-
"text-wrap": "wrap",
|
| 111 |
-
"text-max-width": 120,
|
| 112 |
-
"text-valign": "center",
|
| 113 |
-
"background-color": "#aed6f1"
|
| 114 |
-
}},
|
| 115 |
-
{"selector": "edge", "style": {
|
| 116 |
-
"curve-style": "bezier",
|
| 117 |
-
"target-arrow-shape": "triangle",
|
| 118 |
-
"arrow-scale": 2,
|
| 119 |
-
"line-color": "#888",
|
| 120 |
-
"target-arrow-color": "#888"
|
| 121 |
-
}}
|
| 122 |
-
],
|
| 123 |
-
userZoomingEnabled=True,
|
| 124 |
-
userPanningEnabled=True,
|
| 125 |
-
minZoom=0.2,
|
| 126 |
-
maxZoom=2,
|
| 127 |
-
wheelSensitivity=0.1
|
| 128 |
-
),
|
| 129 |
-
html.Hr(),
|
| 130 |
-
html.H4("Function Input/Output Table"),
|
| 131 |
-
dash_table.DataTable(
|
| 132 |
-
id="function-table",
|
| 133 |
-
columns=[], # dynamically injected
|
| 134 |
-
style_table={"overflowX": "auto"},
|
| 135 |
-
style_cell={
|
| 136 |
-
"textAlign": "left",
|
| 137 |
-
"whiteSpace": "normal",
|
| 138 |
-
"wordBreak": "break-word",
|
| 139 |
-
"maxWidth": 300
|
| 140 |
-
}
|
| 141 |
-
)
|
| 142 |
])
|
| 143 |
|
| 144 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 145 |
|
| 146 |
@app.callback(
|
| 147 |
Output("main-function-ui", "children"),
|
|
@@ -182,17 +277,25 @@ def update_multi_graph(main_func):
|
|
| 182 |
|
| 183 |
edges = []
|
| 184 |
for src in reachable:
|
| 185 |
-
|
| 186 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 187 |
edge = {
|
| 188 |
"data": {
|
| 189 |
"source": src,
|
| 190 |
-
"target": tgt
|
| 191 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 192 |
}
|
| 193 |
-
if parsed[src]["filename"] != parsed[tgt]["filename"]:
|
| 194 |
-
edge["classes"] = "crossfile"
|
| 195 |
edges.append(edge)
|
|
|
|
| 196 |
|
| 197 |
raw_table_data = [{
|
| 198 |
"Function": fn,
|
|
|
|
| 4 |
from dash import html, dcc, dash_table
|
| 5 |
import dash_cytoscape as cyto
|
| 6 |
from dash.dependencies import Input, Output, State
|
|
|
|
| 7 |
import json
|
| 8 |
import urllib.parse
|
| 9 |
|
| 10 |
app = dash.Dash(__name__, suppress_callback_exceptions=True)
|
| 11 |
server = app.server
|
| 12 |
|
| 13 |
+
uploaded_files = {}
|
| 14 |
+
|
| 15 |
def parse_functions_from_files(file_dict):
|
| 16 |
functions = {}
|
| 17 |
defined_funcs = set()
|
| 18 |
|
| 19 |
+
def infer_type_from_value(value_node):
|
| 20 |
+
if isinstance(value_node, ast.Call) and isinstance(value_node.func, ast.Attribute):
|
| 21 |
+
if value_node.func.attr == "read_csv":
|
| 22 |
+
return "pd.DataFrame"
|
| 23 |
+
elif value_node.func.attr == "DataFrame":
|
| 24 |
+
return "pd.DataFrame"
|
| 25 |
+
elif value_node.func.attr == "array":
|
| 26 |
+
return "np.ndarray"
|
| 27 |
+
elif isinstance(value_node, ast.List):
|
| 28 |
+
return "list"
|
| 29 |
+
elif isinstance(value_node, ast.Dict):
|
| 30 |
+
return "dict"
|
| 31 |
+
elif isinstance(value_node, ast.Set):
|
| 32 |
+
return "set"
|
| 33 |
+
elif isinstance(value_node, ast.Constant):
|
| 34 |
+
if isinstance(value_node.value, str):
|
| 35 |
+
return "str"
|
| 36 |
+
elif isinstance(value_node.value, bool):
|
| 37 |
+
return "bool"
|
| 38 |
+
elif isinstance(value_node.value, int):
|
| 39 |
+
return "int"
|
| 40 |
+
elif isinstance(value_node.value, float):
|
| 41 |
+
return "float"
|
| 42 |
+
return "?"
|
| 43 |
+
|
| 44 |
for fname, code in file_dict.items():
|
| 45 |
tree = ast.parse(code)
|
| 46 |
for node in ast.walk(tree):
|
|
|
|
| 52 |
for node in ast.walk(tree):
|
| 53 |
if isinstance(node, ast.FunctionDef):
|
| 54 |
func_name = node.name
|
| 55 |
+
args = []
|
| 56 |
+
arg_types = {}
|
| 57 |
+
local_assignments = {}
|
| 58 |
returns = []
|
| 59 |
calls = []
|
| 60 |
reads_state = set()
|
| 61 |
writes_state = set()
|
| 62 |
|
| 63 |
+
for arg in node.args.args:
|
| 64 |
+
name = arg.arg
|
| 65 |
+
if arg.annotation:
|
| 66 |
+
arg_type = ast.unparse(arg.annotation)
|
| 67 |
+
else:
|
| 68 |
+
arg_type = "?"
|
| 69 |
+
arg_types[name] = arg_type
|
| 70 |
+
args.append(f"{name}: {arg_type}")
|
| 71 |
+
|
| 72 |
for sub in ast.walk(node):
|
|
|
|
| 73 |
if isinstance(sub, ast.Call) and isinstance(sub.func, ast.Name):
|
| 74 |
if sub.func.id in defined_funcs:
|
| 75 |
calls.append(sub.func.id)
|
| 76 |
+
elif isinstance(sub, ast.Assign):
|
| 77 |
+
for target in sub.targets:
|
| 78 |
+
if isinstance(target, ast.Name):
|
| 79 |
+
local_assignments[target.id] = infer_type_from_value(sub.value)
|
| 80 |
elif isinstance(sub, ast.Return):
|
| 81 |
if sub.value is None:
|
| 82 |
continue
|
| 83 |
elif isinstance(sub.value, ast.Tuple):
|
| 84 |
+
for elt in sub.value.elts:
|
| 85 |
+
label = ast.unparse(elt)
|
| 86 |
+
ret_type = local_assignments.get(label, infer_type_from_value(elt))
|
| 87 |
+
returns.append(f"{label}: {ret_type}")
|
| 88 |
else:
|
| 89 |
+
label = ast.unparse(sub.value)
|
| 90 |
+
ret_type = local_assignments.get(label, infer_type_from_value(sub.value))
|
| 91 |
+
returns.append(f"{label}: {ret_type}")
|
| 92 |
elif isinstance(sub, ast.Subscript):
|
| 93 |
+
if (isinstance(sub.value, ast.Attribute)
|
|
|
|
| 94 |
and isinstance(sub.value.value, ast.Name)
|
| 95 |
and sub.value.value.id == "st"
|
| 96 |
+
and sub.value.attr == "session_state"):
|
|
|
|
| 97 |
key = None
|
| 98 |
try:
|
| 99 |
+
if isinstance(sub.slice, ast.Constant):
|
| 100 |
key = sub.slice.value
|
| 101 |
+
elif isinstance(sub.slice, ast.Index) and isinstance(sub.slice.value, ast.Constant):
|
| 102 |
key = sub.slice.value.value
|
| 103 |
except Exception:
|
| 104 |
pass
|
|
|
|
| 105 |
if key:
|
| 106 |
if isinstance(sub.ctx, ast.Store):
|
| 107 |
writes_state.add(str(key))
|
|
|
|
| 114 |
"calls": calls,
|
| 115 |
"filename": fname,
|
| 116 |
"reads_state": sorted(reads_state),
|
| 117 |
+
"writes_state": sorted(writes_state)
|
| 118 |
}
|
| 119 |
|
| 120 |
return functions
|
|
|
|
| 130 |
return visited
|
| 131 |
|
| 132 |
app.layout = html.Div([
|
| 133 |
+
html.H2("Function Dependency Visualizer (AST-Based)"),
|
| 134 |
+
dcc.Tabs(id="tab-selector", value="intro", children=[
|
| 135 |
+
dcc.Tab(label="π Introduction", value="intro"),
|
| 136 |
+
dcc.Tab(label="π Graph Explorer", value="graph")
|
| 137 |
+
]),
|
| 138 |
+
html.Div(id="tab-content")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 139 |
])
|
| 140 |
|
| 141 |
+
@app.callback(
|
| 142 |
+
Output("tab-content", "children"),
|
| 143 |
+
Input("tab-selector", "value")
|
| 144 |
+
)
|
| 145 |
+
def render_tab(tab):
|
| 146 |
+
if tab == "intro":
|
| 147 |
+
return dcc.Markdown("""
|
| 148 |
+
### π Introduction
|
| 149 |
+
|
| 150 |
+
This tool analyzes uploaded Python files and visualizes how functions call one another.
|
| 151 |
+
|
| 152 |
+
---
|
| 153 |
+
|
| 154 |
+
#### π Features
|
| 155 |
+
- Upload multiple `.py` files
|
| 156 |
+
- Select a top-level function to explore all functions it calls
|
| 157 |
+
- Function table shows:
|
| 158 |
+
- Arguments and return values with inferred types (e.g. `df: pd.DataFrame`)
|
| 159 |
+
- Streamlit `session_state` variables read/written (if applicable)
|
| 160 |
+
- Graph shows:
|
| 161 |
+
- Call order on edges (`1`, `2`, `3`)
|
| 162 |
+
- Thicker lines if a function is called multiple times
|
| 163 |
+
|
| 164 |
+
---
|
| 165 |
+
|
| 166 |
+
#### π How to Use
|
| 167 |
+
1. Switch to the **Graph Explorer** tab
|
| 168 |
+
2. Upload one or more `.py` files
|
| 169 |
+
3. Pick the entry point function
|
| 170 |
+
4. Explore the graph and table dynamically
|
| 171 |
+
|
| 172 |
+
---
|
| 173 |
+
|
| 174 |
+
### π About the Creator
|
| 175 |
+
|
| 176 |
+
This tool was built by **Tomas Larsson**, a data scientist and financial modeler with a passion for making complex topics easy to explore and understand.
|
| 177 |
+
|
| 178 |
+
Tomas is also the creator of [**my.moneytoolbox.com**](https://mymoneytoolbox.com), a blog focused on:
|
| 179 |
+
|
| 180 |
+
- Tax-efficient investing
|
| 181 |
+
- Retirement modeling
|
| 182 |
+
- Personal finance analytics
|
| 183 |
+
- Tools for DIY investors and early retirees
|
| 184 |
+
|
| 185 |
+
Whether you're a fellow data enthusiast or someone planning their financial future, Tomas's blog is a resource-rich destination with transparent tools, clear explanations, and practical guidance.
|
| 186 |
+
|
| 187 |
+
""", style={"padding": "20px", "maxWidth": "900px"})
|
| 188 |
+
|
| 189 |
+
return html.Div([
|
| 190 |
+
html.Div([
|
| 191 |
+
dcc.Upload(id="upload", children=html.Button("Upload Python Files"), multiple=True),
|
| 192 |
+
html.Div(id="main-function-ui", style={"marginRight": "10px", "width": "300px"}),
|
| 193 |
+
html.A("Download Graph JSON", id="download-link", download="graph.json", href="", target="_blank")
|
| 194 |
+
], style={"display": "flex", "flexDirection": "row", "alignItems": "center", "gap": "10px"}),
|
| 195 |
+
html.Div(id="file-name", style={"marginTop": "10px"}),
|
| 196 |
+
cyto.Cytoscape(
|
| 197 |
+
id="cytoscape-graph",
|
| 198 |
+
layout={"name": "breadthfirst", "directed": True, "padding": 30, "spacingFactor": 1.5},
|
| 199 |
+
style={"width": "100%", "height": "600px"},
|
| 200 |
+
elements=[],
|
| 201 |
+
stylesheet=[
|
| 202 |
+
{"selector": "node", "style": {
|
| 203 |
+
"label": "data(label)",
|
| 204 |
+
"text-wrap": "wrap",
|
| 205 |
+
"text-max-width": 120,
|
| 206 |
+
"text-valign": "center",
|
| 207 |
+
"background-color": "#aed6f1"
|
| 208 |
+
}},
|
| 209 |
+
{"selector": "edge", "style": {
|
| 210 |
+
"curve-style": "bezier",
|
| 211 |
+
"target-arrow-shape": "triangle",
|
| 212 |
+
"arrow-scale": 2,
|
| 213 |
+
"line-color": "#888",
|
| 214 |
+
"target-arrow-color": "#888",
|
| 215 |
+
"label": "data(label)",
|
| 216 |
+
"font-size": "14px",
|
| 217 |
+
"text-margin-y": -10
|
| 218 |
+
}}
|
| 219 |
+
],
|
| 220 |
+
userZoomingEnabled=True,
|
| 221 |
+
userPanningEnabled=True,
|
| 222 |
+
minZoom=0.2,
|
| 223 |
+
maxZoom=2,
|
| 224 |
+
wheelSensitivity=0.1
|
| 225 |
+
),
|
| 226 |
+
html.Hr(),
|
| 227 |
+
html.H4("Function Input/Output Table"),
|
| 228 |
+
dash_table.DataTable(
|
| 229 |
+
id="function-table",
|
| 230 |
+
columns=[],
|
| 231 |
+
style_table={"overflowX": "auto"},
|
| 232 |
+
style_cell={
|
| 233 |
+
"textAlign": "left",
|
| 234 |
+
"whiteSpace": "normal",
|
| 235 |
+
"wordBreak": "break-word",
|
| 236 |
+
"maxWidth": 300
|
| 237 |
+
}
|
| 238 |
+
)
|
| 239 |
+
])
|
| 240 |
|
| 241 |
@app.callback(
|
| 242 |
Output("main-function-ui", "children"),
|
|
|
|
| 277 |
|
| 278 |
edges = []
|
| 279 |
for src in reachable:
|
| 280 |
+
call_sequence = parsed[src]["calls"]
|
| 281 |
+
call_index = 1
|
| 282 |
+
for tgt in call_sequence:
|
| 283 |
+
if tgt not in reachable:
|
| 284 |
+
continue
|
| 285 |
+
if not any(e["data"]["source"] == src and e["data"]["target"] == tgt for e in edges):
|
| 286 |
edge = {
|
| 287 |
"data": {
|
| 288 |
"source": src,
|
| 289 |
+
"target": tgt,
|
| 290 |
+
"label": str(call_index)
|
| 291 |
+
},
|
| 292 |
+
"style": {
|
| 293 |
+
"line-width": 4 if call_sequence.count(tgt) > 1 else 2
|
| 294 |
+
},
|
| 295 |
+
"classes": "crossfile" if parsed[src]["filename"] != parsed[tgt]["filename"] else ""
|
| 296 |
}
|
|
|
|
|
|
|
| 297 |
edges.append(edge)
|
| 298 |
+
call_index += 1
|
| 299 |
|
| 300 |
raw_table_data = [{
|
| 301 |
"Function": fn,
|