Bump numpy from 1.18.1 to 1.21.0 in /scripts/data_collection

README.md

@@ -10,8 +10,32 @@ Library for analyzing source code with graphs and NLP. What this repository can
 
 ### Installation
 
+You need to use conda, create virtual environment `SourceCodeTools` with python 3.8
+```bash
+conda create -n SourceCodeTools python=3.8
+```
+
+If you plan to use graphviz
+```python
+conda install -c conda-forge pygraphviz
+```
+
+Install CUDA 11.1 if needed
+```python
+conda install -c nvidia cudatoolkit=11.1
+```
+
+To install SourceCodeTools library run
 ```bash
 git clone https://github.com/VitalyRomanov/method-embedding.git
 cd method-embedding
 pip install -e .
-```
+# pip install -e .[gpu]
+```
+
+### Installing Sourcetrail
+Download a release from [Github repo](https://github.com/CoatiSoftware/Sourcetrail/releases) (latest tested version is 2020.1.117). Add Sourcetrail location to `PATH`
+```bash
+echo 'export PATH=/path/to/Sourcetrail_2020_1_117:$PATH' >> ~/.bashrc
+```
+Scripts that use Sourcetrail work on Linux, some issues were spotted on Macs. 

SourceCodeTools/code/ast_tools.py

@@ -30,9 +30,18 @@ def get_mentions(function, root, mention):
 
 
 def get_descendants(function, children):
+    """
 
+    :param function: function string
+    :param children: List of targets.
+    :return: Offsets for attributes or names that are used as target for assignment operation. Subscript, Tuple and List
+    targets are skipped.
+    """
     descendants = []
 
+    # if isinstance(children, ast.Tuple):
+    #     descendants.extend(get_descendants(function, children.elts))
+    # else:
     for chld in children:
         # for node in ast.walk(chld):
         node = chld
@@ -42,6 +51,10 @@ def get_descendants(function, children):
                                 [(node.lineno-1, node.end_lineno-1, node.col_offset, node.end_col_offset, "new_var")], as_bytes=True)
             # descendants.append((node.id, offset[-1]))
             descendants.append((function[offset[-1][0]:offset[-1][1]], offset[-1]))
+        # elif isinstance(node, ast.Tuple):
+        #     descendants.extend(get_descendants(function, node.elts))
+        elif isinstance(node, ast.Subscript) or isinstance(node, ast.Tuple) or isinstance(node, ast.List):
+            pass # skip for now
         else:
             raise Exception("")