|
1 | | -# Query Jupyter server for the info about a dataframe |
2 | | -import json as _VSCODE_json |
3 | | -import pandas as _VSCODE_pd |
4 | | -import pandas.io.json as _VSCODE_pd_json |
5 | | - |
6 | | -# _VSCode_sub_supportsDataExplorer will contain our list of data explorer supported types |
7 | | -_VSCode_supportsDataExplorer = "['list', 'Series', 'dict', 'ndarray', 'DataFrame']" |
8 | | - |
9 | | -# In IJupyterVariables.getValue this '_VSCode_JupyterTestValue' will be replaced with the json stringified value of the target variable |
10 | | -# Indexes off of _VSCODE_targetVariable need to index types that are part of IJupyterVariable |
11 | | -_VSCODE_targetVariable = _VSCODE_json.loads("""_VSCode_JupyterTestValue""") |
12 | | - |
13 | | -# Function to compute row count for a value |
14 | | -def _VSCODE_getRowCount(var): |
15 | | - if hasattr(var, "shape"): |
16 | | - try: |
17 | | - # Get a bit more restrictive with exactly what we want to count as a shape, since anything can define it |
18 | | - if isinstance(var.shape, tuple): |
19 | | - return var.shape[0] |
20 | | - except TypeError: |
21 | | - return 0 |
22 | | - elif hasattr(var, "__len__"): |
23 | | - try: |
24 | | - return len(var) |
25 | | - except TypeError: |
26 | | - return 0 |
27 | | - |
28 | | - |
29 | | -# First check to see if we are a supported type, this prevents us from adding types that are not supported |
30 | | -# and also keeps our types in sync with what the variable explorer says that we support |
31 | | -if _VSCODE_targetVariable["type"] not in _VSCode_supportsDataExplorer: |
32 | | - del _VSCode_supportsDataExplorer |
33 | | - print(_VSCODE_json.dumps(_VSCODE_targetVariable)) |
34 | | - del _VSCODE_targetVariable |
35 | | -else: |
36 | | - del _VSCode_supportsDataExplorer |
37 | | - _VSCODE_evalResult = eval(_VSCODE_targetVariable["name"]) |
38 | | - |
39 | | - # Figure out shape if not already there. Use the shape to compute the row count |
40 | | - _VSCODE_targetVariable["rowCount"] = _VSCODE_getRowCount(_VSCODE_evalResult) |
41 | | - |
42 | | - # Turn the eval result into a df |
43 | | - _VSCODE_df = _VSCODE_evalResult |
44 | | - if isinstance(_VSCODE_evalResult, list): |
45 | | - _VSCODE_df = _VSCODE_pd.DataFrame(_VSCODE_evalResult) |
46 | | - elif isinstance(_VSCODE_evalResult, _VSCODE_pd.Series): |
47 | | - _VSCODE_df = _VSCODE_pd.Series.to_frame(_VSCODE_evalResult) |
48 | | - elif isinstance(_VSCODE_evalResult, dict): |
49 | | - _VSCODE_evalResult = _VSCODE_pd.Series(_VSCODE_evalResult) |
50 | | - _VSCODE_df = _VSCODE_pd.Series.to_frame(_VSCODE_evalResult) |
51 | | - elif _VSCODE_targetVariable["type"] == "ndarray": |
52 | | - _VSCODE_df = _VSCODE_pd.DataFrame(_VSCODE_evalResult) |
53 | | - elif hasattr(_VSCODE_df, "toPandas"): |
54 | | - _VSCODE_df = _VSCODE_df.toPandas() |
55 | | - _VSCODE_targetVariable["rowCount"] = _VSCODE_getRowCount(_VSCODE_df) |
56 | | - |
57 | | - # If any rows, use pandas json to convert a single row to json. Extract |
58 | | - # the column names and types from the json so we match what we'll fetch when |
59 | | - # we ask for all of the rows |
60 | | - if ( |
61 | | - hasattr(_VSCODE_targetVariable, "rowCount") |
62 | | - and _VSCODE_targetVariable["rowCount"] |
63 | | - ): |
64 | | - try: |
65 | | - _VSCODE_row = _VSCODE_df.iloc[0:1] |
66 | | - _VSCODE_json_row = _VSCODE_pd_json.to_json( |
67 | | - None, _VSCODE_row, date_format="iso" |
68 | | - ) |
69 | | - _VSCODE_columnNames = list(_VSCODE_json.loads(_VSCODE_json_row)) |
70 | | - del _VSCODE_row |
71 | | - del _VSCODE_json_row |
72 | | - except: |
73 | | - _VSCODE_columnNames = list(_VSCODE_df) |
74 | | - else: |
75 | | - _VSCODE_columnNames = list(_VSCODE_df) |
76 | | - |
77 | | - # Compute the index column. It may have been renamed |
78 | | - _VSCODE_indexColumn = _VSCODE_df.index.name if _VSCODE_df.index.name else "index" |
79 | | - _VSCODE_columnTypes = list(_VSCODE_df.dtypes) |
80 | | - del _VSCODE_df |
81 | | - |
82 | | - # Make sure the index column exists |
83 | | - if _VSCODE_indexColumn not in _VSCODE_columnNames: |
84 | | - _VSCODE_columnNames.insert(0, _VSCODE_indexColumn) |
85 | | - _VSCODE_columnTypes.insert(0, "int64") |
86 | | - |
87 | | - # Then loop and generate our output json |
88 | | - _VSCODE_columns = [] |
89 | | - for _VSCODE_n in range(0, len(_VSCODE_columnNames)): |
90 | | - _VSCODE_column_type = _VSCODE_columnTypes[_VSCODE_n] |
91 | | - _VSCODE_column_name = str(_VSCODE_columnNames[_VSCODE_n]) |
92 | | - _VSCODE_colobj = {} |
93 | | - _VSCODE_colobj["key"] = _VSCODE_column_name |
94 | | - _VSCODE_colobj["name"] = _VSCODE_column_name |
95 | | - _VSCODE_colobj["type"] = str(_VSCODE_column_type) |
96 | | - _VSCODE_columns.append(_VSCODE_colobj) |
97 | | - del _VSCODE_column_name |
98 | | - del _VSCODE_column_type |
99 | | - |
100 | | - del _VSCODE_columnNames |
101 | | - del _VSCODE_columnTypes |
102 | | - |
103 | | - # Save this in our target |
104 | | - _VSCODE_targetVariable["columns"] = _VSCODE_columns |
105 | | - _VSCODE_targetVariable["indexColumn"] = _VSCODE_indexColumn |
106 | | - del _VSCODE_columns |
107 | | - del _VSCODE_indexColumn |
108 | | - |
109 | | - # Transform this back into a string |
110 | | - print(_VSCODE_json.dumps(_VSCODE_targetVariable)) |
111 | | - del _VSCODE_targetVariable |
112 | | - |
113 | | - # Cleanup imports |
114 | | - del _VSCODE_json |
115 | | - del _VSCODE_pd |
116 | | - del _VSCODE_pd_json |
| 1 | +# Query Jupyter server for the info about a dataframe |
| 2 | +import json as _VSCODE_json |
| 3 | +import pandas as _VSCODE_pd |
| 4 | +import pandas.io.json as _VSCODE_pd_json |
| 5 | +import builtins as _VSCODE_builtins |
| 6 | + |
| 7 | +# _VSCode_sub_supportsDataExplorer will contain our list of data explorer supported types |
| 8 | +_VSCode_supportsDataExplorer = "['list', 'Series', 'dict', 'ndarray', 'DataFrame']" |
| 9 | + |
| 10 | +# In IJupyterVariables.getValue this '_VSCode_JupyterTestValue' will be replaced with the json stringified value of the target variable |
| 11 | +# Indexes off of _VSCODE_targetVariable need to index types that are part of IJupyterVariable |
| 12 | +_VSCODE_targetVariable = _VSCODE_json.loads("""_VSCode_JupyterTestValue""") |
| 13 | + |
| 14 | +# Function to compute row count for a value |
| 15 | +def _VSCODE_getRowCount(var): |
| 16 | + if hasattr(var, "shape"): |
| 17 | + try: |
| 18 | + # Get a bit more restrictive with exactly what we want to count as a shape, since anything can define it |
| 19 | + if isinstance(var.shape, tuple): |
| 20 | + return var.shape[0] |
| 21 | + except TypeError: |
| 22 | + return 0 |
| 23 | + elif hasattr(var, "__len__"): |
| 24 | + try: |
| 25 | + return _VSCODE_builtins.len(var) |
| 26 | + except TypeError: |
| 27 | + return 0 |
| 28 | + |
| 29 | + |
| 30 | +# First check to see if we are a supported type, this prevents us from adding types that are not supported |
| 31 | +# and also keeps our types in sync with what the variable explorer says that we support |
| 32 | +if _VSCODE_targetVariable["type"] not in _VSCode_supportsDataExplorer: |
| 33 | + del _VSCode_supportsDataExplorer |
| 34 | + print(_VSCODE_json.dumps(_VSCODE_targetVariable)) |
| 35 | + del _VSCODE_targetVariable |
| 36 | +else: |
| 37 | + del _VSCode_supportsDataExplorer |
| 38 | + _VSCODE_evalResult = _VSCODE_builtins.eval(_VSCODE_targetVariable["name"]) |
| 39 | + |
| 40 | + # Figure out shape if not already there. Use the shape to compute the row count |
| 41 | + _VSCODE_targetVariable["rowCount"] = _VSCODE_getRowCount(_VSCODE_evalResult) |
| 42 | + |
| 43 | + # Turn the eval result into a df |
| 44 | + _VSCODE_df = _VSCODE_evalResult |
| 45 | + if isinstance(_VSCODE_evalResult, list): |
| 46 | + _VSCODE_df = _VSCODE_pd.DataFrame(_VSCODE_evalResult) |
| 47 | + elif isinstance(_VSCODE_evalResult, _VSCODE_pd.Series): |
| 48 | + _VSCODE_df = _VSCODE_pd.Series.to_frame(_VSCODE_evalResult) |
| 49 | + elif isinstance(_VSCODE_evalResult, dict): |
| 50 | + _VSCODE_evalResult = _VSCODE_pd.Series(_VSCODE_evalResult) |
| 51 | + _VSCODE_df = _VSCODE_pd.Series.to_frame(_VSCODE_evalResult) |
| 52 | + elif _VSCODE_targetVariable["type"] == "ndarray": |
| 53 | + _VSCODE_df = _VSCODE_pd.DataFrame(_VSCODE_evalResult) |
| 54 | + elif hasattr(_VSCODE_df, "toPandas"): |
| 55 | + _VSCODE_df = _VSCODE_df.toPandas() |
| 56 | + _VSCODE_targetVariable["rowCount"] = _VSCODE_getRowCount(_VSCODE_df) |
| 57 | + |
| 58 | + # If any rows, use pandas json to convert a single row to json. Extract |
| 59 | + # the column names and types from the json so we match what we'll fetch when |
| 60 | + # we ask for all of the rows |
| 61 | + if ( |
| 62 | + hasattr(_VSCODE_targetVariable, "rowCount") |
| 63 | + and _VSCODE_targetVariable["rowCount"] |
| 64 | + ): |
| 65 | + try: |
| 66 | + _VSCODE_row = _VSCODE_df.iloc[0:1] |
| 67 | + _VSCODE_json_row = _VSCODE_pd_json.to_json( |
| 68 | + None, _VSCODE_row, date_format="iso" |
| 69 | + ) |
| 70 | + _VSCODE_columnNames = list(_VSCODE_json.loads(_VSCODE_json_row)) |
| 71 | + del _VSCODE_row |
| 72 | + del _VSCODE_json_row |
| 73 | + except: |
| 74 | + _VSCODE_columnNames = list(_VSCODE_df) |
| 75 | + else: |
| 76 | + _VSCODE_columnNames = list(_VSCODE_df) |
| 77 | + |
| 78 | + # Compute the index column. It may have been renamed |
| 79 | + _VSCODE_indexColumn = _VSCODE_df.index.name if _VSCODE_df.index.name else "index" |
| 80 | + _VSCODE_columnTypes = _VSCODE_builtins.list(_VSCODE_df.dtypes) |
| 81 | + del _VSCODE_df |
| 82 | + |
| 83 | + # Make sure the index column exists |
| 84 | + if _VSCODE_indexColumn not in _VSCODE_columnNames: |
| 85 | + _VSCODE_columnNames.insert(0, _VSCODE_indexColumn) |
| 86 | + _VSCODE_columnTypes.insert(0, "int64") |
| 87 | + |
| 88 | + # Then loop and generate our output json |
| 89 | + _VSCODE_columns = [] |
| 90 | + for _VSCODE_n in _VSCODE_builtins.range( |
| 91 | + 0, _VSCODE_builtins.len(_VSCODE_columnNames) |
| 92 | + ): |
| 93 | + _VSCODE_column_type = _VSCODE_columnTypes[_VSCODE_n] |
| 94 | + _VSCODE_column_name = str(_VSCODE_columnNames[_VSCODE_n]) |
| 95 | + _VSCODE_colobj = {} |
| 96 | + _VSCODE_colobj["key"] = _VSCODE_column_name |
| 97 | + _VSCODE_colobj["name"] = _VSCODE_column_name |
| 98 | + _VSCODE_colobj["type"] = str(_VSCODE_column_type) |
| 99 | + _VSCODE_columns.append(_VSCODE_colobj) |
| 100 | + del _VSCODE_column_name |
| 101 | + del _VSCODE_column_type |
| 102 | + |
| 103 | + del _VSCODE_columnNames |
| 104 | + del _VSCODE_columnTypes |
| 105 | + |
| 106 | + # Save this in our target |
| 107 | + _VSCODE_targetVariable["columns"] = _VSCODE_columns |
| 108 | + _VSCODE_targetVariable["indexColumn"] = _VSCODE_indexColumn |
| 109 | + del _VSCODE_columns |
| 110 | + del _VSCODE_indexColumn |
| 111 | + |
| 112 | + # Transform this back into a string |
| 113 | + print(_VSCODE_json.dumps(_VSCODE_targetVariable)) |
| 114 | + del _VSCODE_targetVariable |
| 115 | + |
| 116 | + # Cleanup imports |
| 117 | + del _VSCODE_json |
| 118 | + del _VSCODE_pd |
| 119 | + del _VSCODE_pd_json |
| 120 | + del _VSCODE_builtins |
0 commit comments