From bdb4b6305d4ee4fbe907cf200c2651169e6f9a40 Mon Sep 17 00:00:00 2001
From: Edward Delaporte <delaport@illinois.edu>
Date: Wed, 14 Feb 2024 13:18:29 -0600
Subject: [PATCH] Recipe for cleaning data cells out of Jupyter notebooks

---
 files/gitattributes       |  1 +
 util/jupyter_git_clean.py | 67 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)
 create mode 100644 files/gitattributes
 create mode 100644 util/jupyter_git_clean.py

diff --git a/files/gitattributes b/files/gitattributes
new file mode 100644
index 0000000..2d71c45
--- /dev/null
+++ b/files/gitattributes
@@ -0,0 +1 @@
+*.ipynb filter=~/dotfiles/util/jupyter_git_clean.py
\ No newline at end of file
diff --git a/util/jupyter_git_clean.py b/util/jupyter_git_clean.py
new file mode 100644
index 0000000..301ea78
--- /dev/null
+++ b/util/jupyter_git_clean.py
@@ -0,0 +1,67 @@
+#!/usr/bin/env python
+
+"""
+Clear outputs of IPython notebooks.
+
+By default, it prints the notebooks without outputs into stdout.
+When the --in-place option is given, all files will be overwritten.
+
+From
+https://github.com/ivanov/nbutils/blob/master/nbstrip.py
+
+"""
+
+import sys
+
+from IPython.nbformat import current as nbformat
+
+
+
+
+def clear_outputs(nb):
+    """Clear output of notebook `nb` INPLACE."""
+    for ws in nb.worksheets:
+        for cell in ws.cells:
+            cell.outputs = []
+
+
+def stripoutput(inputs, inplace=False):
+    """
+    Strip output of notebooks.
+
+    Parameters
+    ----------
+    inputs : list of string
+        Path to the notebooks to be processed.
+    inplace : bool
+        If this is `True`, outputs in the input files will be deleted.
+        Default is `False`.
+
+    """
+    for inpath in inputs:
+        with file(inpath) as fp:
+            nb = nbformat.read(fp, 'ipynb')
+        clear_outputs(nb)
+        if inplace:
+            with file(inpath, 'w') as fp:
+                nbformat.write(nb, fp, 'ipynb')
+        else:
+            nbformat.write(nb, sys.stdout, 'ipynb')
+
+
+def main():
+    from argparse import ArgumentParser
+    parser = ArgumentParser(description=__doc__)
+    parser.add_argument('inputs', nargs='+', metavar='input',
+                        help='Paths to notebook files.')
+    #XXX: TODO: handle alternative outputs here, such as to stdout or to file
+    parser.add_argument('-i', '--inplace', '--in-place', default=True,
+            action='store_true', 
+            help='Overwrite existing notebook when given.')
+
+    args = parser.parse_args()
+    stripoutput(**vars(args))
+
+
+if __name__ == '__main__':
+    main()
-- 
2.47.3