Hide notebook metadata

CHANGELOG.md

@@ -3,6 +3,7 @@
 
 **Added**
 - Activated GitHub code scanning alerts
+- New option `hide_notebook_metadata` to encapsulate the notebook metadata in an HTML comment (#527)
 
 **Changed**
 - Install Jupytext from source on MyBinder to avoid cache issues (#567)

docs/config.md

@@ -105,6 +105,8 @@ It is possible to filter nested metadata. For example, if you want to preserve t
 default_notebook_metadata_filter = "-jupytext.text_representation.jupytext_version"
 ```
 
+Finally, to hide the notebook metadata in an HTML comment in Markdown files, use the option `hide_notebook_metadata`.
+
 ### More options
 
 There are a couple more options available - please have a look at the `JupytextConfiguration` class in [config.py](https://github.com/mwouts/jupytext/blob/master/jupytext/config.py).

jupytext/cli.py

@@ -409,9 +409,14 @@ def jupytext_single_file(nb_file, args, log):
 
     # I. ### Read the notebook ###
     fmt = copy(args.input_format) or {}
-    if config:
-        config.set_default_format_options(fmt)
-    set_format_options(fmt, args.format_options)
+    if not fmt:
+        ext = os.path.splitext(nb_file)[1]
+        if ext:
+            fmt = {"extension": ext}
+    if fmt:
+        if config:
+            config.set_default_format_options(fmt)
+        set_format_options(fmt, args.format_options)
     log(
         "[jupytext] Reading {}{}".format(
             nb_file if nb_file != "-" else "stdin",
@@ -422,19 +427,12 @@ def jupytext_single_file(nb_file, args, log):
     )
 
     notebook = read(nb_file, fmt=fmt)
-    if not fmt:
+    if "extension" in fmt and "format_name" not in fmt:
         text_representation = notebook.metadata.get("jupytext", {}).get(
             "text_representation", {}
         )
-        ext = os.path.splitext(nb_file)[1]
-        if text_representation.get("extension") == ext:
-            fmt = {
-                key: text_representation[key]
-                for key in text_representation
-                if key in ["extension", "format_name"]
-            }
-        elif ext:
-            fmt = {"extension": ext}
+        if text_representation.get("extension") == fmt["extension"]:
+            fmt["format_name"] = text_representation["format_name"]
 
     if config and "formats" not in notebook.metadata.get("jupytext", {}):
         default_formats = config.default_formats(nb_file)

jupytext/config.py

@@ -62,14 +62,21 @@ class JupytextConfiguration(Configurable):
 
     default_notebook_metadata_filter = Unicode(
         u"",
-        help="Cell metadata that should be save in the text representations. "
+        help="Notebook metadata that should be save in the text representations. "
         "Examples: 'all', '-all', 'widgets,nteract', 'kernelspec,jupytext-all'",
         config=True,
     )
 
+    hide_notebook_metadata = Enum(
+        values=[True, False],
+        allow_none=True,
+        help="Should the notebook metadata be wrapped into an HTML comment in the Markdown format?",
+        config=True,
+    )
+
     default_cell_metadata_filter = Unicode(
         u"",
-        help="Notebook metadata that should be saved in the text representations. "
+        help="Cell metadata that should be saved in the text representations. "
         "Examples: 'all', 'hide_input,hide_output'",
         config=True,
     )
@@ -124,6 +131,10 @@ def set_default_format_options(self, format_options, read=False):
             format_options.setdefault(
                 "cell_metadata_filter", self.default_cell_metadata_filter
             )
+        if self.hide_notebook_metadata is not None:
+            format_options.setdefault(
+                "hide_notebook_metadata", self.hide_notebook_metadata
+            )
         if self.comment_magics is not None:
             format_options.setdefault("comment_magics", self.comment_magics)
         if self.split_at_heading:

jupytext/formats.py

@@ -692,6 +692,7 @@ def short_form_multiple_formats(jupytext_formats):
 _VALID_FORMAT_INFO = ["extension", "format_name", "suffix", "prefix"]
 _BINARY_FORMAT_OPTIONS = [
     "comment_magics",
+    "hide_notebook_metadata",
     "split_at_heading",
     "rst2md",
     "cell_metadata_json",

jupytext/header.py

@@ -116,6 +116,12 @@ def metadata_and_cell_to_header(notebook, metadata, text_format, ext):
     if header:
         header = ["---"] + header + ["---"]
 
+        if (
+            metadata.get("jupytext", {}).get("hide_notebook_metadata", False)
+            and text_format.format_name == "markdown"
+        ):
+            header = ["<!--", ""] + header + ["", "-->"]
+
     return comment_lines(header, text_format.header_prefix), lines_to_next_cell
 
 
@@ -142,10 +148,13 @@ def header_to_metadata_and_cell(lines, header_prefix, ext=None):
 
     header = []
     jupyter = []
-    injupyter = False
+    in_jupyter = False
+    in_html_div = False
+
+    start = 0
+    started = False
     ended = False
     metadata = {}
-    start = 0
     i = -1
 
     comment = "#" if header_prefix == "#'" else header_prefix
@@ -167,27 +176,37 @@ def header_to_metadata_and_cell(lines, header_prefix, ext=None):
                 metadata.setdefault("jupytext", {})["encoding"] = line
                 start = i + 1
                 continue
-
         if not line.startswith(header_prefix):
             break
+        if not comment:
+            if line.strip().startswith("<!--"):
+                in_html_div = True
+                continue
 
-        line = uncomment_line(line, header_prefix)
-
-        if i == start:
-            if _HEADER_RE.match(line):
+        if in_html_div:
+            if ended:
+                if "-->" in line:
+                    break
+            if not started and not line.strip():
                 continue
-            break
 
-        if i > start and _HEADER_RE.match(line):
-            ended = True
-            break
+        line = uncomment_line(line, header_prefix)
+        if _HEADER_RE.match(line):
+            if not started:
+                started = True
+                continue
+            else:
+                ended = True
+                if in_html_div:
+                    continue
+                break
 
         if _JUPYTER_RE.match(line):
-            injupyter = True
+            in_jupyter = True
         elif line and not _LEFTSPACE_RE.match(line):
-            injupyter = False
+            in_jupyter = False
 
-        if injupyter:
+        if in_jupyter:
             jupyter.append(line)
         else:
             header.append(line)

tests/test_cli_config.py

@@ -3,6 +3,7 @@
 from jupytext.cli import jupytext
 from jupytext.jupytext import read, write
 from jupytext.header import header_to_metadata_and_cell
+from jupytext.compare import compare
 
 
 def test_default_jupytext_formats(tmpdir):
@@ -89,3 +90,38 @@ def test_save_using_preferred_and_default_format(config, tmpdir):
     # read py file
     nb_py = read(str(tmp_py))
     assert nb_py.metadata["jupytext"]["text_representation"]["format_name"] == "percent"
+
+
+def test_hide_notebook_metadata(tmpdir, no_jupytext_version_number):
+    tmpdir.join(".jupytext").write("hide_notebook_metadata = true")
+    tmp_ipynb = tmpdir.join("notebook.ipynb")
+    tmp_md = tmpdir.join("notebook.md")
+
+    nb = new_notebook(
+        cells=[new_code_cell("1 + 1")], metadata={"jupytext": {"formats": "ipynb,md"}}
+    )
+
+    write(nb, str(tmp_ipynb))
+    jupytext([str(tmp_ipynb), "--sync"])
+
+    with open(str(tmp_md)) as stream:
+        text_md = stream.read()
+
+    compare(
+        text_md,
+        """<!--
+
+---
+jupyter:
+  jupytext:
+    formats: ipynb,md
+    hide_notebook_metadata: true
+---
+
+-->
+
+```python
+1 + 1
+```
+""",
+    )

tests/test_header.py

@@ -148,3 +148,41 @@ def test_multiline_metadata(
     compare(actual, markdown)
     nb2 = jupytext.reads(markdown, ".md")
     compare(nb2, notebook)
+
+
+def test_header_in_html_comment():
+    text = """<!--
+
+---
+jupyter:
+  title: Sample header
+---
+
+-->
+"""
+    lines = text.splitlines()
+    metadata, _, cell, _ = header_to_metadata_and_cell(lines, "")
+
+    assert metadata == {"title": "Sample header"}
+    assert cell is None
+
+
+def test_header_to_html_comment(no_jupytext_version_number):
+    metadata = {"jupytext": {"mainlanguage": "python", "hide_notebook_metadata": True}}
+    nb = new_notebook(metadata=metadata, cells=[])
+    header, lines_to_next_cell = metadata_and_cell_to_header(
+        nb, metadata, get_format_implementation(".md"), ".md"
+    )
+    compare(
+        "\n".join(header),
+        """<!--
+
+---
+jupyter:
+  jupytext:
+    hide_notebook_metadata: true
+    mainlanguage: python
+---
+
+-->""",
+    )

tests/test_read_simple_markdown.py

@@ -916,3 +916,37 @@ def test_custom_metadata(
     compare(md2, md)
     nb2 = jupytext.reads(md, "md")
     compare_notebooks(nb2, nb)
+
+
+def test_hide_notebook_metadata(
+    no_jupytext_version_number,
+    nb=new_notebook(
+        metadata={
+            "jupytext": {"hide_notebook_metadata": True},
+            "kernelspec": {
+                "display_name": "Python 3",
+                "language": "python",
+                "name": "python3",
+            },
+        }
+    ),
+    md="""<!--
+
+---
+jupyter:
+  jupytext:
+    hide_notebook_metadata: true
+  kernelspec:
+    display_name: Python 3
+    language: python
+    name: python3
+---
+
+-->
+""",
+):
+    """Test the hide_notebook_metadata option"""
+    md2 = jupytext.writes(nb, "md")
+    compare(md2, md)
+    nb2 = jupytext.reads(md, "md")
+    compare_notebooks(nb2, nb)