svgslide 0.1.0__tar.gz

svgslide-0.1.0/CHANGES.md

@@ -0,0 +1,16 @@
+# Changes
+
+## 0.1.0
+
+First release.
+
+- Converts SVG drawings and PDF pages into PowerPoint (`.pptx`) files whose
+  shapes, text and pictures are separate, editable objects; can also write the
+  converted drawing as SVG for other editors.
+- Python API (`convert`, and `convert_figure` for Matplotlib figures) and a
+  `svgslide` command-line tool.
+- Optional merging of identical shapes and coordinate rounding for large
+  plots (`merge`, `merge_scope`, `precision`).
+- Replace fonts in the output file (e.g. `fonts={"sans-serif": "Arial"}`,
+  `--font sans-serif=Arial` or `fonts="Arial"`, `--font Arial`).
+- PDF input needs `pip install "svgslide[pdf]"` (pypdfium2).

svgslide-0.1.0/LICENSE

@@ -0,0 +1,50 @@
+MIT License
+
+Copyright (c) 2026 Krystian Łazowski
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+---------------------------------------------------------------------------
+Acknowledgement
+
+svgslide is written from scratch, but its DrawingML output follows the
+structure used by mpl-office (https://pypi.org/project/mpl-office/), which
+served as the reference for the DrawingML details. Its licence:
+
+    MIT License
+
+    Copyright (c) 2026 Jerritt Collord
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.

svgslide-0.1.0/MANIFEST.in

@@ -0,0 +1 @@
+include CHANGES.md

svgslide-0.1.0/PKG-INFO

@@ -0,0 +1,138 @@
+Metadata-Version: 2.4
+Name: svgslide
+Version: 0.1.0
+Summary: Convert an SVG drawing or a PDF into an editable PowerPoint (.pptx) file
+Author-email: Krystian Łazowski <k.lazowski@exeter.ac.uk>
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/krystianll/svgslide
+Project-URL: Issues, https://github.com/krystianll/svgslide/issues
+Project-URL: Changelog, https://github.com/krystianll/svgslide/blob/main/CHANGES.md
+Keywords: svg,pdf,pptx,powerpoint,drawingml,matplotlib,figures,converter
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Topic :: Multimedia :: Graphics :: Graphics Conversion
+Classifier: Topic :: Office/Business :: Office Suites
+Classifier: Topic :: Scientific/Engineering :: Visualization
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: pdf
+Requires-Dist: pypdfium2>=5.14; extra == "pdf"
+Dynamic: license-file
+
+# svgslide
+
+Convert an SVG drawing or a PDF into a PowerPoint file (`.pptx`) whose parts
+are separate, editable PowerPoint objects: shapes you can recolour and
+resize, real text boxes, pictures, and groups that follow the drawing's
+groups. Made with scientific figures in mind (Matplotlib, seaborn, ggplot),
+but works well for most simple vector images (see Limitations).
+
+Pure Python (3.9+). SVG input needs nothing else; PDF input uses the
+optional pypdfium2.
+
+```sh
+pip install svgslide          # SVG input
+pip install "svgslide[pdf]"   # SVG and PDF input
+svgslide figure.svg           # writes figure.pptx
+```
+
+## Python
+
+```python
+import svgslide
+
+svgslide.convert("figure.svg", "figure.pptx")
+svgslide.convert("figure.pdf", "figure.pptx", pages=[1, 3])
+svgslide.convert_figure(fig, "figure.pptx")      # a Matplotlib figure
+```
+
+Options (both functions):
+
+- `fonts`: replace fonts in the output file, e.g.
+  `fonts={"sans-serif": "Arial"}`, or `fonts="Arial"` for every font; in a
+  table, `"*"` stands for every font not named.
+- `pages` (PDF): which pages to convert; each page becomes its own slide.
+- `merge`, `merge_scope`: shapes that look the same become one shape when
+  there are more than `merge` of them (default 2000; `0` merges every set,
+  `None` never). `merge_scope` says what may merge: `"group"` (default,
+  within one group of the drawing), `"touching"` (shapes that overlap or
+  touch) or `"figure"` (anywhere). Useful for dense scatter plots and
+  genome-browser tracks.
+- `precision`: round outline coordinates to a grid in points (e.g. `0.01`)
+  for smaller files.
+- `format`: `"pptx"` or `"svg"`; a destination ending in `.svg` writes the
+  converted drawing as SVG, for editors that do not open `.pptx`.
+
+Each call returns a report with the counts of shapes, text boxes, pictures
+and groups, the fonts found, and the SVG or PDF features it could not
+convert.
+
+## Command line
+
+```sh
+svgslide INPUT... [-o OUTPUT] [--font NAME|OLD=NEW]... [--pages LIST]
+         [--merge N|none] [--merge-scope group|touching|figure] [--precision PT]
+```
+
+Each input (`.svg`, `.svgz` or `.pdf`) gives a `.pptx` next to it, or
+`-o OUTPUT` (`.pptx` or `.svg`) for a single input. `--font sans-serif=Arial`
+replaces one font, a bare `--font Arial` every font not named, so
+`--font Arial --font "Symbol=Symbol"` keeps Symbol. The fonts found and
+anything that could not be converted are printed for each file.
+
+## What is converted
+
+- Shapes and paths, with their colours, transparency, line widths, dashes,
+  line ends and gradients; CSS styles; reused symbols and markers.
+- Text as text boxes with font, size, weight, colour, alignment and rotation;
+  text along a curve becomes one text box per letter.
+- Text that a program saved only as letter outlines: Matplotlib and seaborn
+  labels become real text again; Cairo labels (R `svg()`, `pdftocairo`)
+  become one shape per label.
+- Embedded PNG, JPEG and GIF images as pictures; rectangular clipping.
+- PDF pages: shapes, text, images and clipping, read with pypdfium2.
+
+## Limitations
+
+- PowerPoint has no clipping for shapes: clipping by a rectangle is applied
+  exactly, any other clip shape only approximately.
+- Not converted (listed in the report): filters such as blur and drop
+  shadows, masks, pattern fills, Inkscape flowed text, animation, external
+  image files; in PDFs, smooth shadings and patterns (patterns become their
+  average colour), soft masks and blend modes.
+- The focal point of radial gradients is ignored.
+- Text from Cairo-made files (R `svg()`, `pdftocairo`) cannot become editable
+  text: those files hold letter shapes, not characters.
+- Letter spacing of text along a curve is estimated, so it can differ a
+  little from the original.
+- Fonts are not embedded: PowerPoint uses the named font if it is installed,
+  otherwise a substitute. Use `fonts=` / `--font` to pick fonts you have.
+- Merged shapes keep each point's full outline, so a merged `.pptx` can be
+  larger than an unmerged one, although it has far fewer objects.
+
+## Development
+
+```sh
+python -m venv .venv
+.venv/bin/pip install -e ".[pdf]" pytest matplotlib
+.venv/bin/python -m pytest
+```
+
+Tests needing pypdfium2 or Matplotlib are skipped without them.
+`SVGSLIDE_CORPUS=/folder/of/svgs` adds a smoke test over a folder of SVG
+files; picture comparison tests run when LibreOffice (`soffice`) and
+Inkscape are on `PATH`.
+
+## Licence
+
+MIT. The DrawingML structure follows mpl-office (MIT); see `LICENSE`.

svgslide-0.1.0/README.md

@@ -0,0 +1,107 @@
+# svgslide
+
+Convert an SVG drawing or a PDF into a PowerPoint file (`.pptx`) whose parts
+are separate, editable PowerPoint objects: shapes you can recolour and
+resize, real text boxes, pictures, and groups that follow the drawing's
+groups. Made with scientific figures in mind (Matplotlib, seaborn, ggplot),
+but works well for most simple vector images (see Limitations).
+
+Pure Python (3.9+). SVG input needs nothing else; PDF input uses the
+optional pypdfium2.
+
+```sh
+pip install svgslide          # SVG input
+pip install "svgslide[pdf]"   # SVG and PDF input
+svgslide figure.svg           # writes figure.pptx
+```
+
+## Python
+
+```python
+import svgslide
+
+svgslide.convert("figure.svg", "figure.pptx")
+svgslide.convert("figure.pdf", "figure.pptx", pages=[1, 3])
+svgslide.convert_figure(fig, "figure.pptx")      # a Matplotlib figure
+```
+
+Options (both functions):
+
+- `fonts`: replace fonts in the output file, e.g.
+  `fonts={"sans-serif": "Arial"}`, or `fonts="Arial"` for every font; in a
+  table, `"*"` stands for every font not named.
+- `pages` (PDF): which pages to convert; each page becomes its own slide.
+- `merge`, `merge_scope`: shapes that look the same become one shape when
+  there are more than `merge` of them (default 2000; `0` merges every set,
+  `None` never). `merge_scope` says what may merge: `"group"` (default,
+  within one group of the drawing), `"touching"` (shapes that overlap or
+  touch) or `"figure"` (anywhere). Useful for dense scatter plots and
+  genome-browser tracks.
+- `precision`: round outline coordinates to a grid in points (e.g. `0.01`)
+  for smaller files.
+- `format`: `"pptx"` or `"svg"`; a destination ending in `.svg` writes the
+  converted drawing as SVG, for editors that do not open `.pptx`.
+
+Each call returns a report with the counts of shapes, text boxes, pictures
+and groups, the fonts found, and the SVG or PDF features it could not
+convert.
+
+## Command line
+
+```sh
+svgslide INPUT... [-o OUTPUT] [--font NAME|OLD=NEW]... [--pages LIST]
+         [--merge N|none] [--merge-scope group|touching|figure] [--precision PT]
+```
+
+Each input (`.svg`, `.svgz` or `.pdf`) gives a `.pptx` next to it, or
+`-o OUTPUT` (`.pptx` or `.svg`) for a single input. `--font sans-serif=Arial`
+replaces one font, a bare `--font Arial` every font not named, so
+`--font Arial --font "Symbol=Symbol"` keeps Symbol. The fonts found and
+anything that could not be converted are printed for each file.
+
+## What is converted
+
+- Shapes and paths, with their colours, transparency, line widths, dashes,
+  line ends and gradients; CSS styles; reused symbols and markers.
+- Text as text boxes with font, size, weight, colour, alignment and rotation;
+  text along a curve becomes one text box per letter.
+- Text that a program saved only as letter outlines: Matplotlib and seaborn
+  labels become real text again; Cairo labels (R `svg()`, `pdftocairo`)
+  become one shape per label.
+- Embedded PNG, JPEG and GIF images as pictures; rectangular clipping.
+- PDF pages: shapes, text, images and clipping, read with pypdfium2.
+
+## Limitations
+
+- PowerPoint has no clipping for shapes: clipping by a rectangle is applied
+  exactly, any other clip shape only approximately.
+- Not converted (listed in the report): filters such as blur and drop
+  shadows, masks, pattern fills, Inkscape flowed text, animation, external
+  image files; in PDFs, smooth shadings and patterns (patterns become their
+  average colour), soft masks and blend modes.
+- The focal point of radial gradients is ignored.
+- Text from Cairo-made files (R `svg()`, `pdftocairo`) cannot become editable
+  text: those files hold letter shapes, not characters.
+- Letter spacing of text along a curve is estimated, so it can differ a
+  little from the original.
+- Fonts are not embedded: PowerPoint uses the named font if it is installed,
+  otherwise a substitute. Use `fonts=` / `--font` to pick fonts you have.
+- Merged shapes keep each point's full outline, so a merged `.pptx` can be
+  larger than an unmerged one, although it has far fewer objects.
+
+## Development
+
+```sh
+python -m venv .venv
+.venv/bin/pip install -e ".[pdf]" pytest matplotlib
+.venv/bin/python -m pytest
+```
+
+Tests needing pypdfium2 or Matplotlib are skipped without them.
+`SVGSLIDE_CORPUS=/folder/of/svgs` adds a smoke test over a folder of SVG
+files; picture comparison tests run when LibreOffice (`soffice`) and
+Inkscape are on `PATH`.
+
+## Licence
+
+MIT. The DrawingML structure follows mpl-office (MIT); see `LICENSE`.

svgslide-0.1.0/pyproject.toml

@@ -0,0 +1,49 @@
+[build-system]
+requires = ["setuptools>=77"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "svgslide"
+version = "0.1.0"
+description = "Convert an SVG drawing or a PDF into an editable PowerPoint (.pptx) file"
+readme = "README.md"
+license = "MIT"
+license-files = ["LICENSE"]
+authors = [{ name = "Krystian Łazowski", email = "k.lazowski@exeter.ac.uk" }]
+requires-python = ">=3.9"
+dependencies = []
+keywords = ["svg", "pdf", "pptx", "powerpoint", "drawingml", "matplotlib", "figures", "converter"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Science/Research",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3 :: Only",
+    "Programming Language :: Python :: 3.9",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
+    "Topic :: Multimedia :: Graphics :: Graphics Conversion",
+    "Topic :: Office/Business :: Office Suites",
+    "Topic :: Scientific/Engineering :: Visualization",
+]
+
+[project.urls]
+Homepage = "https://github.com/krystianll/svgslide"
+Issues = "https://github.com/krystianll/svgslide/issues"
+Changelog = "https://github.com/krystianll/svgslide/blob/main/CHANGES.md"
+
+[project.optional-dependencies]
+# 5.14: first release whose PDFium tells pattern paints apart
+pdf = ["pypdfium2>=5.14"]
+
+[project.scripts]
+svgslide = "svgslide.__main__:main"
+
+[tool.setuptools]
+packages = ["svgslide"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]

svgslide-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

svgslide-0.1.0/svgslide/__init__.py

@@ -0,0 +1,227 @@
+"""svgslide: convert an SVG drawing or a PDF into an editable PowerPoint file.
+
+    >>> import svgslide
+    >>> report = svgslide.convert("figure.svg", "figure.pptx", fonts={"DejaVu Sans": "Arial"})
+    >>> report.skipped
+    []
+"""
+
+import contextlib
+import gzip
+import io
+import math
+import os
+import sys
+import xml.etree.ElementTree as ET
+
+from . import _convert, _drawingml, _merge, _package
+
+__all__ = ['convert', 'convert_figure']
+__version__ = '0.1.0'
+
+
+class Report:
+    """What :func:`convert` wrote and what it had to leave out.
+
+    Attributes: ``shapes``, ``text_boxes``, ``pictures``, ``groups`` (counts),
+    ``slide_size`` ((width, height) in EMU) and ``skipped``: a list of
+    ``(feature, element)`` pairs such as ``("filter", "#shadow")``; for a
+    PDF the counts are summed over the slides and each element names its
+    page, e.g. ``("shading (smooth colour gradient)", "page 2 object 14")``.
+    ``merged`` is the number of shapes that went into merged shapes (a set of
+    n shapes merged into one counts n - 1).  ``fonts`` lists the font names
+    of the drawing's text as written (PDF subset prefixes removed), in the
+    order first seen; text drawn as glyph outlines has none.
+    """
+
+    __slots__ = ('shapes', 'text_boxes', 'pictures', 'groups', 'skipped', 'slide_size', 'merged', 'fonts')
+
+    def __init__(self, counts, skipped, slide_size, merged=0, fonts=()):
+        self.merged = merged
+        self.fonts = list(fonts)
+        self.shapes = counts['shape']
+        self.text_boxes = counts['text']
+        self.pictures = counts['picture']
+        self.groups = counts['group']
+        self.skipped = skipped
+        self.slide_size = slide_size
+
+    def __repr__(self):
+        return ('Report(shapes=%d, text_boxes=%d, pictures=%d, groups=%d, merged=%d, fonts=%r, skipped=%r)'
+                % (self.shapes, self.text_boxes, self.pictures, self.groups, self.merged, self.fonts,
+                   self.skipped))
+
+
+def _read_source(source):
+    if isinstance(source, (bytes, bytearray)):
+        data = bytes(source)
+    elif isinstance(source, str) and source.lstrip(' \t\r\n').startswith('<'):
+        data = source.encode('utf-8')
+        # an XML declaration naming another encoding would now be wrong
+        if data.lstrip().startswith(b'<?xml'):
+            end = data.find(b'?>')
+            data = data[end + 2:] if end > 0 else data
+    else:
+        with open(os.fspath(source), 'rb') as fh:
+            data = fh.read()
+    if data[:2] == b'\x1f\x8b':
+        data = gzip.decompress(data)
+    return data
+
+
+def _is_pdf(data):
+    # a PDF starts with %PDF-; readers accept a byte order mark or white
+    # space before it
+    return data[:1024].lstrip(b'\xef\xbb\xbf \t\r\n\f\x00').startswith(b'%PDF-')
+
+
+def _check_merge(merge, merge_scope):
+    if merge is not None and (isinstance(merge, bool) or not isinstance(merge, int) or merge < 0):
+        raise ValueError('merge must be a whole number >= 0 or None, got %r' % (merge,))
+    if merge_scope not in _merge.SCOPES:
+        raise ValueError('merge_scope must be one of %s, got %r' % (', '.join(_merge.SCOPES), merge_scope))
+
+
+def _check_fonts(fonts):
+    if fonts is None:
+        return None
+    if isinstance(fonts, str):
+        return {'*': fonts}
+    if isinstance(fonts, dict) and all(isinstance(k, str) and isinstance(v, str) for k, v in fonts.items()):
+        return dict(fonts)
+    raise ValueError('fonts must be a font name, a {name: replacement} dict or None, got %r' % (fonts,))
+
+
+def _check_precision(precision):
+    if precision is None:
+        return None
+    if isinstance(precision, bool) or not isinstance(precision, (int, float)) or \
+            not math.isfinite(precision) or precision <= 0:
+        raise ValueError('precision must be a positive number of points or None, got %r' % (precision,))
+    return float(precision)
+
+
+def _format(destination, format):
+    if format is None:
+        name = '' if hasattr(destination, 'write') else os.fspath(destination)
+        return 'svg' if str(name).lower().endswith('.svg') else 'pptx'
+    if format not in ('pptx', 'svg'):
+        raise ValueError('format must be "pptx" or "svg", got %r' % (format,))
+    return format
+
+
+def convert(source, destination, *, fonts=None, pages=None, merge=2000, merge_scope='group', format=None,
+            precision=None):
+    """Convert an SVG or a PDF into a .pptx whose parts are editable shapes.
+
+    source: a file path, or the SVG itself as ``str`` or ``bytes``, or the
+        PDF as ``bytes`` (recognised by its content, not the file name;
+        reading PDFs needs ``pip install svgslide[pdf]``).
+    destination: a file path or a binary file object.
+    fonts: optional ``{font_name: office_font_name}`` replacement table,
+        e.g. ``{"DejaVu Sans": "Arial"}`` (names matched without regard to
+        case); the key ``"*"`` stands for every font not named, and a plain
+        string replaces every font (``"Arial"``).  Names not in it are kept
+        as written.
+    pages: PDF only: the 1-based page number or list of page numbers to
+        convert, e.g. ``2`` or ``[1, 3]``; default all pages.  Each page
+        becomes a slide; the slides have the first page's size and other
+        pages are scaled to fit and centred.  An SVG gives one slide.
+    merge: shapes that look the same (fill, line, opacity) become one shape
+        (one outline per member) when a set has more than this many members;
+        0 merges every set of two or more, None never merges.  Text boxes and
+        pictures never merge.  A member joins only if nothing that looks
+        different, drawn in between, overlaps it, so the picture is unchanged.
+    merge_scope: what forms a set: ``"group"`` (shapes in one SVG group,
+        the default), ``"touching"`` (shapes whose boxes overlap or touch,
+        chained) or ``"figure"`` (anywhere in the drawing).
+    format: ``"pptx"`` or ``"svg"`` (the converted drawing as an SVG file,
+        one page only); default from the destination's suffix (``.svg``),
+        else ``"pptx"``.
+    precision: None (full precision), or a grid size in points, e.g. 0.01:
+        the coordinates of free-form and merged shapes are rounded to that
+        grid and written with fewer digits (smaller files); shape positions
+        and sizes, text boxes and pictures stay exact.
+
+    Returns a :class:`Report` with counts and the skipped features.
+    """
+    _check_merge(merge, merge_scope)
+    fonts = _check_fonts(fonts)
+    precision = _check_precision(precision)
+    grid = precision * 12700 if precision else None      # EMU
+    fmt = _format(destination, format)
+    data = _read_source(source)
+    if _is_pdf(data):
+        from . import _pdf
+        docs = _pdf.read(data, pages)
+    else:
+        if pages is not None:
+            raise ValueError('pages= applies to PDF input only; this source is an SVG')
+        # comments are kept: Matplotlib writes each outlined text's string in one
+        parser = ET.XMLParser(target=ET.TreeBuilder(insert_comments=True))
+        root = ET.fromstring(data, parser=parser)
+        if _convert.css.local_name(root.tag) != 'svg':
+            raise ValueError('not an SVG document (root element is <%s>)' % _convert.css.local_name(root.tag))
+        docs = [(root, None, [])]
+    if fmt == 'svg' and len(docs) > 1:
+        raise ValueError('SVG output holds one page; this PDF has %d pages here: choose one with pages=' % len(docs))
+    series = max(merge, 1) if merge is not None and merge_scope == 'group' else None
+    slides = []
+    trees = []
+    merged = 0
+    found = {}
+    counts = {'shape': 0, 'text': 0, 'picture': 0, 'group': 0}
+    skipped = []
+    size = None
+    for root, page, own_skipped in docs:
+        conv = _convert.Converter(root, fonts, series, grid)
+        nodes, absorbed = _merge.merge(conv.run(), merge, merge_scope, grid)
+        merged += conv.absorbed + absorbed
+        for name in conv.found_fonts:
+            found.setdefault(name, None)
+        trees.append(nodes)
+        ser = _drawingml.Serializer()
+        parts = []
+        ser.write(nodes, parts)
+        slides.append((''.join(parts), conv.media))
+        for k, v in ser.counts.items():
+            counts[k] += v
+        skipped.extend(own_skipped)
+        if page is None:
+            skipped.extend(conv.skipped)
+        else:
+            skipped.extend((f, ('page %d %s' % (page, e)).rstrip()) for f, e in conv.skipped)
+        if size is None:
+            size = (conv.slide_w, conv.slide_h)
+    if fmt == 'svg':
+        from . import _svgout
+        data = _svgout.write(trees[0], size[0], size[1], slides[0][1], precision)
+    else:
+        data = _package.build(slides, size[0], size[1])
+    if hasattr(destination, 'write'):
+        destination.write(data)
+    else:
+        with open(os.fspath(destination), 'wb') as fh:
+            fh.write(data)
+    return Report(counts, skipped, size, merged, found)
+
+
+def convert_figure(figure, destination, *, fonts=None, merge=2000, merge_scope='group', format=None,
+                   precision=None):
+    """Convert a Matplotlib figure into a one-slide .pptx with editable parts.
+
+    figure: a Matplotlib ``Figure``, or any object with a Matplotlib-like
+        ``savefig(file, format='svg', dpi=...)``.  svgslide does not import
+        Matplotlib: when Matplotlib is already loaded, the figure is saved
+        with ``svg.fonttype = 'none'`` so that its text stays text.
+    destination, fonts, merge, merge_scope, format, precision: as for :func:`convert`.
+
+    Returns the :class:`Report`.
+    """
+    buf = io.BytesIO()
+    mpl = sys.modules.get('matplotlib')
+    ctx = mpl.rc_context({'svg.fonttype': 'none'}) if mpl is not None else contextlib.nullcontext()
+    with ctx:
+        figure.savefig(buf, format='svg', dpi='figure')
+    return convert(buf.getvalue(), destination, fonts=fonts, merge=merge, merge_scope=merge_scope, format=format,
+                   precision=precision)