Import AITURK IDE 1.0.0-beta.1 from Hermes 63279301; preserve MIT license

This commit is contained in:
2026-09-05 13:26:46 +03:00
commit 03634b1ca3
11340 changed files with 3442369 additions and 0 deletions
@@ -0,0 +1,214 @@
#!/usr/bin/env python3
"""Create a .pptx presentation from a JSON deck spec.
Spec format (all positions/sizes in inches, colors as RRGGBB hex):
{
"slide_size": "16:9", // or "4:3" (default "16:9")
"slides": [
{"layout": "title", "title": "My Deck", "subtitle": "Q3 review"},
{"layout": "title_content", "title": "Agenda",
"bullets": ["Top item",
{"text": "Sub item", "level": 1, "bold": true,
"size": 18, "color": "CC0000", "font": "Arial",
"italic": false,
"link": "https://example.com/agenda"}],
"background": "1F2937", // solid slide background (hex)
"footer": "Confidential", // footer placeholder text
"slide_number": true, // enable slide-number placeholder
"notes": "Speaker notes for this slide"},
{"layout": "blank", "title": "Widgets",
"images": [{"path": "logo.png", "left": 1, "top": 1, "width": 3}],
"tables": [{"left": 1, "top": 2, "width": 6, "height": 2,
"rows": [["H1", "H2"], ["a", "b"]]}],
"shapes": [{"type": "rounded_rectangle", "left": 8, "top": 1,
"width": 3, "height": 1, "fill": "4472C4",
"text": "Callout", "text_color": "FFFFFF"}],
"charts": [{"type": "bar", "left": 1, "top": 3, "width": 6,
"height": 3.5, "title": "Sales",
"categories": ["Q1", "Q2"],
"series": {"North": [10, 20], "South": [7, 13]}}]}
]
}
Layouts: title, title_content, section, two_content, title_only, blank
Chart types: bar, bar_h, line, pie Shape types: rectangle,
rounded_rectangle, oval, diamond, right_arrow, chevron
"""
import argparse
import copy
import json
import sys
from pptx import Presentation
from pptx.chart.data import CategoryChartData
from pptx.dml.color import RGBColor
from pptx.enum.chart import XL_CHART_TYPE
from pptx.enum.shapes import MSO_SHAPE
from pptx.util import Inches, Pt
LAYOUTS = {"title": 0, "title_content": 1, "section": 2,
"two_content": 3, "title_only": 5, "blank": 6}
CHART_TYPES = {"bar": XL_CHART_TYPE.COLUMN_CLUSTERED,
"bar_h": XL_CHART_TYPE.BAR_CLUSTERED,
"line": XL_CHART_TYPE.LINE_MARKERS,
"pie": XL_CHART_TYPE.PIE}
SHAPE_TYPES = {"rectangle": MSO_SHAPE.RECTANGLE,
"rounded_rectangle": MSO_SHAPE.ROUNDED_RECTANGLE,
"oval": MSO_SHAPE.OVAL, "diamond": MSO_SHAPE.DIAMOND,
"right_arrow": MSO_SHAPE.RIGHT_ARROW,
"chevron": MSO_SHAPE.CHEVRON}
def style_run(run, spec):
"""Apply font styling from a bullet/text spec dict to a run."""
font = run.font
if spec.get("size"):
font.size = Pt(spec["size"])
if spec.get("bold") is not None:
font.bold = spec["bold"]
if spec.get("italic") is not None:
font.italic = spec["italic"]
if spec.get("font"):
font.name = spec["font"]
if spec.get("color"):
font.color.rgb = RGBColor.from_string(spec["color"])
if spec.get("link"):
run.hyperlink.address = spec["link"]
def add_bullets(text_frame, bullets):
text_frame.clear()
for i, item in enumerate(bullets):
if isinstance(item, str):
item = {"text": item}
para = text_frame.paragraphs[0] if i == 0 else text_frame.add_paragraph()
para.level = int(item.get("level", 0))
run = para.add_run()
run.text = item.get("text", "")
style_run(run, item)
def copy_layout_placeholder(slide, ph_idx):
"""Copy a layout placeholder (footer=11, slide number=12) onto the
slide so it actually renders; returns the shape or None if the layout
does not provide it."""
for ph in slide.slide_layout.placeholders:
if ph.placeholder_format.idx == ph_idx:
slide.shapes._spTree.append(copy.deepcopy(ph._element))
for shape in slide.placeholders:
if shape.placeholder_format.idx == ph_idx:
return shape
return None
def build_slide(prs, spec):
layout_idx = LAYOUTS.get(spec.get("layout", "title_content"), 1)
slide = prs.slides.add_slide(prs.slide_layouts[layout_idx])
if spec.get("background"):
fill = slide.background.fill
fill.solid()
fill.fore_color.rgb = RGBColor.from_string(spec["background"])
if spec.get("slide_number"):
copy_layout_placeholder(slide, 12)
if spec.get("footer"):
shape = copy_layout_placeholder(slide, 11)
if shape is not None:
shape.text_frame.text = spec["footer"]
if spec.get("title") is not None and slide.shapes.title is not None:
slide.shapes.title.text = spec["title"]
if spec.get("subtitle") is not None:
for ph in slide.placeholders:
if ph.placeholder_format.idx == 1:
ph.text = spec["subtitle"]
break
if spec.get("bullets"):
body = next((ph for ph in slide.placeholders
if ph.placeholder_format.idx != 0), None)
if body is None:
body = slide.shapes.add_textbox(Inches(0.5), Inches(1.5),
Inches(9), Inches(5))
add_bullets(body.text_frame, spec["bullets"])
for img in spec.get("images", []):
kwargs = {}
if img.get("width"):
kwargs["width"] = Inches(img["width"])
if img.get("height"):
kwargs["height"] = Inches(img["height"])
slide.shapes.add_picture(img["path"], Inches(img.get("left", 1)),
Inches(img.get("top", 1)), **kwargs)
for tbl in spec.get("tables", []):
rows = tbl["rows"]
shape = slide.shapes.add_table(
len(rows), len(rows[0]), Inches(tbl.get("left", 1)),
Inches(tbl.get("top", 2)), Inches(tbl.get("width", 6)),
Inches(tbl.get("height", 2)))
for r, row in enumerate(rows):
for c, val in enumerate(row):
shape.table.cell(r, c).text = str(val)
for shp in spec.get("shapes", []):
shape = slide.shapes.add_shape(
SHAPE_TYPES.get(shp.get("type", "rectangle"), MSO_SHAPE.RECTANGLE),
Inches(shp.get("left", 1)), Inches(shp.get("top", 1)),
Inches(shp.get("width", 2)), Inches(shp.get("height", 1)))
if shp.get("fill"):
shape.fill.solid()
shape.fill.fore_color.rgb = RGBColor.from_string(shp["fill"])
if shp.get("text"):
shape.text_frame.text = shp["text"]
if shp.get("text_color"):
run = shape.text_frame.paragraphs[0].runs[0]
run.font.color.rgb = RGBColor.from_string(shp["text_color"])
for cht in spec.get("charts", []):
data = CategoryChartData()
data.categories = cht["categories"]
for name, values in cht["series"].items():
data.add_series(name, values)
frame = slide.shapes.add_chart(
CHART_TYPES.get(cht.get("type", "bar"),
XL_CHART_TYPE.COLUMN_CLUSTERED),
Inches(cht.get("left", 1)), Inches(cht.get("top", 2)),
Inches(cht.get("width", 6)), Inches(cht.get("height", 4)), data)
if cht.get("title"):
frame.chart.has_title = True
frame.chart.chart_title.text_frame.text = cht["title"]
if spec.get("notes"):
slide.notes_slide.notes_text_frame.text = spec["notes"]
return slide
def main(argv=None):
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
parser = argparse.ArgumentParser(
description="Create a .pptx deck from a JSON spec.",
epilog=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument("spec", help="path to JSON deck spec")
parser.add_argument("output", help="output .pptx path")
args = parser.parse_args(argv)
with open(args.spec, encoding="utf-8") as fh:
spec = json.load(fh)
prs = Presentation()
if spec.get("slide_size", "16:9") == "16:9":
prs.slide_width, prs.slide_height = Inches(13.333), Inches(7.5)
else:
prs.slide_width, prs.slide_height = Inches(10), Inches(7.5)
for slide_spec in spec.get("slides", []):
build_slide(prs, slide_spec)
prs.save(args.output)
print(json.dumps({"ok": True, "output": args.output,
"slides": len(prs.slides._sldIdLst)}))
return 0
if __name__ == "__main__":
sys.exit(main())
@@ -0,0 +1,436 @@
#!/usr/bin/env python3
"""Edit a .pptx in place (or save to --output).
Operations (repeatable / combinable):
--replace-text OLD NEW Replace text everywhere (slides, tables, notes).
Adjacent runs with identical formatting are
merged first, so matches PowerPoint split across
identically-formatted runs keep their formatting.
Only a match spanning genuinely different
formats falls back to a paragraph rewrite with
the first run's font (documented caveat).
--chart-data SPEC.json Update a chart. Full replace spec:
{"slide": 0, "chart": 0,
"categories": ["Q1", "Q2"],
"series": {"North": [1, 2], "South": [3, 4]}}
Or surgical ops (existing data is read, modified,
and written back via replace_data):
{"slide": 0, "chart": 0, "ops": [
{"op": "update_series", "name": "North",
"values": [5, 6]},
{"op": "add_series", "name": "East",
"values": [1, 2]},
{"op": "remove_series", "name": "South"},
{"op": "rename_category", "from": "Q1",
"to": "Q1 FY26"},
{"op": "set_title", "title": "New title"}]}
--swap-image SLIDE SHAPE_NAME NEW_IMAGE
Replace a picture's bits, keeping position/size.
--remove-slide N Delete slide at index N (0-based).
--move-slide FROM TO Reorder: move slide FROM to position TO.
--duplicate-slide N Append an independent deep copy of slide N
(text, images, tables, shapes, notes). Refuses
slides containing charts (a chart embeds an XLSX
workbook part that cannot be cloned reliably).
--set-background N HEX Solid background color for slide N.
--hyperlink N TEXT URL Make runs containing TEXT on slide N links.
--enable-slide-number N Copy the layout's slide-number placeholder in.
--set-footer N TEXT Enable the layout's footer placeholder with TEXT.
--set-notes N TEXT Replace slide N's speaker notes.
--append-notes N TEXT Append a paragraph to slide N's speaker notes.
"""
import argparse
import copy
import json
import sys
from lxml import etree
from pptx import Presentation
from pptx.chart.data import CategoryChartData
from pptx.dml.color import RGBColor
from pptx.enum.shapes import MSO_SHAPE_TYPE
from pptx.oxml.ns import qn
R_EMBED = "{http://schemas.openxmlformats.org/officeDocument/2006/relationships}"
def _run_format_key(r):
"""Canonical string for a run's <a:rPr>; None when absent."""
rPr = r.find(qn("a:rPr"))
if rPr is None:
return None
return etree.tostring(rPr)
def normalize_runs(para):
"""Merge adjacent runs whose formatting is byte-identical.
PowerPoint splits paragraph text into runs at spell-check and edit
boundaries even when formatting never changes; merging them back makes
cross-run text replacement lossless for the common case.
"""
runs = list(para.runs)
i = 0
while i + 1 < len(runs):
a, b = runs[i], runs[i + 1]
if (_run_format_key(a._r) == _run_format_key(b._r)
and a._r.getnext() is b._r):
a.text = a.text + b.text
b._r.getparent().remove(b._r)
runs.pop(i + 1)
else:
i += 1
def replace_in_text_frame(text_frame, old, new):
count = 0
for para in text_frame.paragraphs:
joined = "".join(run.text for run in para.runs)
if old not in joined:
continue
if not any(old in run.text for run in para.runs):
# Match spans runs: merge identically-formatted neighbours
# first, which resolves pure spell-check splits losslessly.
normalize_runs(para)
if any(old in run.text for run in para.runs):
# Run-level replace: preserves each run's formatting exactly.
for run in para.runs:
if old in run.text:
count += run.text.count(old)
run.text = run.text.replace(old, new)
else:
# Match spans genuinely differently-formatted runs -> rewrite
# paragraph, keeping only the first run's formatting (caveat).
joined = "".join(run.text for run in para.runs)
count += joined.count(old)
first = para.runs[0]
first.text = joined.replace(old, new)
for run in para.runs[1:]:
run._r.getparent().remove(run._r)
return count
def iter_text_frames(slide):
for shape in slide.shapes:
if shape.has_text_frame:
yield shape.text_frame
if shape.has_table:
for row in shape.table.rows:
for cell in row.cells:
yield cell.text_frame
if slide.has_notes_slide:
yield slide.notes_slide.notes_text_frame
def replace_text(prs, old, new):
total = 0
for slide in prs.slides:
for tf in iter_text_frames(slide):
total += replace_in_text_frame(tf, old, new)
return total
def _read_chart_data(chart):
"""Current categories and ordered (name, values) pairs of a chart."""
categories = [str(c) for c in chart.plots[0].categories]
series = []
for plot in chart.plots:
for s in plot.series:
try:
name = s.name
except (AttributeError, KeyError):
name = ""
series.append([name, list(s.values)])
return categories, series
def update_chart(prs, spec_path):
"""Full replace ("categories"+"series") or surgical "ops".
python-pptx can only swap a chart's entire dataset (replace_data), so
surgical ops are implemented as read-existing -> modify -> replace.
"""
with open(spec_path, encoding="utf-8") as fh:
spec = json.load(fh)
slide = prs.slides[spec.get("slide", 0)]
charts = [s.chart for s in slide.shapes if s.has_chart]
if not charts:
raise SystemExit(f"no chart on slide {spec.get('slide', 0)}")
chart = charts[spec.get("chart", 0)]
if "ops" in spec:
categories, series = _read_chart_data(chart)
dirty = False
for op in spec["ops"]:
kind = op["op"]
if kind == "update_series":
match = [s for s in series if s[0] == op["name"]]
if not match:
raise SystemExit(f"no series named {op['name']!r}")
match[0][1] = op["values"]
dirty = True
elif kind == "add_series":
series.append([op["name"], op["values"]])
dirty = True
elif kind == "remove_series":
before = len(series)
series = [s for s in series if s[0] != op["name"]]
if len(series) == before:
raise SystemExit(f"no series named {op['name']!r}")
dirty = True
elif kind == "rename_category":
if "index" in op:
idx = int(op["index"])
else:
if op["from"] not in categories:
raise SystemExit(
f"no category named {op['from']!r}")
idx = categories.index(op["from"])
categories[idx] = op["to"]
dirty = True
elif kind == "set_title":
chart.has_title = True
chart.chart_title.text_frame.text = op["title"]
else:
raise SystemExit(f"unknown chart op {kind!r}")
if dirty:
data = CategoryChartData()
data.categories = categories
for name, values in series:
data.add_series(name, values)
chart.replace_data(data)
return
data = CategoryChartData()
data.categories = spec["categories"]
for name, values in spec["series"].items():
data.add_series(name, values)
chart.replace_data(data)
def swap_image(prs, slide_idx, shape_name, new_path):
slide = prs.slides[int(slide_idx)]
for shape in slide.shapes:
if (shape.shape_type == MSO_SHAPE_TYPE.PICTURE
and shape.name == shape_name):
image_part, rid = slide.part.get_or_add_image_part(new_path)
blip = shape._element.blipFill.blip
blip.set(R_EMBED + "embed", rid)
return True
raise SystemExit(f"no picture named {shape_name!r} on slide {slide_idx}")
def remove_slide(prs, index):
sldIdLst = prs.slides._sldIdLst
slide_id = list(sldIdLst)[int(index)]
rid = slide_id.get(R_EMBED + "id")
prs.part.drop_rel(rid)
sldIdLst.remove(slide_id)
def move_slide(prs, src, dst):
"""Reorder by moving the <p:sldId> element inside <p:sldIdLst>."""
sldIdLst = prs.slides._sldIdLst
ids = list(sldIdLst)
element = ids[int(src)]
sldIdLst.remove(element)
sldIdLst.insert(int(dst), element)
def duplicate_slide(prs, index):
"""Append an independent deep copy of slide `index`.
Copies the shape tree XML and re-creates image/media relationships on
the new slide part, remapping rIds. Charts are refused: each chart
relationship embeds a separate XLSX workbook part, and cloning that
graph reliably is not supported — better to refuse than corrupt.
"""
source = prs.slides[int(index)]
if any(sh.has_chart for sh in source.shapes):
raise SystemExit(
f"slide {index} contains a chart; duplication of chart slides "
"is not supported (chart XML embeds a workbook part that "
"cannot be cloned safely). Rebuild the chart on a new slide "
"with pptx_create.py / pptx_from_template.py instead.")
dest = prs.slides.add_slide(source.slide_layout)
# drop the placeholders add_slide seeded from the layout
for shape in list(dest.shapes):
shape._element.getparent().remove(shape._element)
for shape in source.shapes:
dest.shapes._spTree.append(copy.deepcopy(shape._element))
# re-create the source slide's part relationships on the copy
rid_map = {}
for rel in list(source.part.rels.values()):
if rel.reltype.endswith(("/slideLayout", "/notesSlide")):
continue
if rel.is_external:
new_rid = dest.part.rels.get_or_add_ext_rel(
rel.reltype, rel.target_ref)
else:
new_rid = dest.part.relate_to(rel.target_part, rel.reltype)
rid_map[rel.rId] = new_rid
for el in dest.shapes._spTree.iter():
for attr, val in el.attrib.items():
if attr.startswith(R_EMBED) and val in rid_map:
el.set(attr, rid_map[val])
if source.has_notes_slide:
dest.notes_slide.notes_text_frame.text = (
source.notes_slide.notes_text_frame.text)
return len(prs.slides._sldIdLst) - 1
def set_background(slide, hex_color):
fill = slide.background.fill
fill.solid()
fill.fore_color.rgb = RGBColor.from_string(hex_color)
def add_hyperlink(prs, slide_idx, text, url):
"""Turn every run containing `text` on the slide into a hyperlink.
The link applies to the whole run (python-pptx links whole runs).
"""
slide = prs.slides[int(slide_idx)]
hits = 0
for shape in slide.shapes:
if not shape.has_text_frame:
continue
for para in shape.text_frame.paragraphs:
for run in para.runs:
if text in run.text:
run.hyperlink.address = url
hits += 1
if not hits:
raise SystemExit(f"no run containing {text!r} on slide {slide_idx}")
return hits
def _copy_layout_placeholder(slide, ph_idx):
"""Copy the layout placeholder with idx `ph_idx` onto the slide.
Slide-number (idx 12) and footer (idx 11) placeholders exist on the
layout but are not inherited by a slide until the slide carries its
own copy — this enables them. Returns the new shape, or None when the
layout does not provide that placeholder.
"""
for ph in slide.slide_layout.placeholders:
if ph.placeholder_format.idx == ph_idx:
el = copy.deepcopy(ph._element)
slide.shapes._spTree.append(el)
for shape in slide.placeholders:
if shape.placeholder_format.idx == ph_idx:
return shape
return None
return None
def enable_slide_number(slide):
if any(ph.placeholder_format.idx == 12 for ph in slide.placeholders):
return True
return _copy_layout_placeholder(slide, 12) is not None
def set_footer(slide, text):
shape = next((ph for ph in slide.placeholders
if ph.placeholder_format.idx == 11), None)
if shape is None:
shape = _copy_layout_placeholder(slide, 11)
if shape is None:
raise SystemExit("layout provides no footer placeholder; add a "
"textbox instead")
shape.text_frame.text = text
return True
def set_notes(slide, text, append=False):
tf = slide.notes_slide.notes_text_frame
if append and tf.text:
para = tf.add_paragraph()
para.text = text
else:
tf.text = text
def main(argv=None):
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
parser = argparse.ArgumentParser(
description="Edit a .pptx: replace text, update chart data, swap "
"images, duplicate/remove/reorder slides, backgrounds, "
"hyperlinks, footers, slide numbers, speaker notes.",
epilog=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument("pptx", help="path to the .pptx file")
parser.add_argument("--output", help="save to this path instead of "
"overwriting the input")
parser.add_argument("--replace-text", nargs=2, action="append",
metavar=("OLD", "NEW"), default=[])
parser.add_argument("--chart-data", metavar="SPEC_JSON")
parser.add_argument("--swap-image", nargs=3,
metavar=("SLIDE", "SHAPE_NAME", "IMAGE"))
parser.add_argument("--remove-slide", type=int, metavar="N")
parser.add_argument("--move-slide", nargs=2, type=int,
metavar=("FROM", "TO"))
parser.add_argument("--duplicate-slide", type=int, metavar="N")
parser.add_argument("--set-background", nargs=2,
metavar=("SLIDE", "HEX"))
parser.add_argument("--hyperlink", nargs=3,
metavar=("SLIDE", "TEXT", "URL"))
parser.add_argument("--enable-slide-number", type=int, metavar="N")
parser.add_argument("--set-footer", nargs=2, metavar=("SLIDE", "TEXT"))
parser.add_argument("--set-notes", nargs=2, metavar=("SLIDE", "TEXT"))
parser.add_argument("--append-notes", nargs=2, metavar=("SLIDE", "TEXT"))
args = parser.parse_args(argv)
prs = Presentation(args.pptx)
report = {"ok": True, "replacements": 0}
for old, new in args.replace_text:
report["replacements"] += replace_text(prs, old, new)
if args.chart_data:
update_chart(prs, args.chart_data)
report["chart_updated"] = True
if args.swap_image:
swap_image(prs, *args.swap_image)
report["image_swapped"] = True
if args.duplicate_slide is not None:
report["duplicated_to"] = duplicate_slide(prs, args.duplicate_slide)
if args.set_background:
set_background(prs.slides[int(args.set_background[0])],
args.set_background[1])
report["background_set"] = True
if args.hyperlink:
report["hyperlinked_runs"] = add_hyperlink(prs, *args.hyperlink)
if args.enable_slide_number is not None:
report["slide_number_enabled"] = enable_slide_number(
prs.slides[args.enable_slide_number])
if args.set_footer:
set_footer(prs.slides[int(args.set_footer[0])], args.set_footer[1])
report["footer_set"] = True
if args.set_notes:
set_notes(prs.slides[int(args.set_notes[0])], args.set_notes[1])
report["notes_set"] = True
if args.append_notes:
set_notes(prs.slides[int(args.append_notes[0])],
args.append_notes[1], append=True)
report["notes_appended"] = True
if args.remove_slide is not None:
remove_slide(prs, args.remove_slide)
report["slide_removed"] = args.remove_slide
if args.move_slide:
move_slide(prs, *args.move_slide)
report["slide_moved"] = args.move_slide
out = args.output or args.pptx
prs.save(out)
report["output"] = out
print(json.dumps(report))
return 0
if __name__ == "__main__":
sys.exit(main())
@@ -0,0 +1,88 @@
#!/usr/bin/env python3
"""Build a deck from a .pptx template (brand deck) and fill placeholders.
Two modes:
1) Token fill (default): open TEMPLATE, replace every {{token}} across
slides, tables, and notes using --values JSON ({"token": "value"}),
save to OUTPUT. Formatting of the token's run is preserved.
2) --add-slides SPEC.json: additionally append slides built from the
template's own layouts (referenced by layout name or index), so new
slides inherit the brand master. Spec:
{"slides": [{"layout": "Title and Content", "title": "New",
"bullets": ["a", {"text": "b", "level": 1}],
"notes": "presenter text"}]}
"""
import argparse
import json
import sys
from pptx import Presentation
def fill_tokens(prs, values):
from pptx_edit import replace_text # same scripts/ directory
total = 0
for token, value in values.items():
total += replace_text(prs, "{{%s}}" % token, str(value))
return total
def find_layout(prs, ref):
if isinstance(ref, int):
return prs.slide_layouts[ref]
for layout in prs.slide_layouts:
if layout.name == ref:
return layout
raise SystemExit(f"layout {ref!r} not found; available: "
f"{[la.name for la in prs.slide_layouts]}")
def add_slides(prs, spec):
from pptx_create import add_bullets
for slide_spec in spec.get("slides", []):
layout = find_layout(prs, slide_spec.get("layout", 1))
slide = prs.slides.add_slide(layout)
if slide_spec.get("title") is not None and slide.shapes.title:
slide.shapes.title.text = slide_spec["title"]
if slide_spec.get("bullets"):
body = next((ph for ph in slide.placeholders
if ph.placeholder_format.idx != 0), None)
if body is not None:
add_bullets(body.text_frame, slide_spec["bullets"])
if slide_spec.get("notes"):
slide.notes_slide.notes_text_frame.text = slide_spec["notes"]
def main(argv=None):
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
parser = argparse.ArgumentParser(
description="Fill {{tokens}} in a .pptx template and optionally "
"append slides using the template's own layouts.",
epilog=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument("template", help="path to the template .pptx")
parser.add_argument("output", help="output .pptx path")
parser.add_argument("--values", metavar="JSON",
help="JSON file mapping token -> replacement value")
parser.add_argument("--add-slides", metavar="SPEC_JSON",
help="JSON spec of slides to append")
args = parser.parse_args(argv)
prs = Presentation(args.template)
filled = 0
if args.values:
with open(args.values, encoding="utf-8") as fh:
filled = fill_tokens(prs, json.load(fh))
if args.add_slides:
with open(args.add_slides, encoding="utf-8") as fh:
add_slides(prs, json.load(fh))
prs.save(args.output)
print(json.dumps({"ok": True, "output": args.output,
"tokens_filled": filled}))
return 0
if __name__ == "__main__":
import os
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
sys.exit(main())
@@ -0,0 +1,131 @@
#!/usr/bin/env python3
"""Read a .pptx file: JSON outline, notes, or export embedded images.
Modes:
--outline JSON with per-slide layout, texts, tables, notes,
chart data, and image inventory (default mode).
--notes JSON list of speaker notes per slide.
--images DIR Export every embedded picture to DIR as files.
"""
import argparse
import json
import os
import sys
from pptx import Presentation
from pptx.enum.shapes import MSO_SHAPE_TYPE
from pptx.util import Emu
def iter_shapes(shapes):
"""Yield shapes, descending into groups."""
for shape in shapes:
if shape.shape_type == MSO_SHAPE_TYPE.GROUP:
yield from iter_shapes(shape.shapes)
else:
yield shape
def chart_info(chart):
info = {"type": str(chart.chart_type),
"categories": [str(c) for c in chart.plots[0].categories],
"series": []}
for plot in chart.plots:
for series in plot.series:
try:
name = series.name
except (AttributeError, KeyError):
name = None
info["series"].append({"name": name,
"values": list(series.values)})
return info
def slide_record(index, slide):
rec = {"index": index, "layout": slide.slide_layout.name,
"texts": [], "tables": [], "images": [], "charts": [],
"notes": None}
for shape in iter_shapes(slide.shapes):
if shape.has_text_frame and shape.text_frame.text.strip():
rec["texts"].append(shape.text_frame.text)
if shape.has_table:
rec["tables"].append(
[[cell.text for cell in row.cells]
for row in shape.table.rows])
if shape.shape_type == MSO_SHAPE_TYPE.PICTURE:
try:
img = shape.image
rec["images"].append({"filename": img.filename,
"ext": img.ext,
"size_bytes": len(img.blob)})
except (KeyError, ValueError):
rec["images"].append({"filename": None, "ext": None,
"size_bytes": None,
"note": "linked or unreadable"})
if shape.has_chart:
rec["charts"].append(chart_info(shape.chart))
if slide.has_notes_slide:
rec["notes"] = slide.notes_slide.notes_text_frame.text
return rec
def export_images(prs, out_dir):
os.makedirs(out_dir, exist_ok=True)
written = []
for i, slide in enumerate(prs.slides):
for j, shape in enumerate(iter_shapes(slide.shapes)):
if shape.shape_type != MSO_SHAPE_TYPE.PICTURE:
continue
try:
img = shape.image
except (KeyError, ValueError):
continue
path = os.path.join(out_dir, f"slide{i}_img{j}.{img.ext}")
with open(path, "wb") as fh:
fh.write(img.blob)
written.append(path)
return written
def main(argv=None):
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
parser = argparse.ArgumentParser(
description="Read a .pptx: outline/notes as JSON, export images.")
parser.add_argument("pptx", help="path to the .pptx file")
parser.add_argument("--outline", action="store_true",
help="print full JSON outline (default)")
parser.add_argument("--notes", action="store_true",
help="print speaker notes only")
parser.add_argument("--images", metavar="DIR",
help="export embedded images into DIR")
args = parser.parse_args(argv)
prs = Presentation(args.pptx)
if args.images:
written = export_images(prs, args.images)
print(json.dumps({"ok": True, "exported": written}, indent=2))
return 0
if args.notes:
notes = [slide.notes_slide.notes_text_frame.text
if slide.has_notes_slide else None
for slide in prs.slides]
print(json.dumps({"ok": True, "notes": notes},
indent=2, ensure_ascii=True))
return 0
outline = {
"ok": True,
"slide_size_inches": [round(Emu(prs.slide_width).inches, 3),
round(Emu(prs.slide_height).inches, 3)],
"slide_count": len(prs.slides._sldIdLst),
"layouts_available": [lay.name for lay in prs.slide_layouts],
"slides": [slide_record(i, s) for i, s in enumerate(prs.slides)],
}
print(json.dumps(outline, indent=2, ensure_ascii=True))
return 0
if __name__ == "__main__":
sys.exit(main())
@@ -0,0 +1,93 @@
#!/usr/bin/env python3
"""Render every slide of a .pptx to per-slide PNG images.
Pipeline: LibreOffice (soffice --headless --convert-to pdf) turns the deck
into a PDF, then poppler (pdftoppm, or pdftocairo as an alternate) splits
the PDF into one PNG per slide.
Output is JSON. When both tools are present:
{"rendered": true, "files": ["render/slide-1.png", ...]}
When either tool is missing the script still exits 0 and reports:
{"rendered": false, "missing": ["soffice"], "guidance": "..."}
so callers can degrade gracefully (fall back to pptx_read.py --outline).
"""
import argparse
import glob
import json
import os
import shutil
import subprocess
import sys
import tempfile
def find_tools():
"""Return (soffice, splitter, missing) using shutil.which."""
soffice = shutil.which("soffice")
splitter = shutil.which("pdftoppm") or shutil.which("pdftocairo")
missing = []
if not soffice:
missing.append("soffice")
if not splitter:
missing.append("pdftoppm (or pdftocairo)")
return soffice, splitter, missing
def render(pptx_path, out_dir, prefix, dpi):
soffice, splitter, missing = find_tools()
if missing:
return {
"rendered": False, "missing": missing,
"guidance": "Install LibreOffice (soffice) and poppler-utils "
"(pdftoppm/pdftocairo) to render slides. Without "
"them, verify decks with pptx_read.py --outline "
"instead.",
}
os.makedirs(out_dir, exist_ok=True)
with tempfile.TemporaryDirectory() as tmp:
proc = subprocess.run(
[soffice, "--headless", "--convert-to", "pdf",
"--outdir", tmp, pptx_path],
capture_output=True, text=True, encoding="utf-8",
errors="replace", timeout=300)
pdfs = glob.glob(os.path.join(tmp, "*.pdf"))
if proc.returncode != 0 or not pdfs:
raise SystemExit(f"soffice PDF conversion failed: {proc.stderr}")
pdf = pdfs[0]
out_prefix = os.path.join(out_dir, prefix)
proc = subprocess.run(
[splitter, "-png", "-r", str(dpi), pdf, out_prefix],
capture_output=True, text=True, encoding="utf-8",
errors="replace", timeout=300)
if proc.returncode != 0:
raise SystemExit(f"{os.path.basename(splitter)} failed: "
f"{proc.stderr}")
files = sorted(glob.glob(out_prefix + "*.png"))
return {"rendered": True, "files": files}
def main(argv=None):
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
parser = argparse.ArgumentParser(
description="Render each slide of a .pptx to a PNG via "
"soffice + pdftoppm/pdftocairo.",
epilog=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument("pptx", help="path to the .pptx file")
parser.add_argument("--outdir", default="render",
help="directory for PNGs (default: ./render)")
parser.add_argument("--prefix", default="slide",
help="PNG filename prefix (default: slide)")
parser.add_argument("--dpi", type=int, default=100,
help="render resolution (default: 100)")
args = parser.parse_args(argv)
result = render(args.pptx, args.outdir, args.prefix, args.dpi)
print(json.dumps(result, indent=2))
return 0
if __name__ == "__main__":
sys.exit(main())