MHoroszowski
diff --git a/‎CLAUDE.md‎
Lines changed: 2 additions & 1 deletion b/‎CLAUDE.md‎
Lines changed: 2 additions & 1 deletion
diff --git a/‎features/sld-comments.feature‎
Lines changed: 35 additions & 0 deletions b/‎features/sld-comments.feature‎
Lines changed: 35 additions & 0 deletions
diff --git a/‎features/steps/comments.py‎
Lines changed: 221 additions & 0 deletions b/‎features/steps/comments.py‎
Lines changed: 221 additions & 0 deletions
diff --git a/‎src/pptx/__init__.py‎
Lines changed: 14 additions & 0 deletions b/‎src/pptx/__init__.py‎
Lines changed: 14 additions & 0 deletions
@@ -56,8 +56,9 @@ Reference `git log 253dbc87 --format=%B -n1` for the canonical shape: scope, des
 
 Concretely:
 - Agents **author** the UAT script (it's part of the deliverable per §6.3).
-- Agents **may run** the UAT — once, after the §7 trinity is green — to verify the script doesn't crash, exits non-zero on failure, and actually asserts what it claims to assert. This is QA on the test itself, not acceptance of the feature.
+- Agents **may run** the UAT — after the §7 trinity is green — to verify the script doesn't crash, exits non-zero on failure, and actually asserts what it claims to assert. This is QA on the test itself, not acceptance of the feature.
 - Agents **must not report or imply UAT signoff** in commit messages, PR bodies, or summaries. Acceptable: *"UAT script runs clean — pending maintainer visual signoff in PowerPoint/Keynote."* Not acceptable: *"UAT: PASS — round-trip confirmed."*
+- UAT files should be stored under the `./uat` subdirectory.
 - The §7 reporting trinity below (pytest + ruff + behave) is the agent's full self-verification surface. UAT execution output may be included as an attachment but it is **not** the fourth gate.
 - Round-trip / behavioral evidence within agent-runnable scope still goes through **pytest integration tests** that exercise save+reopen. If pytest can fully cover it, the UAT is just a maintainer convenience; if pytest can't, UAT becomes the maintainer's primary acceptance path.
 - If unsure whether the maintainer has delegated signoff for a particular case, the agent **stops and asks**.
 
@@ -0,0 +1,35 @@
+Feature: Modern threaded comments on a slide (issue #25)
+  In order to round-trip review feedback in a presentation
+  As a developer using python-pptx
+  I need to add, read, reply to, and remove threaded comments on a slide
+
+  Scenario: Add a threaded comment and read it back after save/reopen
+    Given a blank slide with no comments
+    When I add a comment "Looks great" by author "Ada"
+    Then the slide has 1 comment
+    And after save and reopen the first comment text is "Looks great" by author "Ada"
+
+  Scenario: Reply to a comment threads under it and round-trips
+    Given a blank slide with no comments
+    When I add a comment "Question?" by author "Ada"
+    And I reply "Here is the answer" by author "Babbage"
+    Then after save and reopen the first comment has 1 reply with text "Here is the answer"
+
+  Scenario: Removing the last comment leaves a valid reopenable file
+    Given a blank slide with no comments
+    When I add a comment "Temporary" by author "Ada"
+    And I remove that comment
+    Then the slide has 0 comments
+    And after save and reopen the slide has 0 comments
+
+  Scenario: Resolving a threaded comment round-trips
+    Given a blank slide with no comments
+    When I add a comment "Please review" by author "Ada"
+    And I resolve that comment
+    Then after save and reopen the first comment is resolved
+
+  Scenario: A legacy comment and a modern comment coexist on round-trip
+    Given a slide carrying a pre-existing legacy comment
+    When I add a modern comment "Modern note" by author "Mary"
+    Then after save and reopen both the legacy and modern comments are present
+    And after save and reopen the legacy author ids are intact
@@ -0,0 +1,221 @@
+"""Gherkin step implementations for modern threaded comments (issue #25)."""
+
+from __future__ import annotations
+
+import io
+
+from behave import given, then, when
+
+from pptx import Presentation
+
+
+# given ===================================================
+
+
+@given("a blank slide with no comments")
+def given_blank_slide_no_comments(context):
+    prs = Presentation()
+    slide = prs.slides.add_slide(prs.slide_layouts[6])
+    context.prs = prs
+    context.slide = slide
+    assert len(slide.comments) == 0
+
+
+# when ====================================================
+
+
+@when('I add a comment "{text}" by author "{author}"')
+def when_add_comment(context, text, author):
+    context.comment = context.slide.comments.add(text, author)
+
+
+@when('I reply "{text}" by author "{author}"')
+def when_add_reply(context, text, author):
+    context.comment.replies.add(text, author)
+
+
+@when("I remove that comment")
+def when_remove_comment(context):
+    context.slide.comments.remove(context.comment)
+
+
+# then ====================================================
+
+
+@then("the slide has {count:d} comment")
+@then("the slide has {count:d} comments")
+def then_slide_has_n_comments(context, count):
+    assert len(context.slide.comments) == count, "expected %d comments, got %d" % (
+        count,
+        len(context.slide.comments),
+    )
+
+
+@then('after save and reopen the first comment text is "{text}" by author "{author}"')
+def then_reopen_first_comment(context, text, author):
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    comment = list(Presentation(stream).slides[0].comments)[0]
+    assert comment.text == text, "text was %r" % comment.text
+    assert comment.author == author, "author was %r" % comment.author
+
+
+@then('after save and reopen the first comment has {count:d} reply with text "{text}"')
+def then_reopen_reply(context, count, text):
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    comment = list(Presentation(stream).slides[0].comments)[0]
+    replies = list(comment.replies)
+    assert len(replies) == count, "expected %d replies, got %d" % (count, len(replies))
+    assert replies[0].text == text, "reply text was %r" % replies[0].text
+
+
+@then("after save and reopen the slide has {count:d} comments")
+def then_reopen_comment_count(context, count):
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    reopened = Presentation(stream).slides[0]
+    assert len(reopened.comments) == count, "expected %d comments after reopen, got %d" % (
+        count,
+        len(reopened.comments),
+    )
+
+
+# -- Wave 3: SF6 resolve + SF8 legacy/modern coexistence ------------------------
+
+
+@when("I resolve that comment")
+def when_resolve_comment(context):
+    context.comment.resolve()
+
+
+@then("after save and reopen the first comment is resolved")
+def then_reopen_first_comment_resolved(context):
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    comment = list(Presentation(stream).slides[0].comments)[0]
+    assert comment.resolved is True, "comment.resolved was %r" % comment.resolved
+
+
+def _build_legacy_plus_modern(prs0):
+    """Inject a legacy <p:cm> + author + slide rel into a saved package and
+    return the reopened Presentation (modern not yet added)."""
+    import zipfile
+
+    base = io.BytesIO()
+    prs0.save(base)
+    base.seek(0)
+
+    legacy_authors = (
+        '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>\n'
+        '<p:cmAuthorLst xmlns:p="http://schemas.openxmlformats.org/'
+        'presentationml/2006/main">'
+        '<p:cmAuthor id="0" name="Legacy Larry" initials="LL" '
+        'lastIdx="1" clrIdx="0"/></p:cmAuthorLst>'
+    )
+    legacy_comments = (
+        '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>\n'
+        '<p:cmLst xmlns:p="http://schemas.openxmlformats.org/'
+        'presentationml/2006/main">'
+        '<p:cm authorId="0" dt="2026-05-16T00:00:00" idx="1">'
+        '<p:pos x="100" y="200"/><p:text>Legacy feedback</p:text>'
+        "</p:cm></p:cmLst>"
+    )
+    src = zipfile.ZipFile(base, "r")
+    names = src.namelist()
+    ct = (
+        src.read("[Content_Types].xml")
+        .decode("utf-8")
+        .replace(
+            "</Types>",
+            '<Override PartName="/ppt/commentAuthors.xml" '
+            'ContentType="application/vnd.openxmlformats-officedocument.'
+            'presentationml.commentAuthors+xml"/>'
+            '<Override PartName="/ppt/comments/comment1.xml" '
+            'ContentType="application/vnd.openxmlformats-officedocument.'
+            'presentationml.comments+xml"/></Types>',
+        )
+    )
+    pres_rels = (
+        src.read("ppt/_rels/presentation.xml.rels")
+        .decode("utf-8")
+        .replace(
+            "</Relationships>",
+            '<Relationship Id="rIdLegacyAuth" '
+            'Type="http://schemas.openxmlformats.org/officeDocument/2006/'
+            'relationships/commentAuthors" Target="commentAuthors.xml"/>'
+            "</Relationships>",
+        )
+    )
+    slide_rels = (
+        src.read("ppt/slides/_rels/slide1.xml.rels")
+        .decode("utf-8")
+        .replace(
+            "</Relationships>",
+            '<Relationship Id="rIdLegacyCmt" '
+            'Type="http://schemas.openxmlformats.org/officeDocument/2006/'
+            'relationships/comments" Target="../comments/comment1.xml"/>'
+            "</Relationships>",
+        )
+    )
+    patched = {
+        "[Content_Types].xml": ct,
+        "ppt/_rels/presentation.xml.rels": pres_rels,
+        "ppt/slides/_rels/slide1.xml.rels": slide_rels,
+        "ppt/commentAuthors.xml": legacy_authors,
+        "ppt/comments/comment1.xml": legacy_comments,
+    }
+    out = io.BytesIO()
+    dst = zipfile.ZipFile(out, "w", zipfile.ZIP_DEFLATED)
+    for name in names:
+        dst.writestr(name, patched.get(name, src.read(name)))
+    for name, data in patched.items():
+        if name not in names:
+            dst.writestr(name, data)
+    src.close()
+    dst.close()
+    out.seek(0)
+    return Presentation(out)
+
+
+@given("a slide carrying a pre-existing legacy comment")
+def given_legacy_comment_slide(context):
+    prs0 = Presentation()
+    prs0.slides.add_slide(prs0.slide_layouts[6])
+    context.prs = _build_legacy_plus_modern(prs0)
+    context.slide = context.prs.slides[0]
+    assert any(c.is_legacy for c in context.slide.comments)
+
+
+@when('I add a modern comment "{text}" by author "{author}"')
+def when_add_modern_comment(context, text, author):
+    context.comment = context.slide.comments.add(text, author)
+
+
+@then("after save and reopen both the legacy and modern comments are present")
+def then_reopen_both_families(context):
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    comments = list(Presentation(stream).slides[0].comments)
+    texts = sorted(c.text for c in comments)
+    assert texts == ["Legacy feedback", "Modern note"], "texts were %r" % texts
+
+
+@then("after save and reopen the legacy author ids are intact")
+def then_reopen_legacy_author_ids(context):
+    import zipfile
+
+    stream = io.BytesIO()
+    context.prs.save(stream)
+    stream.seek(0)
+    zf = zipfile.ZipFile(stream, "r")
+    legacy_xml = zf.read("ppt/comments/comment1.xml").decode("utf-8")
+    authors_xml = zf.read("ppt/commentAuthors.xml").decode("utf-8")
+    zf.close()
+    assert 'authorId="0"' in legacy_xml, "legacy authorId mangled: %r" % legacy_xml
+    assert 'name="Legacy Larry"' in authors_xml, "legacy author lost: %r" % authors_xml
@@ -10,6 +10,12 @@
 from pptx.opc.constants import CONTENT_TYPE as CT
 from pptx.opc.package import PartFactory
 from pptx.parts.chart import ChartPart
+from pptx.parts.comments import (
+    AuthorsPart,
+    CommentAuthorsPart,
+    CommentsPart,
+    ModernCommentsPart,
+)
 from pptx.parts.coreprops import CorePropertiesPart
 from pptx.parts.custom_properties import CustomPropertiesPart
 from pptx.parts.custom_xml import CustomXmlPropertiesPart
@@ -41,6 +47,10 @@
     CT.PML_TEMPLATE_MAIN: PresentationPart,
     CT.PML_SLIDESHOW_MAIN: PresentationPart,
     CT.OPC_CORE_PROPERTIES: CorePropertiesPart,
+    CT.PML_COMMENT_AUTHORS: CommentAuthorsPart,
+    CT.PML_COMMENTS: CommentsPart,
+    CT.PML_THREADED_COMMENTS: ModernCommentsPart,
+    CT.PML_AUTHORS: AuthorsPart,
     CT.OFC_CUSTOM_PROPERTIES: CustomPropertiesPart,
     CT.OFC_CUSTOM_XML_PROPERTIES: CustomXmlPropertiesPart,
     # NOTE: CT.XML is intentionally NOT mapped to CustomXmlPart — see
@@ -80,6 +90,10 @@
 
 del (
     ChartPart,
+    AuthorsPart,
+    CommentAuthorsPart,
+    CommentsPart,
+    ModernCommentsPart,
     CorePropertiesPart,
     CustomPropertiesPart,
     CustomXmlPropertiesPart,