python-substack 0.1.22__tar.gz → 0.1.24__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: python-substack
3
- Version: 0.1.22
3
+ Version: 0.1.24
4
4
  Summary: A Python wrapper around the Substack API.
5
5
  License: MIT
6
6
  License-File: LICENSE
@@ -16,6 +16,8 @@ Classifier: Programming Language :: Python :: 3.12
16
16
  Classifier: Programming Language :: Python :: 3.13
17
17
  Classifier: Programming Language :: Python :: 3.14
18
18
  Requires-Dist: PyYAML (>=6.0,<7.0)
19
+ Requires-Dist: markdown-it-py (>=3.0,<4.0)
20
+ Requires-Dist: mdit-py-plugins (>=0.4,<0.5)
19
21
  Requires-Dist: python-dotenv (>=1.2.1,<2.0.0)
20
22
  Requires-Dist: requests (>=2.32.0,<3.0.0)
21
23
  Project-URL: Homepage, https://github.com/ma2za/python-substack
@@ -180,6 +182,22 @@ This is a paragraph with **bold** and *italic* text.
180
182
  """
181
183
  post.from_markdown(markdown_content, api=api)
182
184
 
185
+ # Markdown footnotes are supported too. References become inline anchors and
186
+ # definitions become footnote blocks, numbered by order of first appearance.
187
+ # Labels can be numbers or names (e.g. [^1] or [^source]).
188
+ footnote_markdown = """
189
+ A claim that needs support.[^1] Another, with a named label.[^source]
190
+
191
+ [^1]: The supporting detail, with a [link](https://example.com).
192
+ [^source]: Author, *Title* (2025).
193
+ """
194
+ post.from_markdown(footnote_markdown, api=api)
195
+
196
+ # Or build footnotes manually:
197
+ post.paragraph(content=[{"content": "Some claim."}]).footnote_anchor(1)
198
+ post.footnote(1, "The note text, with **formatting** allowed.")
199
+
200
+
183
201
  draft = api.post_draft(post.get_draft())
184
202
 
185
203
  # set section (can only be done after first posting the draft)
@@ -156,6 +156,22 @@ This is a paragraph with **bold** and *italic* text.
156
156
  """
157
157
  post.from_markdown(markdown_content, api=api)
158
158
 
159
+ # Markdown footnotes are supported too. References become inline anchors and
160
+ # definitions become footnote blocks, numbered by order of first appearance.
161
+ # Labels can be numbers or names (e.g. [^1] or [^source]).
162
+ footnote_markdown = """
163
+ A claim that needs support.[^1] Another, with a named label.[^source]
164
+
165
+ [^1]: The supporting detail, with a [link](https://example.com).
166
+ [^source]: Author, *Title* (2025).
167
+ """
168
+ post.from_markdown(footnote_markdown, api=api)
169
+
170
+ # Or build footnotes manually:
171
+ post.paragraph(content=[{"content": "Some claim."}]).footnote_anchor(1)
172
+ post.footnote(1, "The note text, with **formatting** allowed.")
173
+
174
+
159
175
  draft = api.post_draft(post.get_draft())
160
176
 
161
177
  # set section (can only be done after first posting the draft)
@@ -1,6 +1,6 @@
1
1
  [tool.poetry]
2
2
  name = "python-substack"
3
- version = "0.1.22"
3
+ version = "0.1.24"
4
4
  description = "A Python wrapper around the Substack API."
5
5
  authors = ["Paolo Mazza <mazzapaolo2019@gmail.com>"]
6
6
  license = "MIT"
@@ -21,6 +21,8 @@ python = "<4.0,>=3.10"
21
21
  requests = "^2.32.0"
22
22
  python-dotenv = "^1.2.1"
23
23
  PyYAML = "^6.0"
24
+ markdown-it-py = "^3.0"
25
+ mdit-py-plugins = "^0.4"
24
26
 
25
27
  [tool.poetry.group.dev.dependencies]
26
28
 
@@ -3,7 +3,7 @@
3
3
  __author__ = "Paolo Mazza"
4
4
  __email__ = "mazzapaolo2019@gmail.com"
5
5
  __license__ = "MIT License"
6
- __version__ = "0.1.21"
6
+ __version__ = "0.1.24"
7
7
  __url__ = "https://github.com/ma2za/python-substack"
8
8
  __download_url__ = "https://pypi.python.org/pypi/python-substack"
9
9
  __description__ = "A Python wrapper around the Substack API"
@@ -0,0 +1,210 @@
1
+ """Markdown -> Substack ProseMirror via markdown-it-py.
2
+
3
+ Implements Post.from_markdown() using a real CommonMark parser (markdown-it-py)
4
+ plus the standard footnote plugin, with a small renderer that walks the syntax
5
+ tree into Substack's node schema.
6
+
7
+ Node construction goes through ``substack.nodes`` so the (undocumented) schema
8
+ lives in exactly one place.
9
+
10
+ Footnotes: Substack numbers footnote anchors by their position in the document
11
+ and pairs them one-to-one, in order, with the footnote blocks at the end (it
12
+ ignores any explicit number and does not support one block serving several
13
+ anchors). So each reference is emitted as its own sequentially-numbered anchor,
14
+ and a matching footnote block is appended for each -- a definition referenced
15
+ more than once is duplicated, which mirrors how Substack's own editor behaves.
16
+ """
17
+
18
+ from __future__ import annotations
19
+
20
+ import copy
21
+ from typing import Dict, List, Optional
22
+
23
+ from markdown_it import MarkdownIt
24
+ from markdown_it.tree import SyntaxTreeNode
25
+ from mdit_py_plugins.footnote import footnote_plugin
26
+
27
+ from substack import nodes
28
+ from substack.nodes import MarkType, NodeType
29
+
30
+ _MARK_FOR = {
31
+ "strong": {"type": MarkType.STRONG},
32
+ "em": {"type": MarkType.EM},
33
+ "s": {"type": MarkType.STRIKETHROUGH},
34
+ }
35
+
36
+
37
+ def _make_parser() -> MarkdownIt:
38
+ return MarkdownIt("commonmark").use(footnote_plugin).enable("strikethrough")
39
+
40
+
41
+ def _coalesce(out_nodes: List[Dict]) -> List[Dict]:
42
+ """Merge adjacent text nodes that carry identical marks (e.g. softbreaks)."""
43
+ merged: List[Dict] = []
44
+ for node in out_nodes:
45
+ if (
46
+ merged
47
+ and node.get("type") == NodeType.TEXT
48
+ and merged[-1].get("type") == NodeType.TEXT
49
+ and node.get("marks") == merged[-1].get("marks")
50
+ ):
51
+ merged[-1]["text"] += node["text"]
52
+ else:
53
+ merged.append(node)
54
+ return merged
55
+
56
+
57
+ def _render_inline(node: SyntaxTreeNode, marks: List[Dict], ctx: Dict) -> List[Dict]:
58
+ """Render an inline subtree into a flat list of text / anchor nodes."""
59
+ out: List[Dict] = []
60
+ for child in node.children:
61
+ t = child.type
62
+ if t == "text":
63
+ if child.content:
64
+ out.append(nodes.text(child.content, marks))
65
+ elif t == "code_inline":
66
+ out.append(nodes.text(child.content, marks + [nodes.code_mark()]))
67
+ elif t in _MARK_FOR:
68
+ out.extend(_render_inline(child, marks + [_MARK_FOR[t]], ctx))
69
+ elif t == "link":
70
+ href = child.attrs.get("href", "")
71
+ out.extend(_render_inline(child, marks + [nodes.link_mark(href)], ctx))
72
+ elif t in ("softbreak", "hardbreak"):
73
+ out.append(nodes.text(" ", marks))
74
+ elif t == "footnote_ref":
75
+ # Number anchors by document position and record which definition each
76
+ # one points to, so matching blocks can be emitted 1:1 afterwards.
77
+ ctx["order"].append(child.meta["id"])
78
+ out.append(nodes.footnote_anchor(len(ctx["order"])))
79
+ elif t == "image":
80
+ # Inline images are rare in this schema; fall back to alt text.
81
+ alt = child.attrs.get("alt") or "".join(
82
+ c.content for c in child.children if c.type == "text"
83
+ )
84
+ if alt:
85
+ out.append(nodes.text(alt, marks))
86
+ return _coalesce(out)
87
+
88
+
89
+ def _only_image(inline: SyntaxTreeNode) -> Optional[SyntaxTreeNode]:
90
+ """If an inline node is just an image (optionally wrapped in a link), return it."""
91
+ kids = [c for c in inline.children if c.type != "softbreak"]
92
+ if len(kids) == 1 and kids[0].type == "image":
93
+ return kids[0]
94
+ if len(kids) == 1 and kids[0].type == "link":
95
+ inner = [c for c in kids[0].children if c.type != "softbreak"]
96
+ if len(inner) == 1 and inner[0].type == "image":
97
+ img = inner[0]
98
+ img._link_href = kids[0].attrs.get("href") # type: ignore[attr-defined]
99
+ return img
100
+ return None
101
+
102
+
103
+ def _captioned_image(img: SyntaxTreeNode, api) -> Dict:
104
+ src = img.attrs.get("src", "")
105
+ if src.startswith("/"):
106
+ src = src[1:]
107
+ if api is not None and not src.startswith("http"):
108
+ try:
109
+ src = api.get_image(src).get("url")
110
+ except Exception:
111
+ pass
112
+ # markdown-it stores the image alt text as the node's content, not in attrs.
113
+ alt = img.content or img.attrs.get("alt") or None
114
+ # Standard markdown image title `![alt](src "caption")` maps to Substack's caption node.
115
+ title = img.attrs.get("title") or None
116
+ caption = [nodes.text(title)] if title else None
117
+ return nodes.captioned_image(
118
+ src,
119
+ alt=alt,
120
+ href=getattr(img, "_link_href", None),
121
+ caption=caption,
122
+ )
123
+
124
+
125
+ def _render_block(node: SyntaxTreeNode, api, ctx: Dict) -> List[Dict]:
126
+ """Render a block-level node into zero or more Substack nodes."""
127
+ t = node.type
128
+
129
+ if t == "paragraph":
130
+ inline = node.children[0]
131
+ img = _only_image(inline)
132
+ if img is not None:
133
+ return [_captioned_image(img, api)]
134
+ return [nodes.paragraph(_render_inline(inline, [], ctx))]
135
+
136
+ if t == "heading":
137
+ level = int(node.tag[1])
138
+ return [nodes.heading(_render_inline(node.children[0], [], ctx), level=level)]
139
+
140
+ if t == "hr":
141
+ return [nodes.horizontal_rule()]
142
+
143
+ if t in ("fence", "code_block"):
144
+ return [
145
+ nodes.code_block(
146
+ node.content.rstrip("\n"), language=node.info.strip() or None
147
+ )
148
+ ]
149
+
150
+ if t == "blockquote":
151
+ paras: List[Dict] = []
152
+ for child in node.children:
153
+ paras.extend(_render_block(child, api, ctx))
154
+ return [nodes.blockquote(paras)]
155
+
156
+ if t == "bullet_list":
157
+ return [nodes.bullet_list(_render_list_items(node, api, ctx))]
158
+
159
+ if t == "ordered_list":
160
+ return [nodes.ordered_list(_render_list_items(node, api, ctx))]
161
+
162
+ # footnote_block is handled separately in markdown_to_doc; ignore it here.
163
+ return []
164
+
165
+
166
+ def _render_list_items(list_node: SyntaxTreeNode, api, ctx: Dict) -> List[Dict]:
167
+ items = []
168
+ for li in list_node.children:
169
+ content: List[Dict] = []
170
+ for child in li.children:
171
+ content.extend(_render_block(child, api, ctx))
172
+ items.append({"type": NodeType.LIST_ITEM, "content": content})
173
+ return items
174
+
175
+
176
+ def _footnote_definitions(tree: SyntaxTreeNode, api) -> Dict[int, List[Dict]]:
177
+ """Map each footnote id to its rendered block content."""
178
+ definitions: Dict[int, List[Dict]] = {}
179
+ for node in tree.children:
180
+ if node.type != "footnote_block":
181
+ continue
182
+ for fn in node.children:
183
+ # A footnote's own content should not register anchors of its own.
184
+ local_ctx = {"order": []}
185
+ content: List[Dict] = []
186
+ for child in fn.children:
187
+ content.extend(_render_block(child, api, local_ctx))
188
+ definitions[fn.meta["id"]] = content
189
+ return definitions
190
+
191
+
192
+ def markdown_to_doc(markdown_content: str, api=None) -> List[Dict]:
193
+ """Convert Markdown into a list of Substack ProseMirror block nodes."""
194
+ tree = SyntaxTreeNode(_make_parser().parse(markdown_content))
195
+
196
+ definitions = _footnote_definitions(tree, api)
197
+
198
+ ctx: Dict = {"order": []}
199
+ out: List[Dict] = []
200
+ for node in tree.children:
201
+ if node.type == "footnote_block":
202
+ continue
203
+ out.extend(_render_block(node, api, ctx))
204
+
205
+ # Emit one footnote block per reference, in anchor order, numbered to match.
206
+ for number, footnote_id in enumerate(ctx["order"], start=1):
207
+ content = copy.deepcopy(definitions.get(footnote_id, []))
208
+ out.append(nodes.footnote(number, content))
209
+
210
+ return out
@@ -0,0 +1,148 @@
1
+ """ProseMirror node builders for Substack documents.
2
+
3
+ Centralises the (undocumented) Substack ProseMirror schema in one place.
4
+ The node-type strings ("paragraph", "footnoteAnchor", "image2", ...) and
5
+ their shapes live here rather than as inline dict literals scattered across
6
+ post.py, giving:
7
+
8
+ * one source of truth for node shapes (so a schema change is a one-line fix),
9
+ * discoverable, typed constructors instead of bare dict literals,
10
+ * a natural seam for validation.
11
+
12
+ The builders intentionally return plain dicts so they stay 100% compatible with
13
+ the existing draft_body structure.
14
+ """
15
+
16
+ from __future__ import annotations
17
+
18
+ from typing import Dict, List, Optional
19
+
20
+
21
+ class NodeType:
22
+ DOC = "doc"
23
+ PARAGRAPH = "paragraph"
24
+ HEADING = "heading"
25
+ TEXT = "text"
26
+ BLOCKQUOTE = "blockquote"
27
+ CODE_BLOCK = "codeBlock"
28
+ HORIZONTAL_RULE = "horizontal_rule"
29
+ BULLET_LIST = "bullet_list"
30
+ ORDERED_LIST = "ordered_list"
31
+ LIST_ITEM = "list_item"
32
+ FOOTNOTE = "footnote"
33
+ FOOTNOTE_ANCHOR = "footnoteAnchor"
34
+ CAPTIONED_IMAGE = "captionedImage"
35
+ CAPTION = "caption"
36
+
37
+
38
+ class MarkType:
39
+ STRONG = "strong"
40
+ EM = "em"
41
+ CODE = "code"
42
+ STRIKETHROUGH = "strikethrough"
43
+ LINK = "link"
44
+
45
+
46
+ def code_mark() -> Dict:
47
+ return {"type": MarkType.CODE}
48
+
49
+
50
+ def text(value: str, marks: Optional[List[Dict]] = None) -> Dict:
51
+ node: Dict = {"type": NodeType.TEXT, "text": value}
52
+ if marks:
53
+ node["marks"] = marks
54
+ return node
55
+
56
+
57
+ def link_mark(href: str) -> Dict:
58
+ return {"type": MarkType.LINK, "attrs": {"href": href}}
59
+
60
+
61
+ def paragraph(content: Optional[List[Dict]] = None) -> Dict:
62
+ return {"type": NodeType.PARAGRAPH, "content": content or []}
63
+
64
+
65
+ def heading(content: List[Dict], level: int = 1) -> Dict:
66
+ return {"type": NodeType.HEADING, "content": content, "attrs": {"level": level}}
67
+
68
+
69
+ def horizontal_rule() -> Dict:
70
+ return {"type": NodeType.HORIZONTAL_RULE}
71
+
72
+
73
+ def blockquote(paragraphs: List[Dict]) -> Dict:
74
+ node: Dict = {"type": NodeType.BLOCKQUOTE}
75
+ if paragraphs:
76
+ node["content"] = paragraphs
77
+ return node
78
+
79
+
80
+ def list_item(content_nodes: List[Dict]) -> Dict:
81
+ return {
82
+ "type": NodeType.LIST_ITEM,
83
+ "content": [paragraph(content_nodes)],
84
+ }
85
+
86
+
87
+ def bullet_list(items: List[Dict]) -> Dict:
88
+ return {"type": NodeType.BULLET_LIST, "content": items}
89
+
90
+
91
+ def ordered_list(items: List[Dict]) -> Dict:
92
+ return {"type": NodeType.ORDERED_LIST, "content": items}
93
+
94
+
95
+ def code_block(code: str, language: Optional[str] = None) -> Dict:
96
+ node: Dict = {"type": NodeType.CODE_BLOCK, "content": [text(code)]}
97
+ if language:
98
+ node["attrs"] = {"language": language}
99
+ return node
100
+
101
+
102
+ def captioned_image(
103
+ src: str,
104
+ alt: Optional[str] = None,
105
+ href: Optional[str] = None,
106
+ caption: Optional[List[Dict]] = None,
107
+ image_size: str = "normal",
108
+ ) -> Dict:
109
+ content: List[Dict] = [
110
+ {
111
+ "type": "image2",
112
+ "attrs": {
113
+ "src": src,
114
+ "srcNoWatermark": None,
115
+ "fullscreen": False,
116
+ "imageSize": image_size,
117
+ "height": 819,
118
+ "width": 1456,
119
+ "resizeWidth": 728 if image_size == "normal" else None,
120
+ "bytes": None,
121
+ "alt": alt,
122
+ "title": None,
123
+ "type": None,
124
+ "href": href,
125
+ "belowTheFold": False,
126
+ "topImage": False,
127
+ "internalRedirect": None,
128
+ "isProcessing": False,
129
+ "align": None,
130
+ "offset": False,
131
+ },
132
+ }
133
+ ]
134
+ if caption:
135
+ content.append({"type": NodeType.CAPTION, "content": caption})
136
+ return {"type": NodeType.CAPTIONED_IMAGE, "content": content}
137
+
138
+
139
+ def footnote_anchor(number: int) -> Dict:
140
+ return {"type": NodeType.FOOTNOTE_ANCHOR, "attrs": {"number": number}}
141
+
142
+
143
+ def footnote(number: int, paragraphs: List[Dict]) -> Dict:
144
+ return {
145
+ "type": NodeType.FOOTNOTE,
146
+ "attrs": {"number": number},
147
+ "content": paragraphs or [paragraph()],
148
+ }
@@ -11,6 +11,7 @@ from typing import Dict, List
11
11
  __all__ = ["Post", "parse_inline", "tokens_to_text_nodes"]
12
12
 
13
13
  from substack.exceptions import SectionNotExistsException
14
+ from substack import nodes
14
15
 
15
16
 
16
17
  def tokens_to_text_nodes(tokens: List[Dict]) -> List[Dict]:
@@ -543,6 +544,58 @@ class Post:
543
544
 
544
545
  return self
545
546
 
547
+ def footnote_anchor(self, number: int):
548
+ """
549
+
550
+ Add an inline footnote reference (the superscript marker) to the last block.
551
+
552
+ Args:
553
+ number: The footnote number this anchor points to.
554
+
555
+ Returns:
556
+ Self for method chaining.
557
+
558
+ """
559
+ content = self.draft_body["content"][-1].get("content", [])
560
+ content += [nodes.footnote_anchor(number)]
561
+ self.draft_body["content"][-1]["content"] = content
562
+ return self
563
+
564
+ def footnote(self, number: int, content=None):
565
+ """
566
+
567
+ Append a footnote block (the note shown at the foot of the post).
568
+
569
+ Args:
570
+ number: The footnote number, matching a footnote_anchor.
571
+ content: Text string or list of inline token dicts. A plain string is
572
+ parsed for inline Markdown and may contain blank-line-separated
573
+ paragraphs; a parse_inline() token list or a list of ready text
574
+ nodes is also accepted (single paragraph).
575
+
576
+ Returns:
577
+ Self for method chaining.
578
+
579
+ """
580
+ paragraphs: List[Dict] = []
581
+ if isinstance(content, str):
582
+ # Blank lines separate paragraphs within the footnote.
583
+ for chunk in re.split(r"\n\s*\n", content):
584
+ chunk = chunk.strip()
585
+ if chunk:
586
+ paragraphs.append(nodes.paragraph(tokens_to_text_nodes(parse_inline(chunk))))
587
+ elif isinstance(content, list):
588
+ # Accept either parse_inline tokens ({"content": ...}) or text nodes.
589
+ if content and content[0].get("type") == "text":
590
+ text_nodes = content
591
+ else:
592
+ text_nodes = tokens_to_text_nodes(content)
593
+ paragraphs.append(nodes.paragraph(text_nodes))
594
+
595
+ node: Dict = nodes.footnote(number, paragraphs)
596
+ self.draft_body["content"] = self.draft_body.get("content", []) + [node]
597
+ return self
598
+
546
599
  def from_markdown(self, markdown_content: str, api=None):
547
600
  """
548
601
  Parse Markdown content and add it to the post.
@@ -559,6 +612,10 @@ class Post:
559
612
  - Ordered lists: Lines starting with '1.', '2.', etc.
560
613
  - Horizontal rules: Lines with ---, ***, or ___
561
614
  - Inline formatting: **bold**, *italic*, ***bold+italic***, `code`, ~~strikethrough~~
615
+ - Footnotes: ``text.[^label]`` references plus ``[^label]: definition``
616
+ lines. References become inline anchors and definitions become
617
+ footnote blocks, numbered by order of first appearance. Labels may be
618
+ numbers or names (e.g. ``[^1]`` or ``[^agi-book]``).
562
619
 
563
620
  Args:
564
621
  markdown_content: Markdown string to parse and add to the post.
@@ -572,276 +629,8 @@ class Post:
572
629
  >>> post = Post("Title", "Subtitle", user_id)
573
630
  >>> post.from_markdown("# Heading\\n\\nThis is **bold** text with [a link](https://example.com).")
574
631
  """
575
- lines = markdown_content.split("\n")
576
- blocks = []
577
- current_block: List[str] = []
578
- in_code_block = False
579
- code_block_language = None
580
-
581
- for line in lines:
582
- # Check for fenced code block start/end
583
- if line.strip().startswith("```"):
584
- if in_code_block:
585
- # End of code block
586
- if current_block:
587
- blocks.append({
588
- "type": "code",
589
- "language": code_block_language,
590
- "content": "\n".join(current_block)
591
- })
592
- current_block = []
593
- in_code_block = False
594
- code_block_language = None
595
- else:
596
- # Start of code block
597
- if current_block:
598
- blocks.append({"type": "text", "content": "\n".join(current_block)})
599
- current_block = []
600
- # Extract language if specified
601
- language = line.strip()[3:].strip()
602
- code_block_language = language if language else None
603
- in_code_block = True
604
- continue
605
-
606
- if in_code_block:
607
- # Inside code block - collect lines as-is
608
- current_block.append(line)
609
- else:
610
- # Regular content
611
- if line.strip() == "":
612
- # Empty line - end current block if it has content
613
- if current_block:
614
- blocks.append({"type": "text", "content": "\n".join(current_block)})
615
- current_block = []
616
- else:
617
- current_block.append(line)
618
-
619
- # Add any remaining content
620
- if current_block:
621
- if in_code_block:
622
- blocks.append({
623
- "type": "code",
624
- "language": code_block_language,
625
- "content": "\n".join(current_block)
626
- })
627
- else:
628
- blocks.append({"type": "text", "content": "\n".join(current_block)})
629
-
630
- # Process blocks
631
- for block in blocks:
632
- if block["type"] == "code":
633
- # Add code block
634
- code_content = block.get("content", "").strip()
635
- if code_content:
636
- # Substack uses "codeBlock" type
637
- code_attrs = {}
638
- if block.get("language"):
639
- code_attrs["language"] = block["language"]
640
- self.add({
641
- "type": "codeBlock",
642
- "content": code_content, # Pass as string, code_block method will handle it
643
- "attrs": code_attrs
644
- })
645
- else:
646
- # Process text block
647
- text_content = block.get("content", "").strip()
648
- if not text_content:
649
- continue
650
-
651
- # Check for horizontal rule: ---, ***, ___
652
- if re.match(r'^(\*{3,}|-{3,}|_{3,})\s*$', text_content):
653
- self.horizontal_rule()
654
- continue
655
-
656
- # Process headings (lines starting with '#' characters)
657
- if text_content.startswith("#"):
658
- level = len(text_content) - len(text_content.lstrip("#"))
659
- heading_text = text_content.lstrip("#").strip()
660
- if heading_text: # Only add if there's actual text
661
- self.heading(content=heading_text, level=min(level, 6))
662
-
663
- # Process images using Markdown image syntax: ![Alt](URL)
664
- # Also handle linked images: [![Alt](image_url)](link_url)
665
- elif text_content.startswith("!") or (text_content.startswith("[") and "![" in text_content):
666
- # Check for linked image first: [![alt](img)](link)
667
- linked_image_match = re.match(r'\[!\[([^\]]*)\]\(([^)]+)\)\]\(([^)]+)\)', text_content)
668
- if linked_image_match:
669
- # Linked image - create image with href
670
- alt_text = linked_image_match.group(1)
671
- image_url = linked_image_match.group(2)
672
- link_url = linked_image_match.group(3)
673
-
674
- # Adjust image URL if it starts with a slash
675
- image_url = image_url[1:] if image_url.startswith("/") else image_url
676
-
677
- # If api is provided and image_url is a local file, upload it
678
- if api is not None:
679
- try:
680
- image = api.get_image(image_url)
681
- image_url = image.get("url")
682
- except Exception:
683
- # If upload fails, use original URL
684
- pass
685
-
686
- self.add({
687
- "type": "captionedImage",
688
- "src": image_url,
689
- "alt": alt_text,
690
- "href": link_url
691
- })
692
- else:
693
- # Regular image: ![Alt](URL)
694
- match = re.match(r"!\[.*?\]\((.*?)\)", text_content)
695
- if match:
696
- image_url = match.group(1)
697
- # Adjust image URL if it starts with a slash
698
- image_url = image_url[1:] if image_url.startswith("/") else image_url
699
-
700
- # If api is provided and image_url is a local file, upload it
701
- if api is not None:
702
- try:
703
- image = api.get_image(image_url)
704
- image_url = image.get("url")
705
- except Exception:
706
- # If upload fails, use original URL
707
- pass
708
-
709
- self.add({"type": "captionedImage", "src": image_url})
710
-
711
- # Process paragraphs, bullet lists, ordered lists, or blockquotes
712
- else:
713
- if "\n" in text_content:
714
- # Process each line, grouping consecutive bullets/ordered items
715
- # into list nodes and consecutive blockquote lines into a
716
- # single blockquote node.
717
- pending_bullets: List[List[Dict]] = []
718
- pending_quotes: List[str] = []
719
- pending_ordered: List[List[Dict]] = []
720
-
721
- def flush_bullets():
722
- if not pending_bullets:
723
- return
724
- list_items = []
725
- for bullet_nodes in pending_bullets:
726
- list_items.append({
727
- "type": "list_item",
728
- "content": [{"type": "paragraph", "content": bullet_nodes}],
729
- })
730
- self.draft_body["content"].append(
731
- {"type": "bullet_list", "content": list_items}
732
- )
733
- pending_bullets.clear()
734
-
735
- def flush_quotes():
736
- if not pending_quotes:
737
- return
738
- paragraphs: List[Dict] = []
739
- for quote_line in pending_quotes:
740
- tokens = parse_inline(quote_line)
741
- text_nodes = tokens_to_text_nodes(tokens)
742
- if text_nodes:
743
- paragraphs.append({"type": "paragraph", "content": text_nodes})
744
- node: Dict = {"type": "blockquote"}
745
- if paragraphs:
746
- node["content"] = paragraphs
747
- self.draft_body["content"].append(node)
748
- pending_quotes.clear()
749
-
750
- def flush_ordered():
751
- if not pending_ordered:
752
- return
753
- list_items = []
754
- for item_nodes in pending_ordered:
755
- list_items.append({
756
- "type": "list_item",
757
- "content": [{"type": "paragraph", "content": item_nodes}],
758
- })
759
- self.draft_body["content"].append(
760
- {"type": "ordered_list", "content": list_items}
761
- )
762
- pending_ordered.clear()
763
-
764
- for line in text_content.split("\n"):
765
- line = line.strip()
766
- if not line:
767
- flush_bullets()
768
- flush_ordered()
769
- flush_quotes()
770
- continue
771
-
772
- # Check for blockquote marker
773
- if line.startswith("> ") or line == ">":
774
- flush_bullets()
775
- flush_ordered()
776
- quote_text = line[2:] if line.startswith("> ") else ""
777
- pending_quotes.append(quote_text)
778
- continue
779
-
780
- # Check for ordered list marker
781
- ordered_match = re.match(r'^(\d+)\.\s+(.*)', line)
782
- if ordered_match:
783
- flush_bullets()
784
- flush_quotes()
785
- item_text = ordered_match.group(2).strip()
786
- tokens = parse_inline(item_text)
787
- text_nodes = tokens_to_text_nodes(tokens)
788
- if text_nodes:
789
- pending_ordered.append(text_nodes)
790
- continue
791
-
792
- # Check for bullet marker
793
- bullet_text = None
794
- if line.startswith("* "):
795
- bullet_text = line[2:].strip()
796
- elif line.startswith("- "):
797
- bullet_text = line[2:].strip()
798
- elif line.startswith("*") and not line.startswith("**"):
799
- bullet_text = line[1:].strip()
800
-
801
- if bullet_text is not None:
802
- flush_ordered()
803
- flush_quotes()
804
- tokens = parse_inline(bullet_text)
805
- text_nodes = tokens_to_text_nodes(tokens)
806
- if text_nodes:
807
- pending_bullets.append(text_nodes)
808
- else:
809
- flush_bullets()
810
- flush_ordered()
811
- flush_quotes()
812
- tokens = parse_inline(line)
813
- self.add({"type": "paragraph", "content": tokens})
814
-
815
- flush_bullets()
816
- flush_ordered()
817
- flush_quotes()
818
- else:
819
- # Single line — blockquote, ordered list, or paragraph
820
- if text_content.startswith("> ") or text_content == ">":
821
- quote_text = text_content[2:] if text_content.startswith("> ") else ""
822
- tokens = parse_inline(quote_text)
823
- text_nodes = tokens_to_text_nodes(tokens)
824
- para = {"type": "paragraph", "content": text_nodes} if text_nodes else {"type": "paragraph"}
825
- self.draft_body["content"] = self.draft_body.get("content", []) + [
826
- {"type": "blockquote", "content": [para]}
827
- ]
828
-
829
- elif re.match(r'^(\d+)\.\s+(.*)', text_content):
830
- ordered_match = re.match(r'^(\d+)\.\s+(.*)', text_content)
831
- item_text = ordered_match.group(2).strip()
832
- tokens = parse_inline(item_text)
833
- text_nodes = tokens_to_text_nodes(tokens)
834
- if text_nodes:
835
- list_item = {
836
- "type": "list_item",
837
- "content": [{"type": "paragraph", "content": text_nodes}],
838
- }
839
- self.draft_body["content"].append(
840
- {"type": "ordered_list", "content": [list_item]}
841
- )
842
-
843
- else:
844
- tokens = parse_inline(text_content)
845
- self.add({"type": "paragraph", "content": tokens})
632
+ from substack import mdrender
846
633
 
634
+ rendered = mdrender.markdown_to_doc(markdown_content, api=api)
635
+ self.draft_body["content"] = self.draft_body.get("content", []) + rendered
847
636
  return self