|
| 1 | +"""Converter from CHANGELOG.md (Markdown) to HTML suitable for a mypy blog post. |
| 2 | +
|
| 3 | +How to use: |
| 4 | +
|
| 5 | +1. Write release notes in CHANGELOG.md. |
| 6 | +2. Make sure the heading for the next release is of form `## Mypy X.Y`. |
| 7 | +2. Run `misc/gen_blog_post_html.py X.Y > target.html`. |
| 8 | +4. Manually inspect and tweak the result. |
| 9 | +
|
| 10 | +Notes: |
| 11 | +
|
| 12 | +* There are some fragile assumptions. Double check the output. |
| 13 | +""" |
| 14 | + |
| 15 | +import argparse |
| 16 | +import html |
| 17 | +import os |
| 18 | +import re |
| 19 | +import sys |
| 20 | + |
| 21 | + |
| 22 | +def format_lists(h: str) -> str: |
| 23 | + a = h.splitlines() |
| 24 | + r = [] |
| 25 | + i = 0 |
| 26 | + bullets = ("- ", "* ", " * ") |
| 27 | + while i < len(a): |
| 28 | + if a[i].startswith(bullets): |
| 29 | + r.append("<p><ul>") |
| 30 | + while i < len(a) and a[i].startswith(bullets): |
| 31 | + r.append("<li>%s" % a[i][2:].lstrip()) |
| 32 | + i += 1 |
| 33 | + r.append("</ul>") |
| 34 | + else: |
| 35 | + r.append(a[i]) |
| 36 | + i += 1 |
| 37 | + return "\n".join(r) |
| 38 | + |
| 39 | + |
| 40 | +def format_code(h: str) -> str: |
| 41 | + a = h.splitlines() |
| 42 | + r = [] |
| 43 | + i = 0 |
| 44 | + while i < len(a): |
| 45 | + if a[i].startswith(" ") or a[i].startswith("```"): |
| 46 | + indent = a[i].startswith(" ") |
| 47 | + if not indent: |
| 48 | + i += 1 |
| 49 | + r.append("<pre>") |
| 50 | + while i < len(a) and ( |
| 51 | + (indent and a[i].startswith(" ")) or (not indent and not a[i].startswith("```")) |
| 52 | + ): |
| 53 | + # Undo > and < |
| 54 | + line = a[i].replace(">", ">").replace("<", "<") |
| 55 | + if not indent: |
| 56 | + line = " " + line |
| 57 | + r.append(html.escape(line)) |
| 58 | + i += 1 |
| 59 | + r.append("</pre>") |
| 60 | + if not indent and a[i].startswith("```"): |
| 61 | + i += 1 |
| 62 | + else: |
| 63 | + r.append(a[i]) |
| 64 | + i += 1 |
| 65 | + return "\n".join(r) |
| 66 | + |
| 67 | + |
| 68 | +def convert(src: str) -> str: |
| 69 | + h = src |
| 70 | + |
| 71 | + # Replace < and >. |
| 72 | + h = re.sub(r"<", "<", h) |
| 73 | + h = re.sub(r">", ">", h) |
| 74 | + |
| 75 | + # Title |
| 76 | + h = re.sub(r"^## (Mypy [0-9.]+)", r"<h1>\1 Released</h1>", h, flags=re.MULTILINE) |
| 77 | + |
| 78 | + # Subheadings |
| 79 | + h = re.sub(r"\n#### ([A-Z`].*)\n", r"\n<h2>\1</h2>\n", h) |
| 80 | + |
| 81 | + # Sub-subheadings |
| 82 | + h = re.sub(r"\n\*\*([A-Z_`].*)\*\*\n", r"\n<h3>\1</h3>\n", h) |
| 83 | + h = re.sub(r"\n`\*\*([A-Z_`].*)\*\*\n", r"\n<h3>`\1</h3>\n", h) |
| 84 | + |
| 85 | + # Translate `**` |
| 86 | + h = re.sub(r"`\*\*`", "<tt>**</tt>", h) |
| 87 | + |
| 88 | + # Paragraphs |
| 89 | + h = re.sub(r"\n([A-Z])", r"\n<p>\1", h) |
| 90 | + |
| 91 | + # Bullet lists |
| 92 | + h = format_lists(h) |
| 93 | + |
| 94 | + # Code blocks |
| 95 | + h = format_code(h) |
| 96 | + |
| 97 | + # Code fragments |
| 98 | + h = re.sub(r"`([^`]+)`", r"<tt>\1</tt>", h) |
| 99 | + |
| 100 | + # Remove **** noise |
| 101 | + h = re.sub(r"\*\*\*\*", "", h) |
| 102 | + |
| 103 | + # Bold text |
| 104 | + h = re.sub(r"\*\*([A-Za-z].*?)\*\*", r" <b>\1</b>", h) |
| 105 | + |
| 106 | + # Emphasized text |
| 107 | + h = re.sub(r" \*([A-Za-z].*?)\*", r" <i>\1</i>", h) |
| 108 | + |
| 109 | + # Remove redundant PR links to avoid double links (they will be generated below) |
| 110 | + h = re.sub(r"\[(#[0-9]+)\]\(https://github.com/python/mypy/pull/[0-9]+/?\)", r"\1", h) |
| 111 | + |
| 112 | + # Issue and PR links |
| 113 | + h = re.sub(r"\((#[0-9]+)\) +\(([^)]+)\)", r"(\2, \1)", h) |
| 114 | + h = re.sub( |
| 115 | + r"fixes #([0-9]+)", |
| 116 | + r'fixes issue <a href="https://github.com/python/mypy/issues/\1">\1</a>', |
| 117 | + h, |
| 118 | + ) |
| 119 | + h = re.sub(r"#([0-9]+)", r'PR <a href="https://github.com/python/mypy/pull/\1">\1</a>', h) |
| 120 | + h = re.sub(r"\) \(PR", ", PR", h) |
| 121 | + |
| 122 | + # Markdown links |
| 123 | + h = re.sub(r"\[([^]]*)\]\(([^)]*)\)", r'<a href="\2">\1</a>', h) |
| 124 | + |
| 125 | + # Add random links in case they are missing |
| 126 | + h = re.sub( |
| 127 | + r"contributors to typeshed:", |
| 128 | + 'contributors to <a href="https://github.com/python/typeshed">typeshed</a>:', |
| 129 | + h, |
| 130 | + ) |
| 131 | + |
| 132 | + # Add missing top-level HTML tags |
| 133 | + h = '<html>\n<meta charset="utf-8" />\n<body>\n' + h + "</body>\n</html>" |
| 134 | + |
| 135 | + return h |
| 136 | + |
| 137 | + |
| 138 | +def extract_version(src: str, version: str) -> str: |
| 139 | + a = src.splitlines() |
| 140 | + i = 0 |
| 141 | + heading = f"## Mypy {version}" |
| 142 | + while i < len(a): |
| 143 | + if a[i].strip() == heading: |
| 144 | + break |
| 145 | + i += 1 |
| 146 | + else: |
| 147 | + raise RuntimeError(f"Can't find heading {heading!r}") |
| 148 | + j = i + 1 |
| 149 | + while not a[j].startswith("## "): |
| 150 | + j += 1 |
| 151 | + return "\n".join(a[i:j]) |
| 152 | + |
| 153 | + |
| 154 | +def main() -> None: |
| 155 | + parser = argparse.ArgumentParser( |
| 156 | + description="Generate HTML release blog post based on CHANGELOG.md and write to stdout." |
| 157 | + ) |
| 158 | + parser.add_argument("version", help="mypy version, in form X.Y or X.Y.Z") |
| 159 | + args = parser.parse_args() |
| 160 | + version: str = args.version |
| 161 | + if not re.match(r"[0-9]+(\.[0-9]+)+$", version): |
| 162 | + sys.exit(f"error: Version must be of form X.Y or X.Y.Z, not {version!r}") |
| 163 | + changelog_path = os.path.join(os.path.dirname(__file__), os.path.pardir, "CHANGELOG.md") |
| 164 | + src = open(changelog_path).read() |
| 165 | + src = extract_version(src, version) |
| 166 | + dst = convert(src) |
| 167 | + sys.stdout.write(dst) |
| 168 | + |
| 169 | + |
| 170 | +if __name__ == "__main__": |
| 171 | + main() |
0 commit comments