466460112b
The anchor regex only matched quoted hrefs; unquoted (href=URL) dropped the URL from the plaintext part. Now handles double/single/unquoted. Added scripts/test_email_plaintext.py (11 cases: link forms, mailto, template-tag preservation, tag stripping, entity unescape, blank-line collapse).
110 lines
4 KiB
Python
110 lines
4 KiB
Python
"""Shared HTML -> plaintext conversion for outbound campaigns.
|
|
|
|
Every campaign we build was HTML-only (no plaintext MIME part). A missing
|
|
text/plain alternative is a spam-score signal: legitimate bulk senders ship
|
|
multipart/alternative, and several filters (and most "this looks like spam"
|
|
heuristics) penalise HTML-only mail. It also degrades the experience for
|
|
plaintext-only clients and accessibility tooling.
|
|
|
|
Listmonk only emits multipart/alternative when a campaign's `altbody` is set;
|
|
otherwise it sends text/html alone. So we generate a readable plaintext
|
|
rendition from the HTML body and pass it as `altbody`.
|
|
|
|
This is intentionally dependency-free (no bs4/html2text on the prod box): a
|
|
small, well-tested regex pipeline that:
|
|
- drops <script>/<style>/<head> blocks,
|
|
- turns <a href=...>text</a> into "text (url)" so links survive,
|
|
- maps <br>, </p>, </div>, <li>, headings, <tr> to newlines,
|
|
- prefixes <li> with "- ",
|
|
- strips all remaining tags,
|
|
- unescapes HTML entities,
|
|
- collapses runs of blank lines / trailing whitespace.
|
|
|
|
Listmonk template tags ({{ .Subscriber... }}, {{ UnsubscribeURL }}) are left
|
|
untouched so they still render per-subscriber in the plaintext part too.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import html as _html
|
|
import re
|
|
|
|
__all__ = ["html_to_text"]
|
|
|
|
_RE_FLAGS = re.IGNORECASE | re.DOTALL
|
|
|
|
# Whole blocks whose *content* must be discarded, not just the tags.
|
|
_DROP_BLOCKS = re.compile(
|
|
r"<(script|style|head|title|noscript)\b[^>]*>.*?</\1>", _RE_FLAGS
|
|
)
|
|
# HTML comments (Listmonk/MSO conditional comments etc.).
|
|
_COMMENTS = re.compile(r"<!--.*?-->", _RE_FLAGS)
|
|
# <a href="URL" ...>TEXT</a> -> TEXT (URL) (skip mailto:/tel:/anchors)
|
|
# Handles quoted ("..." / '...') and unquoted (href=URL) hrefs.
|
|
_ANCHORS = re.compile(
|
|
r'<a\b[^>]*?\bhref\s*=\s*'
|
|
r'(?:"([^"]*)"|\'([^\']*)\'|([^\s">]+))'
|
|
r'[^>]*>(.*?)</a>',
|
|
_RE_FLAGS,
|
|
)
|
|
# Tags that should become a line break.
|
|
_BR = re.compile(r"<br\s*/?>", re.IGNORECASE)
|
|
_BLOCK_END = re.compile(
|
|
r"</(p|div|h[1-6]|tr|table|ul|ol|blockquote|section|header|footer)>",
|
|
re.IGNORECASE,
|
|
)
|
|
_LI = re.compile(r"<li\b[^>]*>", re.IGNORECASE)
|
|
_HR = re.compile(r"<hr\s*/?>", re.IGNORECASE)
|
|
_ANY_TAG = re.compile(r"<[^>]+>")
|
|
_MANY_BLANKS = re.compile(r"\n[ \t]*\n[ \t]*(\n[ \t]*)+")
|
|
_TRAIL_WS = re.compile(r"[ \t]+\n")
|
|
_MANY_SPACES = re.compile(r"[ \t]{2,}")
|
|
|
|
|
|
def _anchor_repl(m: "re.Match[str]") -> str:
|
|
# href is whichever of the 3 alternatives matched (double/single/unquoted).
|
|
url = (m.group(1) or m.group(2) or m.group(3) or "").strip()
|
|
text = _ANY_TAG.sub("", m.group(4)).strip()
|
|
low = url.lower()
|
|
# mailto:/tel: -> surface the address (with link text if it adds info).
|
|
# Bare in-page anchors -> keep text only. Templated hrefs (e.g.
|
|
# {{ UnsubscribeURL }}) ARE kept as "text (url)" so the per-subscriber link
|
|
# still renders in the plaintext part.
|
|
if low.startswith(("mailto:", "tel:")):
|
|
addr = url.split(":", 1)[1].split("?", 1)[0]
|
|
if text and text != addr:
|
|
return f"{text} ({addr})"
|
|
return addr
|
|
if low.startswith("#"):
|
|
return text
|
|
if not text:
|
|
return url
|
|
if text == url:
|
|
return url
|
|
return f"{text} ({url})"
|
|
|
|
|
|
def html_to_text(html: str) -> str:
|
|
"""Convert an HTML email body to a readable text/plain rendition.
|
|
|
|
Returns "" for empty input. Listmonk template tags are preserved verbatim.
|
|
"""
|
|
if not html:
|
|
return ""
|
|
s = html
|
|
s = _DROP_BLOCKS.sub("", s)
|
|
s = _COMMENTS.sub("", s)
|
|
s = _ANCHORS.sub(_anchor_repl, s)
|
|
s = _HR.sub("\n----------\n", s)
|
|
s = _BR.sub("\n", s)
|
|
s = _LI.sub("\n- ", s)
|
|
s = _BLOCK_END.sub("\n", s)
|
|
s = _ANY_TAG.sub("", s)
|
|
s = _html.unescape(s)
|
|
# Normalise whitespace: trim trailing spaces, collapse runs of spaces and
|
|
# blank lines, strip leading/trailing blank space overall.
|
|
s = _TRAIL_WS.sub("\n", s)
|
|
s = _MANY_SPACES.sub(" ", s)
|
|
s = _MANY_BLANKS.sub("\n\n", s)
|
|
s = "\n".join(line.rstrip() for line in s.splitlines())
|
|
return s.strip() + "\n"
|