1# HTML block
2from __future__ import annotations
3
4import logging
5import re
6
7from ..common.html_blocks import block_names
8from ..common.html_re import HTML_OPEN_CLOSE_TAG_STR
9from .state_block import StateBlock
10
11LOGGER = logging.getLogger(__name__)
12
13# An array of opening and corresponding closing sequences for html tags,
14# last argument defines whether it can terminate a paragraph or not
15HTML_SEQUENCES: list[tuple[re.Pattern[str], re.Pattern[str], bool]] = [
16 (
17 re.compile(r"^<(?a:(script|pre|style|textarea))(?=(\s|>|$))", re.IGNORECASE),
18 re.compile(r"<\/(?a:(script|pre|style|textarea))>", re.IGNORECASE),
19 True,
20 ),
21 (re.compile(r"^<!--"), re.compile(r"-->"), True),
22 (re.compile(r"^<\?"), re.compile(r"\?>"), True),
23 (re.compile(r"^<![A-Za-z]"), re.compile(r">"), True),
24 (re.compile(r"^<!\[CDATA\["), re.compile(r"\]\]>"), True),
25 (
26 re.compile(
27 "^</?(?a:(" + "|".join(block_names) + "))(?=(\\s|/?>|$))", re.IGNORECASE
28 ),
29 re.compile(r"^$"),
30 True,
31 ),
32 (re.compile(HTML_OPEN_CLOSE_TAG_STR + "\\s*$"), re.compile(r"^$"), False),
33]
34
35
36def html_block(state: StateBlock, startLine: int, endLine: int, silent: bool) -> bool:
37 LOGGER.debug(
38 "entering html_block: %s, %s, %s, %s", state, startLine, endLine, silent
39 )
40 pos = state.bMarks[startLine] + state.tShift[startLine]
41 maximum = state.eMarks[startLine]
42
43 if state.is_code_block(startLine):
44 return False
45
46 if not state.md.options.get("html", None):
47 return False
48
49 try:
50 if state.src[pos] != "<":
51 return False
52 except IndexError:
53 return False
54
55 lineText = state.src[pos:maximum]
56
57 html_seq = None
58 for HTML_SEQUENCE in HTML_SEQUENCES:
59 if HTML_SEQUENCE[0].search(lineText):
60 html_seq = HTML_SEQUENCE
61 break
62
63 if not html_seq:
64 return False
65
66 if silent:
67 # true if this sequence can be a terminator, false otherwise
68 return html_seq[2]
69
70 nextLine = startLine + 1
71
72 # Block types 6 and 7 (the only ones whose end condition is a blank line)
73 # have `^$` as their closing regexp. For all other types (1-5, e.g.
74 # `<!--` comments), a blank line is regular content and must not terminate
75 # the block - it ends only when its closing sequence is found.
76 endsOnBlankLine = bool(html_seq[1].search(""))
77
78 # If we are here - we detected HTML block.
79 # Let's roll down till block end.
80 if not html_seq[1].search(lineText):
81 while nextLine < endLine:
82 # An outdented blank line shouldn't end a block that doesn't end on a
83 # blank line (e.g. a `<!--` comment inside a list item). Such blocks
84 # must continue until their closing sequence regardless of indent.
85 if state.sCount[nextLine] < state.blkIndent and (
86 endsOnBlankLine or not state.isEmpty(nextLine)
87 ):
88 break
89
90 pos = state.bMarks[nextLine] + state.tShift[nextLine]
91 maximum = state.eMarks[nextLine]
92 lineText = state.src[pos:maximum]
93
94 if html_seq[1].search(lineText):
95 if len(lineText) != 0:
96 nextLine += 1
97 break
98 nextLine += 1
99
100 state.line = nextLine
101
102 token = state.push("html_block", "", 0)
103 token.map = [startLine, nextLine]
104 token.content = state.getLines(startLine, nextLine, state.blkIndent, True)
105
106 return True