python (3.11.7)
1 import re
2 import sys
3 from contextlib import suppress
4 from typing import Iterable, NamedTuple, Optional
5
6 from .color import Color
7 from .style import Style
8 from .text import Text
9
10 re_ansi = re.compile(
11 r"""
12 (?:\x1b\](.*?)\x1b\\)|
13 (?:\x1b([(@-Z\\-_]|\[[0-?]*[ -/]*[@-~]))
14 """,
15 re.VERBOSE,
16 )
17
18
19 class ESC[4;38;5;81m_AnsiToken(ESC[4;38;5;149mNamedTuple):
20 """Result of ansi tokenized string."""
21
22 plain: str = ""
23 sgr: Optional[str] = ""
24 osc: Optional[str] = ""
25
26
27 def _ansi_tokenize(ansi_text: str) -> Iterable[_AnsiToken]:
28 """Tokenize a string in to plain text and ANSI codes.
29
30 Args:
31 ansi_text (str): A String containing ANSI codes.
32
33 Yields:
34 AnsiToken: A named tuple of (plain, sgr, osc)
35 """
36
37 position = 0
38 sgr: Optional[str]
39 osc: Optional[str]
40 for match in re_ansi.finditer(ansi_text):
41 start, end = match.span(0)
42 osc, sgr = match.groups()
43 if start > position:
44 yield _AnsiToken(ansi_text[position:start])
45 if sgr:
46 if sgr == "(":
47 position = end + 1
48 continue
49 if sgr.endswith("m"):
50 yield _AnsiToken("", sgr[1:-1], osc)
51 else:
52 yield _AnsiToken("", sgr, osc)
53 position = end
54 if position < len(ansi_text):
55 yield _AnsiToken(ansi_text[position:])
56
57
58 SGR_STYLE_MAP = {
59 1: "bold",
60 2: "dim",
61 3: "italic",
62 4: "underline",
63 5: "blink",
64 6: "blink2",
65 7: "reverse",
66 8: "conceal",
67 9: "strike",
68 21: "underline2",
69 22: "not dim not bold",
70 23: "not italic",
71 24: "not underline",
72 25: "not blink",
73 26: "not blink2",
74 27: "not reverse",
75 28: "not conceal",
76 29: "not strike",
77 30: "color(0)",
78 31: "color(1)",
79 32: "color(2)",
80 33: "color(3)",
81 34: "color(4)",
82 35: "color(5)",
83 36: "color(6)",
84 37: "color(7)",
85 39: "default",
86 40: "on color(0)",
87 41: "on color(1)",
88 42: "on color(2)",
89 43: "on color(3)",
90 44: "on color(4)",
91 45: "on color(5)",
92 46: "on color(6)",
93 47: "on color(7)",
94 49: "on default",
95 51: "frame",
96 52: "encircle",
97 53: "overline",
98 54: "not frame not encircle",
99 55: "not overline",
100 90: "color(8)",
101 91: "color(9)",
102 92: "color(10)",
103 93: "color(11)",
104 94: "color(12)",
105 95: "color(13)",
106 96: "color(14)",
107 97: "color(15)",
108 100: "on color(8)",
109 101: "on color(9)",
110 102: "on color(10)",
111 103: "on color(11)",
112 104: "on color(12)",
113 105: "on color(13)",
114 106: "on color(14)",
115 107: "on color(15)",
116 }
117
118
119 class ESC[4;38;5;81mAnsiDecoder:
120 """Translate ANSI code in to styled Text."""
121
122 def __init__(self) -> None:
123 self.style = Style.null()
124
125 def decode(self, terminal_text: str) -> Iterable[Text]:
126 """Decode ANSI codes in an iterable of lines.
127
128 Args:
129 lines (Iterable[str]): An iterable of lines of terminal output.
130
131 Yields:
132 Text: Marked up Text.
133 """
134 for line in terminal_text.splitlines():
135 yield self.decode_line(line)
136
137 def decode_line(self, line: str) -> Text:
138 """Decode a line containing ansi codes.
139
140 Args:
141 line (str): A line of terminal output.
142
143 Returns:
144 Text: A Text instance marked up according to ansi codes.
145 """
146 from_ansi = Color.from_ansi
147 from_rgb = Color.from_rgb
148 _Style = Style
149 text = Text()
150 append = text.append
151 line = line.rsplit("\r", 1)[-1]
152 for plain_text, sgr, osc in _ansi_tokenize(line):
153 if plain_text:
154 append(plain_text, self.style or None)
155 elif osc is not None:
156 if osc.startswith("8;"):
157 _params, semicolon, link = osc[2:].partition(";")
158 if semicolon:
159 self.style = self.style.update_link(link or None)
160 elif sgr is not None:
161 # Translate in to semi-colon separated codes
162 # Ignore invalid codes, because we want to be lenient
163 codes = [
164 min(255, int(_code) if _code else 0)
165 for _code in sgr.split(";")
166 if _code.isdigit() or _code == ""
167 ]
168 iter_codes = iter(codes)
169 for code in iter_codes:
170 if code == 0:
171 # reset
172 self.style = _Style.null()
173 elif code in SGR_STYLE_MAP:
174 # styles
175 self.style += _Style.parse(SGR_STYLE_MAP[code])
176 elif code == 38:
177 # Foreground
178 with suppress(StopIteration):
179 color_type = next(iter_codes)
180 if color_type == 5:
181 self.style += _Style.from_color(
182 from_ansi(next(iter_codes))
183 )
184 elif color_type == 2:
185 self.style += _Style.from_color(
186 from_rgb(
187 next(iter_codes),
188 next(iter_codes),
189 next(iter_codes),
190 )
191 )
192 elif code == 48:
193 # Background
194 with suppress(StopIteration):
195 color_type = next(iter_codes)
196 if color_type == 5:
197 self.style += _Style.from_color(
198 None, from_ansi(next(iter_codes))
199 )
200 elif color_type == 2:
201 self.style += _Style.from_color(
202 None,
203 from_rgb(
204 next(iter_codes),
205 next(iter_codes),
206 next(iter_codes),
207 ),
208 )
209
210 return text
211
212
213 if sys.platform != "win32" and __name__ == "__main__": # pragma: no cover
214 import io
215 import os
216 import pty
217 import sys
218
219 decoder = AnsiDecoder()
220
221 stdout = io.BytesIO()
222
223 def read(fd: int) -> bytes:
224 data = os.read(fd, 1024)
225 stdout.write(data)
226 return data
227
228 pty.spawn(sys.argv[1:], read)
229
230 from .console import Console
231
232 console = Console(record=True)
233
234 stdout_result = stdout.getvalue().decode("utf-8")
235 print(stdout_result)
236
237 for line in decoder.decode(stdout_result):
238 console.print(line)
239
240 console.save_html("stdout.html")
Browse
Usage
Versions