fix: add html2text + Rich render_display_data() for display_data output (#58)
Previously text/html display_data output was either ignored (log) or printed raw. Now: - New dependency: html2text (>=2024.2.26) - New util: render_display_data() in utils.py — returns Rich renderables (Markdown or Text) using priority text/markdown > text/html > text/plain. text/html is converted via html2text; text/plain is wrapped with Text.from_ansi to handle embedded ANSI escapes. - All three call sites (exec.py, automation.py, repl.py) refactored to use the shared function instead of duplicated if/elif chains. - Each module uses a shared Console instance (_console / self.console) to avoid re-allocating per-output during streaming. - Removed direct html2text and Markdown imports from the three call sites — they now just pass the renderable to Console.print(). - Tests in test_utils.py cover Markdown return, Text return, priority, and no-text fallback.
Merlin Ran committed
Jun 16, 2026 at 14:25 UTC
c58a851bba025e9cdbf9e66613efa401020853db
7 files changed
+89
-18
pyproject.toml
+1
@@ -27,6 +27,7 @@ dependencies = [
27
"filelock>=3.29.2",
28
"google-auth>=2.49.1",
29
"google-auth-oauthlib>=1.3.0",
30
+ "html2text>=2024.2.26",
31
"jupyter-kernel-client",
32
"nbformat>=5.10.4",
33
"packaging>=24.0",
src/colab_cli/commands/automation.py
+8
-3
@@ -18,12 +18,15 @@ import sys
18
import json
19
from typing import Optional, List
20
import typer
21
+from rich.console import Console
22
from typing_extensions import Annotated
23
24
from colab_cli.runtime import ColabRuntime
25
from colab_cli.contents import ContentsClient
26
from colab_cli.auth import get_credentials
26
-from colab_cli.utils import get_status_code
27
+from colab_cli.utils import get_status_code, render_display_data
28
+
29
+_console = Console()
30
31
32
# Default execute() timeout for human-in-the-loop automations (auth /
@@ -35,6 +38,7 @@ from colab_cli.utils import get_status_code
38
INTERACTIVE_AUTOMATION_TIMEOUT_SEC = 600
39
40
41
+
42
def run_automation(
43
name: str,
44
op: str,
@@ -153,8 +157,9 @@ def run_automation(
157
if "text" in out:
158
sys.stdout.write(out["text"])
159
elif "data" in out:
156
- if "text/plain" in out["data"]:
157
- typer.echo(out["data"]["text/plain"])
160
+ text = render_display_data(out["data"])
161
+ if text is not None:
162
+ _console.print(text)
163
elif out.get("output_type") == "error":
164
ename = out.get("ename", "Error")
165
evalue = out.get("evalue", "")
src/colab_cli/commands/execution.py
+8
-3
@@ -20,13 +20,16 @@ import sys
20
import typer
21
import uuid
22
from nbformat.v4 import new_output
23
+from rich.console import Console
24
from typing import Optional
25
from typing_extensions import Annotated
26
27
from colab_cli.runtime import ColabRuntime
27
-from colab_cli.utils import handle_image, is_terminal_error
28
+from colab_cli.utils import handle_image, is_terminal_error, render_display_data
29
from colab_cli.console import connect_console
30
31
+_console = Console()
32
+
33
TITLE_REGEX = re.compile(r"^\s*#\s*@title\s+(.*)", re.MULTILINE)
34
35
@@ -72,6 +75,7 @@ def save_output(outputs, cell):
75
)
76
77
78
+
79
def display_output(out, output_image=None):
80
if out.get("output_type") == "stream":
81
stream = sys.stderr if out.get("name") == "stderr" else sys.stdout
@@ -79,8 +83,9 @@ def display_output(out, output_image=None):
83
stream.flush()
84
elif "data" in out:
85
data = out["data"]
82
- if text := data.get("text/plain"):
83
- typer.echo(text)
86
+ text = render_display_data(data)
87
+ if text is not None:
88
+ _console.print(text)
89
if png := data.get("image/png"):
90
handle_image(png, "image/png", target_path=output_image)
91
elif jpeg := data.get("image/jpeg"):
src/colab_cli/repl.py
+11
-10
@@ -25,9 +25,8 @@ from rich.console import Console
25
from rich.text import Text
26
27
from colab_cli.runtime import ColabRuntime
28
-from colab_cli.utils import handle_image
28
+from colab_cli.utils import handle_image, render_display_data
29
30
-console = Console()
30
31
32
class ColabREPL:
@@ -43,7 +42,7 @@ class ColabREPL:
42
self.history_logger = history_logger
43
self.output_image = output_image
44
self.kb = KeyBindings()
46
- self.console = console
45
+ self.console = Console()
46
self.repl_history: List[dict] = []
47
48
@self.kb.add("enter")
@@ -91,14 +90,16 @@ class ColabREPL:
90
image_displayed = True
91
break
92
94
- if "text/plain" in data:
95
- text = data["text/plain"]
93
+ text = render_display_data(data)
94
+ if text is not None:
95
# Skip generic IPython object reprs if we already showed an image
97
- if image_displayed and any(
98
- x in text for x in ["<IPython.core.display.Image", "<Figure size"]
99
- ):
100
- return
101
- self.console.print(Text.from_ansi(text))
96
+ if isinstance(text, Text) and image_displayed:
97
+ if any(
98
+ x in text.plain
99
+ for x in ["<IPython.core.display.Image", "<Figure size"]
100
+ ):
101
+ return
102
+ self.console.print(text)
103
elif output.get("output_type") == "error":
104
ename = output.get("ename", "Error")
105
evalue = output.get("evalue", "")
src/colab_cli/utils.py
+20
-1
@@ -13,12 +13,15 @@
13
# limitations under the License.
14
15
import base64
16
+import html2text
17
import logging
18
import sys
19
import tempfile
20
21
+from typing import Optional, Union
22
21
-from typing import Optional
23
+from rich.markdown import Markdown
24
+from rich.text import Text
25
26
27
def get_status_code(e: Exception) -> Optional[int]:
@@ -66,6 +69,22 @@ def print_kitty(image_bytes: bytes):
69
logging.exception("Kitty rendering failed")
70
71
72
+def render_display_data(data: dict) -> Union[Markdown, Text, None]:
73
+ """Extract the best text representation from a display_data dict.
74
+
75
+ Priority: text/markdown > text/html (via html2text) > text/plain.
76
+ Returns a Rich renderable (Markdown or Text) or None when no text mime
77
+ type is present. Callers can pass the result directly to Console.print().
78
+ """
79
+ if "text/markdown" in data:
80
+ return Markdown(data["text/markdown"])
81
+ if "text/html" in data:
82
+ return Markdown(html2text.html2text(data["text/html"]))
83
+ if "text/plain" in data:
84
+ return Text.from_ansi(data["text/plain"])
85
+ return None
86
+
87
+
88
def handle_image(image_b64: str, mime_type: str = "image/png", target_path: str = None):
89
image_bytes = base64.b64decode(image_b64)
90
# Print inline using Kitty protocol
tests/test_utils.py
+30
-1
@@ -15,7 +15,11 @@
15
import base64
16
from unittest.mock import MagicMock, patch
17
18
-from colab_cli.utils import handle_image, print_kitty
18
+import pytest
19
+from rich.markdown import Markdown
20
+from rich.text import Text
21
+
22
+from colab_cli.utils import handle_image, print_kitty, render_display_data
23
24
25
@patch("colab_cli.utils.sys.stdout.isatty", return_value=True)
@@ -56,3 +60,28 @@ def test_handle_image(mock_print_kitty, mock_tempfile, capsys):
60
61
captured = capsys.readouterr()
62
assert "/tmp/fake.png" in captured.out
63
+
64
+
65
+@pytest.mark.parametrize(
66
+ "data, expected_markup",
67
+ [
68
+ ({"text/markdown": "**md**"}, "**md**"),
69
+ ({"text/html": "<b>hi</b>"}, "**hi**\n\n"),
70
+ ({"text/markdown": "**md**", "text/html": "<b>hi</b>"}, "**md**"),
71
+ ({"text/html": "<b>hi</b>", "text/plain": "plain"}, "**hi**\n\n"),
72
+ ],
73
+)
74
+def test_render_display_data_markdown(data, expected_markup):
75
+ result = render_display_data(data)
76
+ assert isinstance(result, Markdown)
77
+ assert result.markup == expected_markup
78
+
79
+
80
+def test_render_display_data_plain():
81
+ result = render_display_data({"text/plain": "plain"})
82
+ assert isinstance(result, Text)
83
+ assert result.plain == "plain"
84
+
85
+
86
+def test_render_display_data_none():
87
+ assert render_display_data({"image/png": "..."}) is None
uv.lock
+11
@@ -319,6 +319,7 @@ dependencies = [
319
{ name = "filelock" },
320
{ name = "google-auth" },
321
{ name = "google-auth-oauthlib" },
322
+ { name = "html2text" },
323
{ name = "jupyter-kernel-client" },
324
{ name = "nbformat" },
325
{ name = "packaging" },
@@ -346,6 +347,7 @@ requires-dist = [
347
{ name = "filelock", specifier = ">=3.29.2" },
348
{ name = "google-auth", specifier = ">=2.49.1" },
349
{ name = "google-auth-oauthlib", specifier = ">=1.3.0" },
350
+ { name = "html2text", specifier = ">=2024.2.26" },
351
{ name = "jupyter-kernel-client", git = "https://github.com/googlecolab/jupyter-kernel-client.git" },
352
{ name = "nbformat", specifier = ">=5.10.4" },
353
{ name = "packaging", specifier = ">=24.0" },
@@ -367,6 +369,15 @@ dev = [
369
{ name = "ruff", specifier = ">=0.15.6" },
370
]
371
372
+[[package]]
373
+name = "html2text"
374
+version = "2025.4.15"
375
+source = { registry = "https://pypi.org/simple" }
376
+sdist = { url = "https://files.pythonhosted.org/packages/f8/27/e158d86ba1e82967cc2f790b0cb02030d4a8bef58e0c79a8590e9678107f/html2text-2025.4.15.tar.gz", hash = "sha256:948a645f8f0bc3abe7fd587019a2197a12436cd73d0d4908af95bfc8da337588", size = 64316, upload-time = "2025-04-15T04:02:30.045Z" }
377
+wheels = [
378
+ { url = "https://files.pythonhosted.org/packages/1d/84/1a0f9555fd5f2b1c924ff932d99b40a0f8a6b12f6dd625e2a47f415b00ea/html2text-2025.4.15-py3-none-any.whl", hash = "sha256:00569167ffdab3d7767a4cdf589b7f57e777a5ed28d12907d8c58769ec734acc", size = 34656, upload-time = "2025-04-15T04:02:28.44Z" },
379
+]
380
+
381
[[package]]
382
name = "idna"
383
version = "3.11"