Add Markdown & HTML to LaTeX conversion

great_tables/_formats.py

@@ -17,15 +17,16 @@
     TypeVar,
     Union,
     cast,
-    overload,
+    overload
 )
 
 import babel
 import faicons
 from babel.dates import format_date, format_datetime, format_time
 from typing_extensions import TypeAlias
 
-from ._gt_data import FormatFn, FormatFns, FormatInfo, FormatterSkipElement, GTData, PFrameData
+from ._gt_data import (FormatFn, FormatFns, FormatInfo, FormatterSkipElement,
+                       GTData, PFrameData)
 from ._helpers import px
 from ._locale import (
     _get_currencies_data,
@@ -44,7 +45,7 @@
     is_series,
     to_list,
 )
-from ._text import _md_html, escape_pattern_str_latex
+from ._text import _md_html, _md_latex, escape_pattern_str_latex
 from ._utils import _str_detect, _str_replace, is_valid_http_schema
 from ._utils_nanoplots import _generate_nanoplot
 
@@ -2668,7 +2669,7 @@ def fmt_markdown_context(
     context: str,
 ) -> str:
     if context == "latex":
-        raise NotImplementedError("fmt_markdown() is not supported in LaTeX.")
+        return _md_latex(x)
 
     if is_na(data._tbl_data, x):
         return x

great_tables/_text.py

@@ -6,6 +6,9 @@
 from typing import Callable
 
 import commonmark
+from docutils.core import publish_parts
+from markdownify import markdownify as mdify
+from myst_parser.docutils_ import Parser
 
 
 class BaseText:
@@ -48,13 +51,7 @@ def to_html(self) -> str:
         return self.text
 
     def to_latex(self) -> str:
-        from ._utils_render_latex import _not_implemented
-
-        _not_implemented(
-            "Using the `html()` helper function won't convert HTML to LaTeX. Escaping HTML string instead."
-        )
-
-        return _latex_escape(self.text)
+        return _html_latex(self.text)
 
 
 def _md_html(x: str) -> str:
@@ -63,10 +60,34 @@ def _md_html(x: str) -> str:
 
 
 def _md_latex(x: str) -> str:
-    # TODO: Implement commonmark to LaTeX conversion (through a different library as
-    # commonmark-py does not support it)
-    raise NotImplementedError("Markdown to LaTeX conversion is not supported yet")
-
+    # USE REGEX TO CONVERT <sub> AND <sup> TO MYST MARKDOWN
+    input = re.sub(r'<sub>(.*?)</sub>', r'{sub}`\1`', x)
+    input = re.sub(r'<sup>(.*?)</sup>', r'{sup}`\1`', input)
+
+    # Use Myst-Parser to convert Markdown to LaTeX
+    raw_output = publish_parts(
+                    source=input,
+                    writer_name="latex",
+                    settings_overrides={
+                        "myst_enable_extensions": ['strikethrough'],
+                        "embed_stylesheet": False,
+                        "legacy_column_widths": True,
+                        "use_latex_citations": False,
+                    },
+                    parser=Parser(),
+                )
+
+    output = raw_output['body'].strip()
+
+    return output
+
+
+def _html_latex(x: str) -> str:
+    # Turn HTML to Markdown first
+    input = mdify(x, strip=['br'], sub_symbol="<sub>", sup_symbol="<sup>")
+
+    # Then render Markdown to LaTeX
+    return _md_latex(input)
 
 def _process_text(x: str | BaseText | None, context: str = "html") -> str:
     if x is None:

pyproject.toml

@@ -44,7 +44,9 @@ dependencies = [
     "importlib-metadata",
     "typing_extensions>=3.10.0.0",
     "Babel>=2.13.1",
-    "importlib-resources"
+    "importlib-resources",
+    "markdownify>=1.2.2",
+    "myst_parser>=4.0.1"
 ]
 requires-python = ">=3.9"
 

tests/test_text.py

@@ -8,6 +8,7 @@
     _latex_escape,
     escape_pattern_str_latex,
     _process_text,
+    _md_latex,
 )
 
 
@@ -30,7 +31,7 @@ def test_md_class():
 
 def test_html_class():
     assert Html("<strong>text</strong>").to_html() == "<strong>text</strong>"
-    assert Html("<strong>text</strong>").to_latex() == "<strong>text</strong>"
+    assert Html("<strong>text</strong>").to_latex() == "\\textbf{text}"
 
 
 def test_latex_escape():
@@ -56,17 +57,17 @@ def test_process_text_html():
 def test_process_text_latex():
     assert _process_text("a & _b_", context="latex") == "a \\& \\_b\\_"
     assert _process_text(Text("\\_\\$"), context="latex") == "\\_\\$"
-    assert _process_text(Html("**a** & <b>"), context="latex") == "**a** \\& <b>"
+    assert _process_text(Html("**a** & <b>bold</b>"), context="latex") == "**a** \\& \\textbf{bold}"
+    assert _process_text(Md("**a**"), context="latex") == "\\textbf{a}"
     assert _process_text(None, context="latex") == ""
 
-    with pytest.raises(NotImplementedError) as exc_info:
-        _process_text(Md("**a** & <b>"), context="latex")
-
-    assert "Markdown to LaTeX conversion is not supported yet" in exc_info.value.args[0]
-
 
 def test_process_text_raises():
     with pytest.raises(TypeError) as exc_info:
         _process_text(1, context="html")  # type: ignore
 
     assert "Invalid type: <class 'int'>" in exc_info.value.args[0]
+
+
+def test_md_latex():
+    assert _md_latex("Testing **bold** text") == "Testing \\textbf{bold} text"