Skip to content

ENH: add LaTeX math mode with parentheses #51903

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 14 commits into from
Mar 31, 2023
Merged

ENH: add LaTeX math mode with parentheses #51903

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
ba72f26
ENH: add math mode with parentheses
natmokval Mar 11, 2023
0801f6a
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 11, 2023
2320fb3
ENH: add math mode with parentheses II
natmokval Mar 14, 2023
c20e795
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 15, 2023
2e283a7
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 15, 2023
29e357c
ENH: add math mode with parentheses III
natmokval Mar 19, 2023
849bf36
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 19, 2023
9be5c22
ENH: add an example to latex-math mode and a line to whatsnew
natmokval Mar 20, 2023
7b4b53a
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 20, 2023
25e6b1f
ENH: update docs for Styler: add description latex-math mode to escape
natmokval Mar 21, 2023
ed27582
improve code style
natmokval Mar 21, 2023
3daf4d5
add an example to test and correct _escape_latex_math
natmokval Mar 24, 2023
78dcdf8
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 24, 2023
c06695e
Merge branch 'main' into 50040-add-math-mode-formatter-escape=latex-p…
natmokval Mar 31, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
89 changes: 72 additions & 17 deletions pandas/io/formats/style_render.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1117,7 +1117,8 @@ def format(
2 & \textbf{\$\%\#} \\
\end{tabular}

Using ``escape`` in 'latex-math' mode.
Applying ``escape`` in 'latex-math' mode. In the example below
we enter math mode using the character ``$``.

>>> df = pd.DataFrame([[r"$\sum_{i=1}^{10} a_i$ a~b $\alpha \
... = \frac{\beta}{\zeta^2}$"], ["%#^ $ \$x^2 $"]])
Expand All @@ -1129,6 +1130,20 @@ def format(
1 & \%\#\textasciicircum \space $ \$x^2 $ \\
\end{tabular}

We can use the character ``\(`` to enter math mode and the character ``\)``
to close math mode.

>>> df = pd.DataFrame([[r"\(\sum_{i=1}^{10} a_i\) a~b \(\alpha \
... = \frac{\beta}{\zeta^2}\)"], ["%#^ \( \$x^2 \)"]])
>>> df.style.format(escape="latex-math").to_latex()
... # doctest: +SKIP
\begin{tabular}{ll}
& 0 \\
0 & \(\sum_{i=1}^{10} a_i\) a\textasciitilde b \(\alpha
= \frac{\beta}{\zeta^2}\) \\
1 & \%\#\textasciicircum \space \( \$x^2 \) \\
\end{tabular}

Pandas defines a `number-format` pseudo CSS attribute instead of the `.format`
method to create `to_excel` permissible formatting. Note that semi-colons are
CSS protected characters but used as separators in Excel's format string.
Expand Down Expand Up @@ -2344,7 +2359,8 @@ def _escape_latex(s):
Escaped string
"""
return (
s.replace("\\", "ab2§=§8yz") # rare string for final conversion: avoid \\ clash
s.replace("\\ ", "ab2§=§8yz")
.replace("\\", "ab2§=§8yz") # rare string for final conversion: avoid \\ clash
.replace("ab2§=§8yz ", "ab2§=§8yz\\space ") # since \backslash gobbles spaces
.replace("&", "\\&")
.replace("%", "\\%")
Expand All @@ -2363,11 +2379,11 @@ def _escape_latex(s):

def _escape_latex_math(s):
r"""
All characters between two characters ``$`` are preserved.
All characters in LaTeX math mode are preserved.

The substrings in LaTeX math mode, which start with the character ``$``
and end with ``$``, are preserved without escaping. Otherwise
regular LaTeX escaping applies. See ``_escape_latex()``.
The substrings in LaTeX math mode, which either are surrounded
by two characters ``$`` or start with the character ``\(`` and end with ``\)``,
are preserved without escaping. Otherwise regular LaTeX escaping applies.

Parameters
----------
Expand All @@ -2379,16 +2395,55 @@ def _escape_latex_math(s):
str :
Escaped string
"""
s = s.replace(r"\$", r"rt8§=§7wz")
pattern = re.compile(r"\$.*?\$")
pos = 0
ps = pattern.search(s, pos)
res = []
while ps:
res.append(_escape_latex(s[pos : ps.span()[0]]))
res.append(ps.group())
pos = ps.span()[1]

def _math_mode_with_dollar(s):
s = s.replace(r"\$", r"rt8§=§7wz")
pattern = re.compile(r"\$.*?\$")
pos = 0
ps = pattern.search(s, pos)
res = []
while ps:
res.append(_escape_latex(s[pos : ps.span()[0]]))
res.append(ps.group())
pos = ps.span()[1]
ps = pattern.search(s, pos)

res.append(_escape_latex(s[pos : len(s)]))
return "".join(res).replace(r"rt8§=§7wz", r"\$")

def _math_mode_with_parentheses(s):
s = s.replace(r"\(", r"LEFT§=§6yzLEFT").replace(r"\)", r"RIGHTab5§=§RIGHT")
res = []
for item in re.split(r"LEFT§=§6yz|ab5§=§RIGHT", s):
if item.startswith("LEFT") and item.endswith("RIGHT"):
res.append(item.replace("LEFT", r"\(").replace("RIGHT", r"\)"))
elif "LEFT" in item and "RIGHT" in item:
res.append(
_escape_latex(item).replace("LEFT", r"\(").replace("RIGHT", r"\)")
)
else:
res.append(
_escape_latex(item)
.replace("LEFT", r"\textbackslash (")
.replace("RIGHT", r"\textbackslash )")
)
return "".join(res)

res.append(_escape_latex(s[pos : len(s)]))
return "".join(res).replace(r"rt8§=§7wz", r"\$")
s = s.replace(r"\$", r"rt8§=§7wz")
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't understand this code.
You are replacing the string \$ with a uuid string, first.
Then, you are searching for a pattern (r"\$.*?\$") that cannot exist, since it was replaced.

Then in line 2490 you are replacing the same string s again with the same uuid, but this is unnecessary since it has already done this in line 2481.
I think your tests pass and this does the correct thing but I think some of these lines are redundant and do nothing for the overall effect??

Copy link
Contributor Author

@natmokval natmokval Mar 25, 2023 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you for the comment.
It’s right, the replacement in line 2490 has a mistake. The right one would be the reverse replacement: s.replace(r"rt8§=§7wz", r"\$")

I can explain what I am trying to do. When I checked the function _escape_latex_math I noticed that for a string like r"$&%#^$" which contains only one sign "$" and only one combination "\$" I got a wrong result, because the function processed this string in math mode. By doing the replacement in line 2481 I exclude from the consideration the string “\$” to avoid confusing it with “$”. Then I get the correct result and do the reverse replacement. If we don’t have a combination of one sign "$" and one sign "\$" we don’t need to do this check, but I prefer to leave it.

I corrected my mistake a made a new commit. I also added an example for this case in the test.

pattern_d = re.compile(r"\$.*?\$")
pattern_p = re.compile(r"\\(.*?\\)")
pos_d = 0
pos_p = 0
ps_d = pattern_d.search(s, pos_d)
ps_p = pattern_p.search(s, pos_p)
mode = []
if ps_d:
mode.append(ps_d.span()[0])
if ps_p:
mode.append(ps_p.span()[0])
if len(mode) == 0:
return _escape_latex(s.replace(r"\$", r"rt8§=§7wz"))
if s[min(mode)] == r"$":
return _math_mode_with_dollar(s.replace(r"\$", r"rt8§=§7wz"))
else:
return _math_mode_with_parentheses(s.replace(r"\$", r"rt8§=§7wz"))
46 changes: 40 additions & 6 deletions pandas/tests/io/formats/style/test_format.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -167,7 +167,7 @@ def test_format_clear(styler, func, attr, kwargs):
"latex",
'<>\\&"\\%\\$\\#\\_\\{\\}\\textasciitilde \\textasciicircum '
"\\textbackslash \\textasciitilde \\space \\textasciicircum \\space "
"\\textbackslash \\space ",
"\\textbackslash ",
),
],
)
Expand All @@ -192,13 +192,47 @@ def test_format_escape_html(escape, exp):
assert styler._translate(True, True)["head"][0][1]["display_value"] == f"&{exp}&"


def test_format_escape_latex_math():
chars = r"$\frac{1}{2} \$ x^2$ ~%#^"
@pytest.mark.parametrize(
"chars, expected",
[
(
r"$ \$&%#_{}~^\ $ &%#_{}~^\ $",
"".join(
[
r"$ \$&%#_{}~^\ $ ",
r"\&\%\#\_\{\}\textasciitilde \textasciicircum \textbackslash \$",
]
),
),
(
r"\( &%#_{}~^\ \) &%#_{}~^\ \(",
"".join(
[
r"\( &%#_{}~^\ \) ",
r"\&\%\#\_\{\}\textasciitilde \textasciicircum ",
r"\textbackslash \textbackslash (",
]
),
),
(
r"$ \frac{1}{2} $ \( \frac{1}{2} \)",
"".join(
[
r"$ \frac{1}{2} $",
r" \textbackslash ( \textbackslash frac\{1\}\{2\} \textbackslash )",
]
),
),
],
)
def test_format_escape_latex_math(chars, expected):
# GH 51903
# latex-math escape works for each DataFrame cell separately.
# If we have a combination of dollar signs and brackets,
# the sign which occurs first would apply.
df = DataFrame([[chars]])

expected = r"$\frac{1}{2} \$ x^2$ \textasciitilde \%\#\textasciicircum "
s = df.style.format("{0}", escape="latex-math")
assert expected == s._translate(True, True)["body"][0][1]["display_value"]
assert s._translate(True, True)["body"][0][1]["display_value"] == expected


def test_format_escape_na_rep():
Expand Down