Support non-ASCII characters in function arguments

examples/gallery/embellishments/colorbar.py

@@ -40,7 +40,7 @@
     # with a length/width (+w) of 4 cm by 0.5 cm, and plotted horizontally (+h)
     position="g0.3/8.7+w4c/0.5c+h",
     box=True,
-    frame=["x+lTemperature", r"y+l\260C"],
+    frame=["x+lTemperature", "y+l°C"],
     scale=100,
 )
 

examples/gallery/symbols/text_symbols.py

@@ -36,9 +36,8 @@
 # plot a lowercase "s" of size 3.5c and use the "Times-Italic" font,
 # color fill is set to "gold"
 fig.plot(x=5.5, y=1.5, style="l3.5c+ts+fTimes-Italic", fill="gold", pen=pen)
-# plot the pi symbol (\160 is octal code for pi) of size 3.5c, for this use
-# the "Symbol" font, the outline color of the symbol is set to
+# plot the pi symbol of size 3.5c, the outline color of the symbol is set to
 # "darkorange", the color fill is set to "magenta4"
-fig.plot(x=7, y=1.5, style="l3.5c+t\160+fSymbol,darkorange", fill="magenta4", pen=pen)
+fig.plot(x=7, y=1.5, style="l3.5c+tπ+fdarkorange", fill="magenta4", pen=pen)
 
 fig.show()

pygmt/helpers/utils.py

@@ -4,6 +4,7 @@
 import os
 import pathlib
 import shutil
+import string
 import subprocess
 import sys
 import time
@@ -145,6 +146,119 @@ def data_kind(data=None, x=None, y=None, z=None, required_z=False):
     return kind
 
 
+def non_ascii_to_octal(argstr):
+    r"""
+    Translate non-ASCII characters to their corresponding octal codes.
+
+    Currently, only characters in the ISOLatin1+ charset and
+    Symbol/ZapfDingbats fonts are supported.
+
+    Parameters
+    ----------
+    argstr : str
+        The string to be translated.
+
+    Returns
+    -------
+    str
+        The translated string.
+
+    Examples
+    --------
+    >>> non_ascii_to_octal("•‰“”±°ÿ")
+    '\\31\\214\\216\\217\\261\\260\\377'
+    >>> non_ascii_to_octal("αζΔΩ∑π∇")
+    '@~\\141@~@~\\172@~@~\\104@~@~\\127@~@~\\344@~@~\\160@~@~\\321@~'
+    >>> non_ascii_to_octal("✁❞❡➾")
+    '@%34%\\41@%%@%34%\\176@%%@%34%\\241@%%@%34%\\376@%%'
+    >>> non_ascii_to_octal("ABC ±120° DEF α ♥")
+    'ABC \\261120\\260 DEF @~\\141@~ @%34%\\252@%%'
+    """
+    # Dictionary mapping non-ASCII characters to octal codes
+    mapping = {}
+
+    # Adobe Symbol charset
+    # References:
+    # 1. https://en.wikipedia.org/wiki/Symbol_(typeface)
+    # 2. https://unicode.org/Public/MAPPINGS/VENDORS/ADOBE/symbol.txt
+    # Notes:
+    # 1. \322 and \342 are "REGISTERED SIGN SERIF" and
+    #    "REGISTERED SIGN SANS SERIF" respectively, but only "REGISTERED SIGN"
+    #    is available in the unicode table. So both are mapped to
+    #    "REGISTERED SIGN". \323, \343, \324 and \344 also have the same
+    #    problem.
+    # 2. Characters for \140, \275, \276 are incorrect.
+    mapping.update(
+        {
+            c: "@~\\" + format(i, "o") + "@~"
+            for c, i in zip(
+                " !∀#∃%&∋()∗+,−./"  # \04x-05x
+                + "0123456789:;<=>?"  # \06x-07x
+                + "≅ΑΒΧΔΕΦΓΗΙϑΚΛΜΝΟ"  # \10x-11x
+                + "ΠΘΡΣΤΥςΩΞΨΖ[∴]⊥_"  # \12x-13x
+                + "αβχδεφγηιϕκλμνο"  # \14x-15x
+                + "πθρστυϖωξψζ{|}∼"  # \16x-17x. \177 is undefined
+                + "€ϒ′≤⁄∞ƒ♣♦♥♠↔←↑→↓"  # \24x-\25x
+                + "°±″≥×∝∂•÷≠≡≈…↵"  # \26x-27x
+                + "ℵℑℜ℘⊗⊕∅∩∪⊃⊇⊄⊂⊆∈∉"  # \30x-31x
+                + "∠∇®©™∏√⋅¬∧∨⇔⇐⇑⇒⇓"  # \32x-33x
+                + "◊〈®©™∑"  # \34x-35x
+                + "〉∫⌠⌡",  # \36x-37x. \360 and \377 are undefined
+                [*range(32, 127), *range(160, 240), *range(241, 255)],
+            )
+        }
+    )
+
+    # Adobe ZapfDingbats charset
+    # References:
+    # 1. https://en.wikipedia.org/wiki/Zapf_Dingbats
+    # 2. https://unicode.org/Public/MAPPINGS/VENDORS/ADOBE/zdingbat.txt
+    mapping.update(
+        {
+            c: "@%34%\\" + format(i, "o") + "@%%"
+            for c, i in zip(
+                " ✁✂✃✄☎✆✇✈✉☛☞✌✍✎✏"  # \04x-\05x
+                + "✐✑✒✓✔✕✖✗✘✙✚✛✜✝✞✟"  # \06x-\07x
+                + "✠✡✢✣✤✥✦✧★✩✪✫✬✭✮✯"  # \10x-\11x
+                + "✰✱✲✳✴✵✶✷✸✹✺✻✼✽✾✿"  # \12x-\13x
+                + "❀❁❂❃❄❅❆❇❈❉❊❋●❍■❏"  # \14x-\15x
+                + "❐❑❒▲▼◆❖◗❘❙❚❛❜❝❞"  # \16x-\17x. \177 is undefined
+                + "❡❢❣❤❥❦❧♣♦♥♠①②③④"  # \24x-\25x. \240 is undefined
+                + "⑤⑥⑦⑧⑨⑩❶❷❸❹❺❻❼❽❾❿"  # \26x-\27x
+                + "➀➁➂➃➄➅➆➇➈➉➊➋➌➍➎➏"  # \30x-\31x
+                + "➐➑➒➓➔→↔↕➘➙➚➛➜➝➞➟"  # \32x-\33x
+                + "➠➡➢➣➤➥➦➧➨➩➪➫➬➭➮➯"  # \34x-\35x
+                + "➱➲➳➴➵➶➷➸➹➺➻➼➽➾",  # \36x-\37x. \360 and \377 are undefined
+                [*range(32, 127), *range(161, 240), *range(241, 255)],
+            )
+        }
+    )
+
+    # Adobe ISOLatin1+ charset (i.e., ISO-8859-1 with extensions)
+    # References:
+    # 1. https://en.wikipedia.org/wiki/ISO/IEC_8859-1
+    # 2. https://docs.generic-mapping-tools.org/dev/cookbook/octal-codes.html
+    # 3. https://www.adobe.com/jp/print/postscript/pdfs/PLRM.pdf
+    mapping.update(
+        {
+            c: "\\" + format(i, "o")
+            for c, i in zip(
+                "•…™—–fiž"  # \03x. \030 is undefined
+                + "š"  # \177
+                + "Œ†‡Ł⁄‹Š›œŸŽł‰„“”"  # \20x-\21x
+                + "ı`´ˆ˜¯˘˙¨‚˚¸'˝˛ˇ",  # \22x-\23x
+                [*range(25, 32), *range(127, 160)],
+            )
+        }
+    )
+    # \240-\377
+    mapping.update({chr(i): "\\" + format(i, "o") for i in range(160, 256)})
+
+    # Remove any printable characters
+    mapping = {k: v for k, v in mapping.items() if k not in string.printable}
+    return argstr.translate(str.maketrans(mapping))
+
+
 def build_arg_string(kwdict, confdict=None, infile=None, outfile=None):
     r"""
     Convert keyword dictionaries and input/output files into a GMT argument
@@ -267,7 +381,7 @@ def build_arg_string(kwdict, confdict=None, infile=None, outfile=None):
         gmt_args = [str(infile)] + gmt_args
     if outfile:
         gmt_args.append("->" + str(outfile))
-    return " ".join(gmt_args)
+    return non_ascii_to_octal(" ".join(gmt_args))
 
 
 def is_nonstr_iter(value):

pygmt/tests/baseline/test_non_ascii_to_octal.png.dvc

@@ -0,0 +1,4 @@
+outs:
+- md5: d93bed7495d77cd2ef7cc1b64edb9b3a
+  size: 19563
+  path: test_non_ascii_to_octal.png

pygmt/tests/test_helpers.py

@@ -6,6 +6,7 @@
 import numpy as np
 import pytest
 import xarray as xr
+from pygmt import Figure
 from pygmt.exceptions import GMTInvalidInput
 from pygmt.helpers import (
     GMTTempFile,
@@ -57,6 +58,24 @@ def test_unique_name():
     assert len(names) == len(set(names))
 
 
+@pytest.mark.mpl_image_compare
+def test_non_ascii_to_octal():
+    """
+    Test support of non-ASCII characters.
+    """
+    fig = Figure()
+    fig.basemap(
+        region=[0, 10, 0, 5],
+        projection="X10c/5c",
+        frame=[
+            "xaf+lISOLatin1: fi‰“”¥",
+            "yaf+lSymbol: αβ∇∋∈",
+            "WSen+tZapfDingbats: ①❷➂➍✦❝❞",
+        ],
+    )
+    return fig
+
+
 def test_kwargs_to_strings_fails():
     """
     Make sure it fails for invalid conversion types.