Merge pull request #2636 from mdboom/font/unicode-python3-pdf

mdboom · mdboom · commit 45ee98ca2ae3 · 2013-12-06T05:37:13.000-08:00
"\usepackage[russian]{babel}" does not work in matplotlib 1.3.1
diff --git a/lib/matplotlib/backends/backend_pdf.py b/lib/matplotlib/backends/backend_pdf.py
@@ -1745,7 +1745,9 @@ def draw_tex(self, gc, x, y, s, prop, angle, ismath='TeX!', mtext=None):
                         dvifont=dvifont)
                 seq += [['font', pdfname, dvifont.size]]
                 oldfont = dvifont
-            seq += [['text', x1, y1, [chr(glyph)], x1+width]]
+            # We need to convert the glyph numbers to bytes, and the easiest
+            # way to do this on both Python 2 and 3 is .encode('latin-1')
+            seq += [['text', x1, y1, [chr(glyph).encode('latin-1')], x1+width]]
 
         # Find consecutive text strings with constant y coordinate and
         # combine into a sequence of strings and kerns, or just one

Original file line number	Diff line number	Diff line change
`@@ -1745,7 +1745,9 @@ def draw_tex(self, gc, x, y, s, prop, angle, ismath='TeX!', mtext=None):`
`1745`	`1745`	`dvifont=dvifont)`
`1746`	`1746`	`seq += [['font', pdfname, dvifont.size]]`
`1747`	`1747`	`oldfont = dvifont`
`1748`		`- seq += [['text', x1, y1, [chr(glyph)], x1+width]]`
	`1748`	`+ # We need to convert the glyph numbers to bytes, and the easiest`
	`1749`	`+ # way to do this on both Python 2 and 3 is .encode('latin-1')`
	`1750`	`+ seq += [['text', x1, y1, [chr(glyph).encode('latin-1')], x1+width]]`
`1749`	`1751`
`1750`	`1752`	`# Find consecutive text strings with constant y coordinate and`
`1751`	`1753`	`# combine into a sequence of strings and kerns, or just one`