tex2png.py
Go to the documentation of this file.
00001 #!/usr/bin/env python
00002 #   Aseba - an event-based framework for distributed robot control
00003 #   Copyright (C) 2007--2011:
00004 #           Stephane Magnenat <stephane at magnenat dot net>
00005 #           (http://stephane.magnenat.net)
00006 #           and other contributors, see authors.txt for details
00007 #
00008 #   This program is free software: you can redistribute it and/or modify
00009 #   it under the terms of the GNU Lesser General Public License as published
00010 #   by the Free Software Foundation, version 3 of the License.
00011 #
00012 #   This program is distributed in the hope that it will be useful,
00013 #   but WITHOUT ANY WARRANTY; without even the implied warranty of
00014 #   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
00015 #   GNU Lesser General Public License for more details.
00016 #
00017 #   You should have received a copy of the GNU Lesser General Public License
00018 #   along with this program. If not, see <http://www.gnu.org/licenses/>.
00019 
00020 import os
00021 import os.path
00022 import shutil
00023 import sys
00024 import re
00025 from string import Template
00026 import subprocess
00027 
00028 tex_file = \
00029 r"""
00030 \documentclass{article}
00031 \pagestyle{empty}
00032 \begin{document}
00033 ${formula}
00034 \end{document}
00035 """
00036 
00037 tex_file_template = Template(tex_file)
00038 tex_math_re = re.compile(r"<span class=\"math-inline\">(.*?)</span>")
00039 
00040 def from_tex(tex_code, filename, density = 100):
00041     """Given a LaTeX string, generates a PNG file at a given density.
00042     Temporary files are generated in the current folder, prefixed by filename, 
00043     then erased.
00044 
00045     Inputs:
00046         tex_code:       String with a valid LaTeX code (ex: $a_i = b_i$)
00047         filename        Filename, without the extension. Can be located in another directory
00048         density         (Optional) Output density, by default at 100
00049 
00050     Output:
00051         No output."""
00052 
00053     basename = os.path.basename(filename)   # process in the current folder, result will be copied later
00054     f = open("{0}.tex".format(basename),'w')
00055     f.write(tex_file_template.substitute(formula=tex_code))
00056     f.close()
00057     devnull = open(os.devnull)
00058     retcode = subprocess.call(["texi2dvi","{0}.tex".format(basename)], stdout=devnull, stderr=devnull)
00059     retcode = subprocess.call(["dvips","-E","{0}.dvi".format(basename)], stdout=devnull, stderr=devnull)
00060     retcode = subprocess.call(["convert","-density","{0}x{0}".format(density),"{0}.ps".format(basename),"{0}.png".format(basename)])
00061     # copy the result to the final file (if directory differes)
00062     if basename != filename:
00063         local_png = "{0}.png".format(basename)
00064         new_png = "{0}.png".format(filename)
00065         shutil.copy(local_png, new_png)
00066         os.remove(local_png)
00067     os.remove("{0}.tex".format(basename))
00068     os.remove("{0}.aux".format(basename))
00069     os.remove("{0}.dvi".format(basename))
00070     os.remove("{0}.log".format(basename))
00071     os.remove("{0}.ps".format(basename))
00072 
00073 def from_html(source_html_file, output_basename):
00074     """Given an HTML file, convert all LaTeX codes enclosed between <span class="math-inline"> tags.
00075 
00076     Inputs:
00077         source_html_file    Path to the HTML file
00078         output_basename     Basename for the output files, without any extension.
00079                             Corresponding .html and .png files will be generated.
00080 
00081     Output:
00082         No output."""
00083 
00084     try:
00085         f = open(source_html_file, 'r')
00086     except IOError:
00087         print >> sys.stderr, "No such file: {0}".format(source_html_file)
00088         return
00089     else:
00090         html = f.read()
00091         f.close()
00092 
00093     # counter
00094     i = 0
00095 
00096     # lookup dictionnary
00097     tex_lookup = dict()
00098 
00099     # loop search for the regular expression
00100     pos = 0
00101     while 1:
00102         match = tex_math_re.search(html, pos)
00103         if match:
00104             tex_expr = match.group(1)
00105             if tex_lookup.has_key(tex_expr):
00106                 # already encountered this expression, retrive the file
00107                 filename = tex_lookup[tex_expr]
00108             else:
00109                 # we must convert to png
00110                 filename = output_basename + "-eq" + str(i)
00111                 i += 1
00112                 print "   " + filename + ".png"
00113                 from_tex(tex_expr, filename)
00114                 # register into the lookup
00115                 tex_lookup[tex_expr] = filename
00116             #html = tex_math_re.sub("<img class=inline-math src=\"" + os.path.basename(filename) + ".png\">", html, 1)
00117             html = tex_math_re.sub("<img src=\"" + os.path.basename(filename) + ".png\">", html, 1)
00118         else:
00119             break
00120 
00121     # Save the file
00122     output_html = output_basename + ".html"
00123     try:
00124         f = open(output_html, 'w')
00125     except IOError:
00126         print >> sys.stderr, "Unable to write the result to: {0}".format(output_html)
00127         return
00128     else:
00129         f.write(html)
00130         f.close()
00131 
00132 # Test the module
00133 if __name__ == "__main__":
00134 #    from_tex(r"$\frac{a}{b} = c$", "test", 200)
00135 #    from_html("test.html", "test-out")
00136     from_html("en_asebastdnative.html", "test-out")


aseba
Author(s): Stéphane Magnenat
autogenerated on Sun Oct 5 2014 23:46:39