Question

This is more of an algorithmic dilemma than a language-specific problem, but since I’m currently using Ruby I’ll tag this as such. I’ve already spent over 20 hours on this and I would’ve never believed it if someone told me writing a LaTeX parser was a walk in the park in comparison.

I have a loop to read hierarchies (that are prefixed with \m) from different files

art.tex: \m{Art}
graphical.tex: \m{Art}{Graphical}
me.tex: \m{About}{Me}
music.tex: \m{Art}{Music}
notes.tex: \m{Art}{Music}{Sheet Music}
site.tex: \m{About}{Site}
something.tex: \m{Something}
whatever.tex: \m{Something}{That}{Does Not}{Matter}

and I need to sort them alphabetically and print them out as a tree

About
    Me (me.tex)
    Site (site.tex)
Art (art.tex)
    Graphical (graphical.tex)
    Music (music.tex)
        Sheet Music (notes.tex)
Something (something.tex)
    That
        Does Not
            Matter (whatever.tex)

in (X)HTML

<ul>
<li>About</li>
<ul>
<li><a href="me.tex">Me</a></li>
<li><a href="site.tex">Site</a></li>
</ul>
<li><a href="art.tex">Art</a></li>
<ul>
<li><a href="graphical.tex">Graphical</a></li>
<li><a href="music.tex">Music</a></li>
<ul>
<li><a href="notes.tex">Sheet Music</a></li>
</ul>
</ul>
<li><a href="something.tex">Something</a></li>
<ul>
<li>That</li>
<ul>
<li>Doesn't</li>
<ul>
<li><a href="whatever.tex">Matter</a></li>
</ul>
</ul>
</ul>
</ul>

using Ruby without Rails, which means that at least Array.sort and Dir.glob are available.

All of my attempts were formed like this (as this part should work just fine).

def fss_brace_array(ss_input)#a concise version of another function; converts {1}{2}...{n} into an array [1, 2, ..., n] or returns an empty array
    ss_output = ss_input[1].scan(%r{\{(.*?)\}})
rescue
    ss_output = []
ensure
    return ss_output
end
#define tree
s_handle = File.join(:content.to_s, "*")
Dir.glob("#{s_handle}.tex").each do |s_handle|
    File.open(s_handle, "r") do |f_handle|
        while s_line = f_handle.gets
            if s_all = s_line.match(%r{\\m\{(\{.*?\})+\}})
                s_all = s_all.to_a
                #do something with tree, fss_brace_array(s_all) and s_handle
                break
            end
        end
    end
end
#do something else with tree

Answer 1

Here’s a working solution in Python after simply pushing all of the entries into an array.

import operator
import itertools
def overput(input):
    for betweenput, output in itertools.groupby(input, key=operator.itemgetter(0)):
        yield '<li>'
        yield betweenput
        output = [x[1:] for x in output if len(x) > 1]
        if output:
            yield '<ul>'
            for part in overput(output):
                yield part
            yield '</ul>'
        yield '</li>'