1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206
|
// Copyright Nat Goodspeed 2013.
// Distributed under the Boost Software License, Version 1.0.
// (See accompanying file LICENSE_1_0.txt or copy at
// http://www.boost.org/LICENSE_1_0.txt)
#include <boost/coroutine/all.hpp>
#include <iostream>
#include <iomanip>
#include <string>
#include <cctype>
#include <sstream>
#include <boost/bind.hpp>
#include <boost/foreach.hpp>
typedef boost::coroutines::asymmetric_coroutine<std::string> coro_t;
// deliver each line of input stream to sink as a separate string
void readlines(coro_t::push_type& sink, std::istream& in)
{
std::string line;
while (std::getline(in, line))
sink(line);
}
void tokenize(coro_t::push_type& sink, coro_t::pull_type& source)
{
// This tokenizer doesn't happen to be stateful: you could reasonably
// implement it with a single call to push each new token downstream. But
// I've worked with stateful tokenizers, in which the meaning of input
// characters depends in part on their position within the input line. At
// the time, I wished for a way to resume at the suspend point!
BOOST_FOREACH(std::string line, source)
{
std::string::size_type pos = 0;
while (pos < line.length())
{
if (line[pos] == '"')
{
std::string token;
++pos; // skip open quote
while (pos < line.length() && line[pos] != '"')
token += line[pos++];
++pos; // skip close quote
sink(token); // pass token downstream
}
else if (std::isspace(line[pos]))
{
++pos; // outside quotes, ignore whitespace
}
else if (std::isalpha(line[pos]))
{
std::string token;
while (pos < line.length() && std::isalpha(line[pos]))
token += line[pos++];
sink(token); // pass token downstream
}
else // punctuation
{
sink(std::string(1, line[pos++]));
}
}
}
}
void only_words(coro_t::push_type& sink, coro_t::pull_type& source)
{
BOOST_FOREACH(std::string token, source)
{
if (! token.empty() && std::isalpha(token[0]))
sink(token);
}
}
void trace(coro_t::push_type& sink, coro_t::pull_type& source)
{
BOOST_FOREACH(std::string token, source)
{
std::cout << "trace: '" << token << "'\n";
sink(token);
}
}
struct FinalEOL
{
~FinalEOL() { std::cout << std::endl; }
};
void layout(coro_t::pull_type& source, int num, int width)
{
// Finish the last line when we leave by whatever means
FinalEOL eol;
// Pull values from upstream, lay them out 'num' to a line
for (;;)
{
for (int i = 0; i < num; ++i)
{
// when we exhaust the input, stop
if (! source)
return;
std::cout << std::setw(width) << source.get();
// now that we've handled this item, advance to next
source();
}
// after 'num' items, line break
std::cout << std::endl;
}
}
int main(int argc, char *argv[])
{
// For example purposes, instead of having a separate text file in the
// local filesystem, construct an istringstream to read.
std::string data(
"This is the first line.\n"
"This, the second.\n"
"The third has \"a phrase\"!\n"
);
{
std::cout << "\nreadlines:\n";
std::istringstream infile(data);
// Each coroutine-function has a small, specific job to do. Instead of
// adding conditional logic to a large, complex input function, the
// caller composes smaller functions into the desired processing
// chain.
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tracer(boost::bind(trace, _1, boost::ref(reader)));
BOOST_FOREACH(std::string line, tracer)
{
std::cout << "got: " << line << "\n";
}
}
{
std::cout << "\ncompose a chain:\n";
std::istringstream infile(data);
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tokenizer(boost::bind(tokenize, _1, boost::ref(reader)));
coro_t::pull_type tracer(boost::bind(trace, _1, boost::ref(tokenizer)));
BOOST_FOREACH(std::string token, tracer)
{
// just iterate, we're already pulling through tracer
}
}
{
std::cout << "\nfilter:\n";
std::istringstream infile(data);
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tokenizer(boost::bind(tokenize, _1, boost::ref(reader)));
coro_t::pull_type filter(boost::bind(only_words, _1, boost::ref(tokenizer)));
coro_t::pull_type tracer(boost::bind(trace, _1, boost::ref(filter)));
BOOST_FOREACH(std::string token, tracer)
{
// just iterate, we're already pulling through tracer
}
}
{
std::cout << "\nlayout() as coroutine::push_type:\n";
std::istringstream infile(data);
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tokenizer(boost::bind(tokenize, _1, boost::ref(reader)));
coro_t::pull_type filter(boost::bind(only_words, _1, boost::ref(tokenizer)));
coro_t::push_type writer(boost::bind(layout, _1, 5, 15));
BOOST_FOREACH(std::string token, filter)
{
writer(token);
}
}
{
std::cout << "\ncalling layout() directly:\n";
std::istringstream infile(data);
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tokenizer(boost::bind(tokenize, _1, boost::ref(reader)));
coro_t::pull_type filter(boost::bind(only_words, _1, boost::ref(tokenizer)));
// Because of the symmetry of the API, we can directly call layout()
// instead of using it as a coroutine-function.
layout(filter, 5, 15);
}
{
std::cout << "\nfiltering output:\n";
std::istringstream infile(data);
coro_t::pull_type reader(boost::bind(readlines, _1, boost::ref(infile)));
coro_t::pull_type tokenizer(boost::bind(tokenize, _1, boost::ref(reader)));
coro_t::push_type writer(boost::bind(layout, _1, 5, 15));
// Because of the symmetry of the API, we can use any of these
// chaining functions in a push_type coroutine chain as well.
coro_t::push_type filter(boost::bind(only_words, boost::ref(writer), _1));
BOOST_FOREACH(std::string token, tokenizer)
{
filter(token);
}
}
std::cout << "\nDone" << std::endl;
return EXIT_SUCCESS;
}
|