Initial commit

.gitignore

@@ -0,0 +1,73 @@
+# This file is used to ignore files which are generated
+# ----------------------------------------------------------------------------
+
+*~
+*.autosave
+*.a
+*.core
+*.moc
+*.o
+*.obj
+*.orig
+*.rej
+*.so
+*.so.*
+*_pch.h.cpp
+*_resource.rc
+*.qm
+.#*
+*.*#
+core
+!core/
+tags
+.DS_Store
+.directory
+*.debug
+Makefile*
+*.prl
+*.app
+moc_*.cpp
+ui_*.h
+qrc_*.cpp
+Thumbs.db
+*.res
+*.rc
+/.qmake.cache
+/.qmake.stash
+
+# qtcreator generated files
+*.pro.user*
+
+# xemacs temporary files
+*.flc
+
+# Vim temporary files
+.*.swp
+
+# Visual Studio generated files
+*.ib_pdb_index
+*.idb
+*.ilk
+*.pdb
+*.sln
+*.suo
+*.vcproj
+*vcproj.*.*.user
+*.ncb
+*.sdf
+*.opensdf
+*.vcxproj
+*vcxproj.*
+
+# MinGW generated files
+*.Debug
+*.Release
+
+# Python byte code
+*.pyc
+
+# Binaries
+# --------
+*.dll
+*.exe
+

CMakeLists.txt

@@ -0,0 +1,6 @@
+project(pdfrender)
+cmake_minimum_required(VERSION 2.8)
+aux_source_directory(. SRC_LIST)
+add_executable(${PROJECT_NAME} ${SRC_LIST})
+target_compile_options(${PROJECT_NAME} PRIVATE -g -std=c++11)
+target_link_libraries(${PROJECT_NAME} poppler-cpp pthread boost_program_options)

main.cpp

@@ -0,0 +1,145 @@
+#include <cstdlib>
+#include <cstdio>
+#include <iostream>
+#include <string>
+#include <future>
+#include <regex>
+
+#include <boost/program_options.hpp>
+#include <poppler/cpp/poppler-page.h>
+#include <poppler/cpp/poppler-document.h>
+#include <poppler/cpp/poppler-page-renderer.h>
+
+using namespace std;
+using namespace poppler;
+
+namespace po = boost::program_options;
+
+int dpi = 75;
+string inputpdf;
+string output;
+int threads = thread::hardware_concurrency();
+
+void render_page(document *doc, int pageno, const string& fmt) {
+    page *p = doc->create_page(pageno);
+    page_renderer r;
+    image img = r.render_page(p);
+    vector<char> buf(1024);
+    int size = snprintf(&buf[0], buf.size(), fmt.c_str(), pageno);
+    if (size >= buf.size()) {
+        buf.reserve(size + 1);
+        snprintf(&buf[0], buf.size(), fmt.c_str(), pageno);
+    }
+    img.save(&buf[0], "JPEG", dpi);
+}
+
+void render_pages(document *doc, int first, int last, int step, const string& fmt) {
+    for (int p=first; p<=last; p+=step) {
+        render_page(doc, p, fmt);
+    }
+}
+
+bool validate_output(const string& output)
+{
+    if (output.empty()) {
+        cout << "Output format must be specified" << endl;
+        return false;
+    }
+
+    regex fmt_regex(R"(%(?:0(\d+))?d)");
+
+    auto beg = sregex_iterator(output.begin(), output.end(), fmt_regex);
+    auto end = sregex_iterator();
+
+    int matches = distance(beg, end);
+    if (matches < 1) {
+        cout << "Output format must contain %d or %0<padding>d" << endl;
+        return false;
+    }
+
+    if (matches > 1) {
+        cout << "Output format must contain only one %d or %0<padding>d" << endl;
+        return false;
+    }
+
+    auto m = *beg;
+    if (!(m[1].str().empty())) {
+        int padding = 0;
+        try {
+            padding = stoi(m[1].str());
+        } catch (...) {
+            cout << "error: invalid interger value for padding: " << m[1].str() << endl;
+            return false;
+        }
+        if (padding > 10) {
+            cout << "error: padding is too large: " << m[1].str() << endl;
+            return false;
+        }
+    }
+    return true;
+}
+
+void parse_args(int argc, char *argv[])
+{
+    po::options_description opts(string("Usage: ") + argv[0] + " [options] input.pdf\nOptions:");
+    opts.add_options()
+        ("help,h", "prints this message")
+        ("output,o", po::value<string>(&output),
+            "output format; must contain exactly one instance of %d or "
+            "%0<padding>d where <padding> is an integer for the page number; "
+            "padding, if used, will always be with zeros")
+        ("dpi", po::value<int>(&dpi)->default_value(75),
+            "render at the specified dpi")
+        ("threads,t", po::value<int>(&threads),
+            "use this many threads to render; by default, uses "
+            "the result of std::thread::hardware_concurrency()")
+        ;
+
+    po::options_description hidden("Hidden");
+    hidden.add_options()
+        ("input", po::value<string>(&inputpdf))
+        ;
+
+    po::options_description allopts("");
+    allopts.add(opts).add(hidden);
+
+    po::positional_options_description pos;
+    pos.add("input", 1);
+
+    po::variables_map vm;
+    po::store(
+        po::command_line_parser(argc, argv)
+            .options(allopts)
+            .positional(pos)
+            .run(),
+        vm
+    );
+    po::notify(vm);
+
+    if (inputpdf.empty() || vm.count("help")) {
+        cout << opts;
+        exit(1);
+    }
+
+    if (!validate_output(output)) {
+        cout << "See " << argv[0] << " -h for more info" << endl;
+        exit(1);
+    }
+}
+
+int main(int argc, char *argv[])
+{
+    parse_args(argc, argv);
+
+    document *doc = document::load_from_file(inputpdf);
+    
+    vector<future<void>> futures;
+    int pagecount = doc->pages();
+    for (int i = 1; i <= threads; i++) {
+        futures.push_back(
+            async(launch::async,
+                render_pages, doc, i, pagecount, threads, output
+            )
+        );
+    }
+}