summaryrefslogtreecommitdiff
path: root/klm/lm/builder/print.cc
diff options
context:
space:
mode:
authorPatrick Simianer <p@simianer.de>2013-05-02 09:09:59 +0200
committerPatrick Simianer <p@simianer.de>2013-05-02 09:09:59 +0200
commit9e50f0237413180fba11b500c9dce5c600e3c157 (patch)
tree556fc31d231353c853a864afffddd43dc525549a /klm/lm/builder/print.cc
parentd18024a41cbc1b54db88d499571349a6234b6db8 (diff)
parent14ed53426726202813a8e82d706b44266f015fe1 (diff)
Merge remote-tracking branch 'upstream/master'
Diffstat (limited to 'klm/lm/builder/print.cc')
-rw-r--r--klm/lm/builder/print.cc74
1 files changed, 3 insertions, 71 deletions
diff --git a/klm/lm/builder/print.cc b/klm/lm/builder/print.cc
index b0323221..84bd81ca 100644
--- a/klm/lm/builder/print.cc
+++ b/klm/lm/builder/print.cc
@@ -1,15 +1,11 @@
#include "lm/builder/print.hh"
-#include "util/double-conversion/double-conversion.h"
-#include "util/double-conversion/utils.h"
+#include "util/fake_ofstream.hh"
#include "util/file.hh"
#include "util/mmap.hh"
#include "util/scoped.hh"
#include "util/stream/timer.hh"
-#define BOOST_LEXICAL_CAST_ASSUME_C_LOCALE
-#include <boost/lexical_cast.hpp>
-
#include <sstream>
#include <string.h>
@@ -28,71 +24,6 @@ VocabReconstitute::VocabReconstitute(int fd) {
map_.push_back(i);
}
-namespace {
-class OutputManager {
- public:
- static const std::size_t kOutBuf = 1048576;
-
- // Does not take ownership of out.
- explicit OutputManager(int out)
- : buf_(util::MallocOrThrow(kOutBuf)),
- builder_(static_cast<char*>(buf_.get()), kOutBuf),
- // Mostly the default but with inf instead. And no flags.
- convert_(double_conversion::DoubleToStringConverter::NO_FLAGS, "inf", "NaN", 'e', -6, 21, 6, 0),
- fd_(out) {}
-
- ~OutputManager() {
- Flush();
- }
-
- OutputManager &operator<<(float value) {
- // Odd, but this is the largest number found in the comments.
- EnsureRemaining(double_conversion::DoubleToStringConverter::kMaxPrecisionDigits + 8);
- convert_.ToShortestSingle(value, &builder_);
- return *this;
- }
-
- OutputManager &operator<<(StringPiece str) {
- if (str.size() > kOutBuf) {
- Flush();
- util::WriteOrThrow(fd_, str.data(), str.size());
- } else {
- EnsureRemaining(str.size());
- builder_.AddSubstring(str.data(), str.size());
- }
- return *this;
- }
-
- // Inefficient!
- OutputManager &operator<<(unsigned val) {
- return *this << boost::lexical_cast<std::string>(val);
- }
-
- OutputManager &operator<<(char c) {
- EnsureRemaining(1);
- builder_.AddCharacter(c);
- return *this;
- }
-
- void Flush() {
- util::WriteOrThrow(fd_, buf_.get(), builder_.position());
- builder_.Reset();
- }
-
- private:
- void EnsureRemaining(std::size_t amount) {
- if (static_cast<std::size_t>(builder_.size() - builder_.position()) < amount) {
- Flush();
- }
- }
-
- util::scoped_malloc buf_;
- double_conversion::StringBuilder builder_;
- double_conversion::DoubleToStringConverter convert_;
- int fd_;
-};
-} // namespace
-
PrintARPA::PrintARPA(const VocabReconstitute &vocab, const std::vector<uint64_t> &counts, const HeaderInfo* header_info, int out_fd)
: vocab_(vocab), out_fd_(out_fd) {
std::stringstream stream;
@@ -112,8 +43,9 @@ PrintARPA::PrintARPA(const VocabReconstitute &vocab, const std::vector<uint64_t>
}
void PrintARPA::Run(const ChainPositions &positions) {
+ util::scoped_fd closer(out_fd_);
UTIL_TIMER("(%w s) Wrote ARPA file\n");
- OutputManager out(out_fd_);
+ util::FakeOFStream out(out_fd_);
for (unsigned order = 1; order <= positions.size(); ++order) {
out << "\\" << order << "-grams:" << '\n';
for (NGramStream stream(positions[order - 1]); stream; ++stream) {