*/
-#include <boost/algorithm/string.hpp>
#include "subrip.h"
-#include "subrip_content.h"
-#include "subrip_subtitle.h"
#include "cross.h"
#include "exceptions.h"
+#include "subrip_content.h"
+#include "data.h"
+#include <sub/subrip_reader.h>
+#include <sub/collect.h>
+#include <unicode/ucsdet.h>
+#include <unicode/ucnv.h>
+#include <iostream>
#include "i18n.h"
-using std::string;
-using std::list;
using std::vector;
using std::cout;
+using std::string;
using boost::shared_ptr;
-using boost::lexical_cast;
-using boost::algorithm::trim;
+using boost::scoped_array;
-SubRip::SubRip (shared_ptr<SubRipContent> content)
+SubRip::SubRip (shared_ptr<const SubRipContent> content)
{
- FILE* f = fopen_boost (content->path (0), "r");
- if (!f) {
- throw OpenFileError (content->path (0));
- }
-
- enum {
- COUNTER,
- METADATA,
- CONTENT
- } state = COUNTER;
-
- char buffer[256];
- int next_count = 1;
-
- boost::optional<SubRipSubtitle> current;
- list<string> lines;
-
- while (!feof (f)) {
- fgets (buffer, sizeof (buffer), f);
- if (feof (f)) {
- break;
- }
-
- string line (buffer);
- trim_right_if (line, boost::is_any_of ("\n\r"));
-
- switch (state) {
- case COUNTER:
- {
- int x = 0;
- try {
- x = lexical_cast<int> (line);
- } catch (...) {
-
- }
-
- if (x == next_count) {
- state = METADATA;
- ++next_count;
- current = SubRipSubtitle ();
- } else {
- throw SubRipError (line, _("a subtitle count"), content->path (0));
- }
- }
- break;
- case METADATA:
- {
- vector<string> p;
- boost::algorithm::split (p, line, boost::algorithm::is_any_of (" "));
- if (p.size() != 3 && p.size() != 7) {
- throw SubRipError (line, _("a time/position line"), content->path (0));
- }
-
- current->from = convert_time (p[0]);
- current->to = convert_time (p[2]);
-
- if (p.size() > 3) {
- current->x1 = convert_coordinate (p[3]);
- current->x2 = convert_coordinate (p[4]);
- current->y1 = convert_coordinate (p[5]);
- current->y2 = convert_coordinate (p[6]);
- }
- state = CONTENT;
- break;
- }
- case CONTENT:
- if (line.empty ()) {
- state = COUNTER;
- current->pieces = convert_content (lines);
- _subtitles.push_back (current.get ());
- current.reset ();
- lines.clear ();
- } else {
- lines.push_back (line);
- }
- break;
- }
- }
-
- if (state == CONTENT) {
- current->pieces = convert_content (lines);
- _subtitles.push_back (current.get ());
- }
-
- fclose (f);
-}
-
-Time
-SubRip::convert_time (string t)
-{
- Time r = 0;
-
- vector<string> a;
- boost::algorithm::split (a, t, boost::is_any_of (":"));
- assert (a.size() == 3);
- r += lexical_cast<int> (a[0]) * 60 * 60 * TIME_HZ;
- r += lexical_cast<int> (a[1]) * 60 * TIME_HZ;
-
- vector<string> b;
- boost::algorithm::split (b, a[2], boost::is_any_of (","));
- r += lexical_cast<int> (b[0]) * TIME_HZ;
- r += lexical_cast<int> (b[1]) * TIME_HZ / 1000;
-
- return r;
-}
-
-int
-SubRip::convert_coordinate (string t)
-{
- vector<string> a;
- boost::algorithm::split (a, t, boost::is_any_of (":"));
- assert (a.size() == 2);
- return lexical_cast<int> (a[1]);
-}
-
-void
-SubRip::maybe_content (list<SubRipSubtitlePiece>& pieces, SubRipSubtitlePiece& p)
-{
- if (!p.text.empty ()) {
- pieces.push_back (p);
- p.text.clear ();
- }
-}
-
-list<SubRipSubtitlePiece>
-SubRip::convert_content (list<string> t)
-{
- list<SubRipSubtitlePiece> pieces;
-
- SubRipSubtitlePiece p;
-
- enum {
- TEXT,
- TAG
- } state = TEXT;
-
- string tag;
-
- /* XXX: missing <font> support */
- /* XXX: nesting of tags e.g. <b>foo<i>bar<b>baz</b>fred</i>jim</b> might
- not work, I think.
- */
-
- for (list<string>::const_iterator i = t.begin(); i != t.end(); ++i) {
- for (size_t j = 0; j < i->size(); ++j) {
- switch (state) {
- case TEXT:
- if ((*i)[j] == '<' || (*i)[j] == '{') {
- state = TAG;
- } else {
- p.text += (*i)[j];
- }
- break;
- case TAG:
- if ((*i)[j] == '>' || (*i)[j] == '}') {
- if (tag == "b") {
- maybe_content (pieces, p);
- p.bold = true;
- } else if (tag == "/b") {
- maybe_content (pieces, p);
- p.bold = false;
- } else if (tag == "i") {
- maybe_content (pieces, p);
- p.italic = true;
- } else if (tag == "/i") {
- maybe_content (pieces, p);
- p.italic = false;
- } else if (tag == "u") {
- maybe_content (pieces, p);
- p.underline = true;
- } else if (tag == "/u") {
- maybe_content (pieces, p);
- p.underline = false;
- }
- tag.clear ();
- state = TEXT;
- } else {
- tag += (*i)[j];
- }
- break;
- }
- }
- }
-
- maybe_content (pieces, p);
-
- return pieces;
+ Data in (content->path (0));
+
+ UErrorCode status = U_ZERO_ERROR;
+ UCharsetDetector* detector = ucsdet_open (&status);
+ ucsdet_setText (detector, reinterpret_cast<const char *> (in.data().get()), in.size(), &status);
+
+ UCharsetMatch const * match = ucsdet_detect (detector, &status);
+ char const * in_charset = ucsdet_getName (match, &status);
+
+ UConverter* to_utf16 = ucnv_open (in_charset, &status);
+ /* This is a guess; I think we should be able to encode any input in 4 times its input size */
+ scoped_array<uint16_t> utf16 (new uint16_t[in.size() * 2]);
+ int const utf16_len = ucnv_toUChars (
+ to_utf16, reinterpret_cast<UChar*>(utf16.get()), in.size() * 2,
+ reinterpret_cast<const char *> (in.data().get()), in.size(),
+ &status
+ );
+
+ UConverter* to_utf8 = ucnv_open ("UTF-8", &status);
+ /* Another guess */
+ scoped_array<char> utf8 (new char[utf16_len * 2]);
+ ucnv_fromUChars (to_utf8, utf8.get(), utf16_len * 2, reinterpret_cast<UChar*>(utf16.get()), utf16_len, &status);
+
+ ucsdet_close (detector);
+ ucnv_close (to_utf16);
+ ucnv_close (to_utf8);
+
+ sub::SubripReader reader (utf8.get());
+ _subtitles = sub::collect<vector<sub::Subtitle> > (reader.subtitles ());
}
-Time
+ContentTime
SubRip::length () const
{
if (_subtitles.empty ()) {
- return 0;
+ return ContentTime ();
}
- return _subtitles.back().to;
+ return ContentTime::from_seconds (_subtitles.back().to.all_as_seconds ());
}
projects / dcpomatic.git / blobdiff