X-Git-Url: https://git.carlh.net/gitweb/?a=blobdiff_plain;f=src%2Fsubrip_reader.cc;h=e5f113da1d294c17d31792ccbd79b103dc9b742f;hb=0a1b4c69c23d6e4838707fe74f3a55aa7b6c8649;hp=644a23cf877eff168b699fd378448e16a3ffb1d8;hpb=d704110228aae2bc7cd2ebc8ccd23d78be4fef9b;p=libsub.git

diff --git a/src/subrip_reader.cc b/src/subrip_reader.cc
index 644a23c..e5f113d 100644
--- a/src/subrip_reader.cc
+++ b/src/subrip_reader.cc
@@ -1,5 +1,5 @@
 /*
-    Copyright (C) 2014 Carl Hetherington <cth@carlh.net>
+    Copyright (C) 2014-2015 Carl Hetherington <cth@carlh.net>
 
     This program is free software; you can redistribute it and/or modify
     it under the terms of the GNU General Public License as published by
@@ -17,24 +17,48 @@
 
 */
 
+/** @file  src/subrip_reader.cc
+ *  @brief SubripReader class.
+ */
+
 #include "subrip_reader.h"
 #include "exceptions.h"
+#include "util.h"
 #include <boost/algorithm/string.hpp>
 #include <boost/lexical_cast.hpp>
 #include <boost/regex.hpp>
+#include <boost/bind.hpp>
 #include <cstdio>
 #include <vector>
+#include <iostream>
 
 using std::string;
 using std::vector;
 using std::list;
 using std::cout;
 using std::hex;
+using std::stringstream;
 using boost::lexical_cast;
 using boost::to_upper;
+using boost::optional;
+using boost::function;
 using namespace sub;
 
+/** @param s Subtitle string encoded in UTF-8 */
+SubripReader::SubripReader (string const & s)
+{
+	stringstream str (s);
+	this->read (boost::bind (&get_line_stringstream, &str));
+}
+
+/** @param f Subtitle file encoded in UTF-8 */
 SubripReader::SubripReader (FILE* f)
+{
+	this->read (boost::bind (&get_line_file, f));
+}
+
+void
+SubripReader::read (function<optional<string> ()> get_line)
 {
 	enum {
 		COUNTER,
@@ -42,55 +66,54 @@ SubripReader::SubripReader (FILE* f)
 		CONTENT
 	} state = COUNTER;
 
-	char buffer[256];
-
-	Time from;
-	Time to;
+	RawSubtitle rs;
 
-	string line;
-	int line_number = 0;
+	/* This reader extracts no information about where the subtitle
+	   should be on screen, so its reference is TOP_OF_SUBTITLE.
+	*/
+	rs.vertical_position.line = 0;
+	rs.vertical_position.reference = TOP_OF_SUBTITLE;
 
-	while (!feof (f)) {
-		char* r = fgets (buffer, sizeof (buffer), f);
-		if (r == 0 || feof (f)) {
+	while (true) {
+		optional<string> line = get_line ();
+		if (!line) {
 			break;
 		}
 
-		line = string (buffer);
-		trim_right_if (line, boost::is_any_of ("\n\r"));
-
-		if (
-			line.length() >= 3 &&
-			static_cast<unsigned char> (line[0]) == 0xef &&
-			static_cast<unsigned char> (line[1]) == 0xbb &&
-			static_cast<unsigned char> (line[2]) == 0xbf
-			) {
-
-			/* Skip Unicode byte order mark */
-			line = line.substr (3);
-		}
+		trim_right_if (*line, boost::is_any_of ("\n\r"));
+		remove_unicode_bom (line);
 
 		switch (state) {
 		case COUNTER:
 		{
-			if (line.empty ()) {
+			if (line->empty ()) {
 				/* a blank line at the start is ok */
 				break;
 			}
 
 			state = METADATA;
+
+			/* Reset stuff that should not persist across separate subtitles */
+			rs.bold = false;
+			rs.italic = false;
+			rs.underline = false;
+			rs.vertical_position.line = 0;
 		}
 		break;
 		case METADATA:
 		{
 			vector<string> p;
-			boost::algorithm::split (p, line, boost::algorithm::is_any_of (" "));
+
+			/* Further trim this line, removing spaces from the end */
+			trim_right_if (*line, boost::is_any_of (" "));
+
+			boost::algorithm::split (p, *line, boost::algorithm::is_any_of (" "));
 			if (p.size() != 3 && p.size() != 7) {
-				throw SubripError (line, "a time/position line");
+				throw SubripError (*line, "a time/position line");
 			}
 
-			from = convert_time (p[0]);
-			to = convert_time (p[2]);
+			rs.from = convert_time (p[0]);
+			rs.to = convert_time (p[2]);
 
 			/* XXX: should not ignore coordinate specifications */
 
@@ -98,12 +121,11 @@ SubripReader::SubripReader (FILE* f)
 			break;
 		}
 		case CONTENT:
-			if (line.empty ()) {
+			if (line->empty ()) {
 				state = COUNTER;
-				line_number = 0;
 			} else {
-				convert_line (line, line_number, from, to);
-				line_number++;
+				convert_line (*line, rs);
+				rs.vertical_position.line = rs.vertical_position.line.get() + 1;
 			}
 			break;
 		}
@@ -131,7 +153,7 @@ SubripReader::convert_time (string t)
 }
 
 void
-SubripReader::convert_line (string t, int line_number, Time from, Time to)
+SubripReader::convert_line (string t, RawSubtitle& p)
 {
 	enum {
 		TEXT,
@@ -140,16 +162,6 @@ SubripReader::convert_line (string t, int line_number, Time from, Time to)
 
 	string tag;
 
-	RawSubtitle p;
-	p.font = "Arial";
-	p.font_size.set_points (48);
-	p.from = from;
-	p.to = to;
-	p.vertical_position.line = line_number;
-	/* XXX: arbitrary */
-	p.vertical_position.lines = 32;
-	p.vertical_position.reference = TOP_OF_SUBTITLE;
-
 	list<Colour> colours;
 	colours.push_back (Colour (1, 1, 1));
 
@@ -212,6 +224,7 @@ SubripReader::convert_line (string t, int line_number, Time from, Time to)
 	maybe_content (p);
 }
 
+/* Push p into _subs if it has some text, and clear the text out of p */
 void
 SubripReader::maybe_content (RawSubtitle& p)
 {