2 Copyright (C) 2014-2020 Carl Hetherington <cth@carlh.net>
4 This program is free software; you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation; either version 2 of the License, or
7 (at your option) any later version.
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
20 /** @file src/subrip_reader.cc
21 * @brief SubripReader class.
24 #include "subrip_reader.h"
25 #include "exceptions.h"
27 #include "sub_assert.h"
28 #include "raw_convert.h"
29 #include "ssa_reader.h"
30 #include <boost/algorithm/string.hpp>
31 #include <boost/lexical_cast.hpp>
32 #include <boost/regex.hpp>
33 #include <boost/bind.hpp>
43 using boost::lexical_cast;
44 using boost::to_upper;
45 using boost::optional;
46 using boost::function;
47 using boost::algorithm::replace_all;
50 /** @param s Subtitle string encoded in UTF-8 */
51 SubripReader::SubripReader (string s)
53 this->read (boost::bind(&get_line_string, &s));
56 /** @param f Subtitle file encoded in UTF-8 */
57 SubripReader::SubripReader (FILE* f)
59 this->read (boost::bind (&get_line_file, f));
63 SubripReader::read (function<optional<string> ()> get_line)
73 rs.vertical_position.line = 0;
74 rs.vertical_position.reference = TOP_OF_SUBTITLE;
77 optional<string> line = get_line ();
82 trim_right_if (*line, boost::is_any_of ("\n\r"));
83 remove_unicode_bom (line);
85 /* Keep some history in case there is an error to report */
86 _context.push_back (*line);
87 if (_context.size() > 5) {
88 _context.pop_front ();
95 /* a blank line at the start is ok */
101 /* Reset stuff that should not persist across separate subtitles */
104 rs.underline = false;
105 rs.vertical_position.line = 0;
106 rs.vertical_position.reference = TOP_OF_SUBTITLE;
113 /* Further trim this line, removing spaces from the end */
114 trim_right_if (*line, boost::is_any_of (" "));
116 boost::algorithm::split (p, *line, boost::algorithm::is_any_of (" "), boost::token_compress_on);
117 if (p.size() != 3 && p.size() != 7) {
118 for (int i = 0; i < 2; ++i) {
119 optional<string> ex = get_line ();
121 _context.push_back (*ex);
124 throw SubripError (*line, "a time/position line", _context);
127 rs.from = convert_time (p[0]);
128 rs.to = convert_time (p[2]);
130 /* XXX: should not ignore coordinate specifications */
136 if (line->empty ()) {
139 convert_line (*line, rs);
140 rs.vertical_position.line = rs.vertical_position.line.get() + 1;
148 SubripReader::convert_time (string t)
151 boost::algorithm::split (a, t, boost::is_any_of (":"));
153 throw SubripError (t, "time in the format h:m:s,ms", _context);
157 boost::algorithm::split (b, a[2], boost::is_any_of (","));
159 throw SubripError (t, "time in the format h:m:s,ms", _context);
165 h = lexical_cast<int>(a[0]);
166 } catch (boost::bad_lexical_cast &) {
167 throw SubripError (t, "integer hour value", _context);
171 m = lexical_cast<int>(a[1]);
172 } catch (boost::bad_lexical_cast &) {
173 throw SubripError (t, "integer minute value", _context);
177 s = lexical_cast<int>(b[0]);
178 } catch (boost::bad_lexical_cast &) {
179 throw SubripError (t, "integer second value", _context);
183 ms = lexical_cast<int>(b[1]);
184 } catch (boost::bad_lexical_cast &) {
185 throw SubripError (t, "integer millisecond value", _context);
188 return Time::from_hms (h, m, s, ms);
192 SubripReader::convert_line (string t, RawSubtitle& p)
201 list<Colour> colours;
202 colours.push_back (Colour (1, 1, 1));
204 /* XXX: missing <font> support */
205 /* XXX: nesting of tags e.g. <b>foo<i>bar<b>baz</b>fred</i>jim</b> might
209 for (size_t i = 0; i < t.size(); ++i) {
212 if (t[i] == '<' || t[i] == '{') {
219 if (t[i] == '>' || t[i] == '}') {
223 } else if (tag == "/b") {
226 } else if (tag == "i") {
229 } else if (tag == "/i") {
232 } else if (tag == "u") {
235 } else if (tag == "/u") {
238 } else if (boost::starts_with (tag, "font")) {
240 boost::regex re (".*color=\"?#([[:xdigit:]]+)\"?");
242 if (boost::regex_search (tag, match, re) && string (match[1]).size() == 6) {
243 p.colour = Colour::from_rgb_hex (match[1]);
244 colours.push_back (p.colour);
248 "[[:space:]]*([[:digit:]]+)[[:space:]]*,"
249 "[[:space:]]*([[:digit:]]+)[[:space:]]*,"
250 "[[:space:]]*([[:digit:]]+)[[:space:]]*,"
251 "[[:space:]]*([[:digit:]]+)[[:space:]]*"
254 if (boost::regex_search (tag, match, re) && match.size() == 5) {
255 p.colour.r = raw_convert<int>(string(match[1])) / 255.0;
256 p.colour.g = raw_convert<int>(string(match[2])) / 255.0;
257 p.colour.b = raw_convert<int>(string(match[3])) / 255.0;
258 colours.push_back (p.colour);
260 throw SubripError (tag, "a colour in the format #rrggbb or rgba(rr,gg,bb,aa)", _context);
263 } else if (tag == "/font") {
265 SUB_ASSERT (!colours.empty());
267 p.colour = colours.back ();
268 } else if (tag.size() > 0 && tag[0] == '\\') {
269 SSAReader::parse_style (p, tag, 288, 288);
274 tag += tolower (t[i]);
280 /* Strip Unicode U+202B (right-to-left embedding) as sometimes it is rendered
281 as a missing character. This may be a hack.
283 replace_all (p.text, "\xe2\x80\xab", "");
288 /* Push p into _subs if it has some text, and clear the text out of p */
290 SubripReader::maybe_content (RawSubtitle& p)
292 if (!p.text.empty ()) {