Added a number of string-related tools.

CMakeLists.txt

@@ -99,6 +99,7 @@ set(SOURCE_FILES
 	src/audio_input/SampleRateConverter.cpp
 	src/audio_input/WaveFileReader.cpp
 	src/audio_input/wave_file_writing.cpp
+	src/string_tools.cpp
 )
 add_executable(rhubarb ${SOURCE_FILES})
 target_link_libraries(rhubarb ${Boost_LIBRARIES} cppFormat sphinxbase pocketSphinx)
@@ -106,7 +107,10 @@ target_compile_options(rhubarb PUBLIC ${enableWarningsFlags})
 
 # Define test project
 #include_directories("${gtest_SOURCE_DIR}/include")
-set(TEST_FILES "")
+set(TEST_FILES
+	tests/string_tools_tests.cpp
+	src/string_tools.cpp
+)
 add_executable(runTests ${TEST_FILES})
 target_link_libraries(runTests gtest gmock gmock_main)
 

src/string_tools.cpp

@@ -0,0 +1,78 @@
+#include "string_tools.h"
+#include <boost/algorithm/string/trim.hpp>
+
+using std::string;
+using std::vector;
+
+vector<string> splitIntoLines(const string& s) {
+	vector<string> lines;
+	auto it = s.cbegin();
+	auto lineBegin = it;
+	auto end = s.cend();
+	// Iterate over input string
+	while (it <= end) {
+		// Add a new result line when we hit a \n character or the end of the string
+		if (it == end || *it == '\n') {
+			string line(lineBegin, it);
+			// Trim \r characters
+			boost::algorithm::trim_if(line, [](char c) { return c == '\r'; });
+			lines.push_back(line);
+			lineBegin = it + 1;
+		}
+		++it;
+	}
+	
+	return lines;
+}
+
+vector<string> wrapSingleLineString(const string& s, int lineLength) {
+	if (lineLength <= 0) throw std::invalid_argument("lineLength must be > 0.");
+	if (s.find('\t') != std::string::npos) throw std::invalid_argument("s must not contain tabs.");
+	if (s.find('\n') != std::string::npos) throw std::invalid_argument("s must not contain line breaks.");
+
+	vector<string> lines;
+	auto it = s.cbegin();
+	auto lineBegin = it;
+	auto lineEnd = it;
+	auto end = s.cend();
+	// Iterate over input string
+	while (it <= end) {
+		// If we're at a word boundary: update safeLineEnd
+		if (it == end || *it == ' ') {
+			lineEnd = it;
+		}
+
+		// If we've hit lineLength or the end of the string: add a new result line
+		if (it == end || it - lineBegin == lineLength) {
+			if (lineEnd == lineBegin) {
+				// The line contains a single word, which is too long. Split mid-word.
+				lineEnd = it;
+			}
+
+			// Add trimmed line to list
+			string line(lineBegin, lineEnd);
+			boost::algorithm::trim_right(line);
+			lines.push_back(line);
+
+			// Resume after the last line, skipping spaces
+			it = lineEnd;
+			while (it != end && *it == ' ') ++it;
+			lineBegin = lineEnd = it;
+		}
+
+		++it;
+	}
+
+	return lines;
+}
+
+vector<string> wrapString(const string& s, int lineLength) {
+	vector<string> lines;
+	for (string paragraph : splitIntoLines(s)) {
+		auto paragraphLines = wrapSingleLineString(paragraph, lineLength);
+		copy(paragraphLines.cbegin(), paragraphLines.cend(), back_inserter(lines));
+	}
+
+	return lines;
+}
+

src/string_tools.h

@@ -0,0 +1,13 @@
+#ifndef RHUBARB_LIP_SYNC_STRING_TOOLS_H
+#define RHUBARB_LIP_SYNC_STRING_TOOLS_H
+
+#include <string>
+#include <vector>
+
+std::vector<std::string> splitIntoLines(const std::string& s);
+
+std::vector<std::string> wrapSingleLineString(const std::string& s, int lineLength);
+
+std::vector<std::string> wrapString(const std::string& s, int lineLength);
+
+#endif //RHUBARB_LIP_SYNC_STRING_TOOLS_H

tests/string_tools_tests.cpp

@@ -0,0 +1,73 @@
+#include <gmock/gmock.h>
+#include "string_tools.h"
+
+using namespace testing;
+
+// splitIntoLines
+
+TEST(splitIntoLines, splitsOnLineBreaks) {
+	EXPECT_THAT(splitIntoLines("this\nis\r\na\r\ntest"), ElementsAre("this", "is", "a", "test"));
+}
+
+TEST(splitIntoLines, handlesEmptyElements) {
+	EXPECT_THAT(splitIntoLines("\n1\n\n\n2\n"), ElementsAre("", "1", "", "", "2", ""));
+	EXPECT_THAT(splitIntoLines("\n"), ElementsAre("", ""));
+	EXPECT_THAT(splitIntoLines(""), ElementsAre(""));
+}
+
+// wrapSingleLineString
+
+TEST(wrapSingleLineString, basic) {
+	const char* lipsum = "Lorem ipsum dolor sit amet, consectetur adipisici elit, sed eiusmod tempor incidunt ut labore et dolore magna aliqua.";
+	EXPECT_THAT(wrapSingleLineString(lipsum, 30), ElementsAre("Lorem ipsum dolor sit amet,", "consectetur adipisici elit,", "sed eiusmod tempor incidunt ut", "labore et dolore magna aliqua."));
+}
+
+TEST(wrapSingleLineString, preciseWrapPosition) {
+	const char* test = "a b c";
+	EXPECT_THAT(wrapSingleLineString(test, 5), ElementsAre("a b c"));
+	EXPECT_THAT(wrapSingleLineString(test, 4), ElementsAre("a b", "c"));
+	EXPECT_THAT(wrapSingleLineString(test, 3), ElementsAre("a b", "c"));
+	EXPECT_THAT(wrapSingleLineString(test, 2), ElementsAre("a", "b", "c"));
+	EXPECT_THAT(wrapSingleLineString(test, 1), ElementsAre("a", "b", "c"));
+}
+
+TEST(wrapSingleLineString, overlongLines) {
+	EXPECT_THAT(wrapSingleLineString("aaa bbbb ccc", 3), ElementsAre("aaa", "bbb", "b", "ccc"));
+	EXPECT_THAT(wrapSingleLineString("aaa bbbb c", 3), ElementsAre("aaa", "bbb", "b c"));
+	EXPECT_THAT(wrapSingleLineString("a bbbb c", 5), ElementsAre("a", "bbbb", "c"));
+	EXPECT_THAT(wrapSingleLineString("aa b", 1), ElementsAre("a", "a", "b"));
+}
+
+TEST(wrapSingleLineString, discardingSpacesAtWrapPositionsAndEnd) {
+	const char* test = "  a  b  c  ";
+	EXPECT_THAT(wrapSingleLineString(test, 20), ElementsAre("  a  b  c"));
+	EXPECT_THAT(wrapSingleLineString(test, 6), ElementsAre("  a  b", "c"));
+	EXPECT_THAT(wrapSingleLineString(test, 5), ElementsAre("  a", "b  c"));
+	EXPECT_THAT(wrapSingleLineString(test, 4), ElementsAre("  a", "b  c"));
+	EXPECT_THAT(wrapSingleLineString(test, 3), ElementsAre("  a", "b", "c"));
+	EXPECT_THAT(wrapSingleLineString(test, 2), ElementsAre("", "a", "b", "c"));
+}
+
+TEST(wrapSingleLineString, errorHandling) {
+	EXPECT_NO_THROW(wrapSingleLineString("test", 1));
+	EXPECT_NO_THROW(wrapSingleLineString("", 1));
+
+	// Throw if lineLength < 1
+	EXPECT_ANY_THROW(wrapSingleLineString("test", 0));
+	EXPECT_ANY_THROW(wrapSingleLineString("", 0));
+	EXPECT_ANY_THROW(wrapSingleLineString("test", -1));
+	EXPECT_ANY_THROW(wrapSingleLineString("", -1));
+
+	// Throw if string contains tabs
+	EXPECT_ANY_THROW(wrapSingleLineString("a\tb", 10));
+
+	// Throw if string contains line breaks
+	EXPECT_ANY_THROW(wrapSingleLineString("a\nb", 10));
+}
+
+// wrapString
+
+TEST(wrapString, basic) {
+	EXPECT_THAT(wrapString("\n\nLine no 3\n\nLine no 4\n", 8), ElementsAre("", "", "Line no", "3", "", "Line no", "4", ""));
+}
+