mirror of
https://github.com/Artikash/Textractor.git
synced 2024-12-23 08:54:12 +08:00
upgrade character deduplication algorithm
This commit is contained in:
parent
50f5b183f1
commit
b28b68d218
@ -5,30 +5,25 @@ bool ProcessSentence(std::wstring& sentence, SentenceInfo sentenceInfo)
|
|||||||
if (sentenceInfo["text number"] == 0) return false;
|
if (sentenceInfo["text number"] == 0) return false;
|
||||||
|
|
||||||
std::vector<int> repeatNumbers(sentence.size() + 1, 0);
|
std::vector<int> repeatNumbers(sentence.size() + 1, 0);
|
||||||
int repeatNumber = 1;
|
for (int i = 0; i < sentence.size(); ++i)
|
||||||
wchar_t prevChar = 0;
|
|
||||||
for (auto nextChar : sentence)
|
|
||||||
{
|
{
|
||||||
if (nextChar == prevChar)
|
if (sentence[i] != sentence[i + 1])
|
||||||
{
|
{
|
||||||
repeatNumber += 1;
|
int j = i;
|
||||||
}
|
while (sentence[j] == sentence[i] && --j >= 0);
|
||||||
else
|
repeatNumbers[i - j] += 1;
|
||||||
{
|
|
||||||
prevChar = nextChar;
|
|
||||||
repeatNumbers.at(repeatNumber) += 1;
|
|
||||||
repeatNumber = 1;
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
if ((repeatNumber = std::distance(repeatNumbers.begin(), std::max_element(repeatNumbers.begin(), repeatNumbers.end()))) == 1) return false;
|
int repeatNumber = std::distance(repeatNumbers.begin(), std::max_element(repeatNumbers.rbegin(), repeatNumbers.rend()).base() - 1);
|
||||||
|
if (repeatNumber < 2) return false;
|
||||||
|
|
||||||
std::wstring newSentence;
|
std::wstring newSentence;
|
||||||
for (int i = 0; i < sentence.size();)
|
for (int i = 0; i < sentence.size();)
|
||||||
{
|
{
|
||||||
newSentence.push_back(sentence.at(i));
|
newSentence.push_back(sentence[i]);
|
||||||
for (int j = i; j <= sentence.size(); ++j)
|
for (int j = i; j <= sentence.size(); ++j)
|
||||||
{
|
{
|
||||||
if (j == sentence.size() || sentence.at(i) != sentence.at(j))
|
if (j == sentence.size() || sentence[i] != sentence[j])
|
||||||
{
|
{
|
||||||
i += (j - i) % repeatNumber == 0 ? repeatNumber : 1;
|
i += (j - i) % repeatNumber == 0 ? repeatNumber : 1;
|
||||||
break;
|
break;
|
||||||
@ -44,8 +39,11 @@ TEST(
|
|||||||
InfoForExtension nonConsole[] = { { "text number", 1 }, {} };
|
InfoForExtension nonConsole[] = { { "text number", 1 }, {} };
|
||||||
|
|
||||||
std::wstring repeatedChars = L"aaaaaaaaaaaabbbbbbcccdddaabbbcccddd";
|
std::wstring repeatedChars = L"aaaaaaaaaaaabbbbbbcccdddaabbbcccddd";
|
||||||
|
std::wstring someRepeatedChars = L"abcdefaabbccddeeff";
|
||||||
ProcessSentence(repeatedChars, { nonConsole });
|
ProcessSentence(repeatedChars, { nonConsole });
|
||||||
|
ProcessSentence(someRepeatedChars, { nonConsole });
|
||||||
assert(repeatedChars.find(L"aaaabbcd") == 0);
|
assert(repeatedChars.find(L"aaaabbcd") == 0);
|
||||||
|
assert(someRepeatedChars == L"abcdefabcdef");
|
||||||
|
|
||||||
std::wstring empty = L"", one = L" ", normal = L"This is a normal sentence. はい";
|
std::wstring empty = L"", one = L" ", normal = L"This is a normal sentence. はい";
|
||||||
ProcessSentence(empty, { nonConsole });
|
ProcessSentence(empty, { nonConsole });
|
||||||
|
@ -30,7 +30,7 @@ bool ProcessSentence(std::wstring& sentence, SentenceInfo sentenceInfo)
|
|||||||
static std::mutex m;
|
static std::mutex m;
|
||||||
m.lock();
|
m.lock();
|
||||||
if (textNumber + 1 > cache.size()) cache.resize(textNumber + 1);
|
if (textNumber + 1 > cache.size()) cache.resize(textNumber + 1);
|
||||||
auto prevSentences = cache.at(textNumber).Acquire();
|
auto prevSentences = cache[textNumber].Acquire();
|
||||||
m.unlock();
|
m.unlock();
|
||||||
auto& inserted = prevSentences->emplace_back(sentence);
|
auto& inserted = prevSentences->emplace_back(sentence);
|
||||||
auto firstLocation = std::find(prevSentences->begin(), prevSentences->end(), sentence);
|
auto firstLocation = std::find(prevSentences->begin(), prevSentences->end(), sentence);
|
||||||
|
Loading…
x
Reference in New Issue
Block a user