mirror of
https://github.com/Artikash/Textractor.git
synced 2024-12-23 08:54:12 +08:00
improve JSON parser
This commit is contained in:
parent
95b145bece
commit
2c9ac1da3c
@ -197,7 +197,7 @@ public:
|
||||
void AddSentence(QString sentence)
|
||||
{
|
||||
if (sentence.size() > maxSentenceSize) sentence = SENTENCE_TOO_BIG;
|
||||
if (!showOriginal && sentence.count(u8"\x200b \n")) sentence = sentence.split(u8"\x200b \n")[1];
|
||||
if (!showOriginal && sentence.contains(u8"\x200b \n")) sentence = sentence.split(u8"\x200b \n")[1];
|
||||
sanitize(sentence);
|
||||
sentence.chop(std::distance(std::remove(sentence.begin(), sentence.end(), QChar::Tabulation), sentence.end()));
|
||||
sentenceHistory.push_back(sentence);
|
||||
|
@ -61,4 +61,4 @@ std::string Escape(const std::string& text)
|
||||
return escaped;
|
||||
}
|
||||
|
||||
//TEST(assert(JSON::Parse(LR"([{"string":"hello world","boolean":false,"number": 1.67e+4,"null": null,"array":[]},"hello world"])")))
|
||||
TEST(assert(JSON::Parse<wchar_t>(LR"([{"string":"hello world","boolean":false,"number":1.67e+4,"null":null,"array":[]},"hello world"])")))
|
||||
|
@ -62,38 +62,6 @@ namespace JSON
|
||||
inline static std::wstring FromCodepoint(int codepoint) { return { (wchar_t)codepoint }; } // TODO: surrogate pairs
|
||||
};
|
||||
|
||||
template <typename C>
|
||||
std::pair<std::basic_string<C>, int> Unescape(std::basic_string_view<C> text)
|
||||
{
|
||||
std::basic_string<C> unescaped;
|
||||
int i = 0;
|
||||
for (; i < text.size(); ++i)
|
||||
{
|
||||
auto ch = text[i];
|
||||
if (ch == '"') return { unescaped, i + 1 };
|
||||
if (ch == '\\')
|
||||
{
|
||||
ch = text[i + 1];
|
||||
if (ch == 'u' && isxdigit(text[i + 2]) && isxdigit(text[i + 3]) && isxdigit(text[i + 4]) && isxdigit(text[i + 5]))
|
||||
{
|
||||
char charCode[] = { text[i + 2], text[i + 3], text[i + 4], text[i + 5], 0 };
|
||||
unescaped += UTF<C>::FromCodepoint(strtol(charCode, nullptr, 16));
|
||||
i += 5;
|
||||
continue;
|
||||
}
|
||||
for (auto [original, value] : Array<char, char>{ { 'b', '\b' }, {'f', '\f'}, {'n', '\n'}, {'r', '\r'}, {'t', '\t'} }) if (ch == original)
|
||||
{
|
||||
unescaped.push_back(value);
|
||||
goto replaced;
|
||||
}
|
||||
unescaped.push_back(ch);
|
||||
replaced: i += 1;
|
||||
}
|
||||
else unescaped.push_back(ch);
|
||||
}
|
||||
return { unescaped, i };
|
||||
}
|
||||
|
||||
template <typename C>
|
||||
struct Value : private std::variant<std::monostate, std::nullopt_t, bool, double, std::basic_string<C>, std::vector<Value<C>>, std::unordered_map<std::basic_string<C>, Value<C>>>
|
||||
{
|
||||
@ -102,6 +70,7 @@ namespace JSON
|
||||
explicit operator bool() const { return index(); }
|
||||
bool IsNull() const { return index() == 1; }
|
||||
auto Boolean() const { return std::get_if<bool>(this); }
|
||||
auto Number() const { return std::get_if<double>(this); }
|
||||
auto String() const { return std::get_if<std::basic_string<C>>(this); }
|
||||
auto Array() const { return std::get_if<std::vector<Value<C>>>(this); }
|
||||
auto Object() const { return std::get_if<std::unordered_map<std::basic_string<C>, Value<C>>>(this); }
|
||||
@ -121,7 +90,7 @@ namespace JSON
|
||||
};
|
||||
|
||||
template <typename C, int maxDepth = 25>
|
||||
Value<C> Parse(std::basic_string_view<C> text, int64_t& i, int depth)
|
||||
Value<C> Parse(const std::basic_string<C>& text, int64_t& i, int depth)
|
||||
{
|
||||
if (depth > maxDepth) return {};
|
||||
C ch;
|
||||
@ -134,10 +103,33 @@ namespace JSON
|
||||
};
|
||||
auto ExtractString = [&]
|
||||
{
|
||||
std::basic_string<C> unescaped;
|
||||
i += 1;
|
||||
auto [string, length] = Unescape(text.substr(i));
|
||||
i += length;
|
||||
return string;
|
||||
for (; i < text.size(); ++i)
|
||||
{
|
||||
auto ch = text[i];
|
||||
if (ch == '"') return i += 1, unescaped;
|
||||
if (ch == '\\')
|
||||
{
|
||||
ch = text[i + 1];
|
||||
if (ch == 'u' && isxdigit(text[i + 2]) && isxdigit(text[i + 3]) && isxdigit(text[i + 4]) && isxdigit(text[i + 5]))
|
||||
{
|
||||
char charCode[] = { text[i + 2], text[i + 3], text[i + 4], text[i + 5], 0 };
|
||||
unescaped += UTF<C>::FromCodepoint(strtol(charCode, nullptr, 16));
|
||||
i += 5;
|
||||
continue;
|
||||
}
|
||||
for (auto [original, value] : Array<char, char>{ { 'b', '\b' }, {'f', '\f'}, {'n', '\n'}, {'r', '\r'}, {'t', '\t'} }) if (ch == original)
|
||||
{
|
||||
unescaped.push_back(value);
|
||||
goto replaced;
|
||||
}
|
||||
unescaped.push_back(ch);
|
||||
replaced: i += 1;
|
||||
}
|
||||
else unescaped.push_back(ch);
|
||||
}
|
||||
return unescaped;
|
||||
};
|
||||
|
||||
if (SkipWhitespace()) return {};
|
||||
@ -155,9 +147,10 @@ namespace JSON
|
||||
|
||||
if (ch == '-' || (ch >= '0' && ch <= '9'))
|
||||
{
|
||||
// no numbers currently used, add an actual parser when needed
|
||||
while (i < text.size() && ((text[i] >= '0' && text[i] <= '9') || text[i] == '-' || text[i] == '+' || text[i] == 'e' || text[i] == 'E' || text[i] == '.')) ++i;
|
||||
return 0.0;
|
||||
std::string number;
|
||||
for (; i < text.size() && ((text[i] >= '0' && text[i] <= '9') || text[i] == '-' || text[i] == '+' || text[i] == 'e' || text[i] == 'E' || text[i] == '.'); ++i)
|
||||
number.push_back(text[i]);
|
||||
return strtod(number.c_str(), NULL);
|
||||
}
|
||||
|
||||
if (ch == '"') return ExtractString();
|
||||
@ -203,6 +196,6 @@ namespace JSON
|
||||
Value<C> Parse(const std::basic_string<C>& text)
|
||||
{
|
||||
int64_t start = 0;
|
||||
return Parse((std::basic_string_view<C>)text, start, 0);
|
||||
return Parse(text, start, 0);
|
||||
}
|
||||
}
|
||||
|
Loading…
x
Reference in New Issue
Block a user