1
0
Fork 0

Codechange: rework StringFilter to not need MallocT anymore

pull/13572/head
Rubidium 2025-02-15 22:23:56 +01:00 committed by rubidium42
parent ff6ae5b87b
commit 8b7460759a
3 changed files with 31 additions and 40 deletions

View File

@ -76,7 +76,9 @@ inline size_t ttd_strnlen(const char *str, size_t maxlen)
bool IsValidChar(char32_t key, CharSetFilter afilter); bool IsValidChar(char32_t key, CharSetFilter afilter);
size_t Utf8Decode(char32_t *c, const char *s); size_t Utf8Decode(char32_t *c, const char *s);
inline size_t Utf8Decode(char32_t *c, std::string::iterator &s) { return Utf8Decode(c, &*s); } /* std::string_view::iterator might be char *, in which case we do not want this templated variant to be taken. */
template <typename T> requires (!std::is_same_v<T, char *> && (std::is_same_v<std::string_view::iterator, T> || std::is_same_v<std::string::iterator, T>))
inline size_t Utf8Decode(char32_t *c, T &s) { return Utf8Decode(c, &*s); }
size_t Utf8Encode(char *buf, char32_t c); size_t Utf8Encode(char *buf, char32_t c);
size_t Utf8Encode(std::ostreambuf_iterator<char> &buf, char32_t c); size_t Utf8Encode(std::ostreambuf_iterator<char> &buf, char32_t c);
size_t Utf8Encode(std::back_insert_iterator<std::string> &buf, char32_t c); size_t Utf8Encode(std::back_insert_iterator<std::string> &buf, char32_t c);

View File

@ -25,34 +25,34 @@ static const char32_t STATE_QUOTE2 = '"';
* Set the term to filter on. * Set the term to filter on.
* @param str Filter term * @param str Filter term
*/ */
void StringFilter::SetFilterTerm(const char *str) void StringFilter::SetFilterTerm(std::string_view str)
{ {
this->word_index.clear(); this->word_index.clear();
this->word_index.shrink_to_fit(); this->word_index.shrink_to_fit();
this->word_matches = 0; this->word_matches = 0;
free(this->filter_buffer);
assert(str != nullptr);
char *dest = MallocT<char>(strlen(str) + 1);
this->filter_buffer = dest;
char32_t state = STATE_WHITESPACE; char32_t state = STATE_WHITESPACE;
const char *pos = str; auto pos = str.begin();
WordState *word = nullptr; auto word_begin = str.end();
size_t len; auto word_end = pos;
for (;; pos += len) {
/* Helper to prevent duplicating code. */
auto add_word = [&] () {
if (word_begin != str.end()) {
this->word_index.emplace_back(std::string(word_begin, word_end + 1), false);
word_begin = str.end();
}
};
for (size_t len; pos < str.end(); pos += len) {
char32_t c; char32_t c;
len = Utf8Decode(&c, pos); len = Utf8Decode(&c, pos);
if (c == 0 || (state == STATE_WORD && IsWhitespace(c))) { if (state == STATE_WORD && IsWhitespace(c)) {
/* Finish word */ /* Finish word */
if (word != nullptr) { add_word();
*(dest++) = '\0';
word = nullptr;
}
state = STATE_WHITESPACE; state = STATE_WHITESPACE;
if (c != 0) continue; else break; continue;
} }
if (state == STATE_WHITESPACE) { if (state == STATE_WHITESPACE) {
@ -74,22 +74,14 @@ void StringFilter::SetFilterTerm(const char *str)
} }
/* Add to word */ /* Add to word */
if (word == nullptr) { if (word_begin == str.end()) {
word = &this->word_index.emplace_back(WordState{ dest, false }); word_begin = pos;
} }
word_end = pos;
memcpy(dest, pos, len);
dest += len;
} }
}
/** /* Add the last word of the string. */
* Set the term to filter on. add_word();
* @param str Filter term
*/
void StringFilter::SetFilterTerm(const std::string &str)
{
this->SetFilterTerm(str.c_str());
} }
/** /**
@ -119,12 +111,12 @@ void StringFilter::AddLine(const char *str)
for (WordState &ws : this->word_index) { for (WordState &ws : this->word_index) {
if (!ws.match) { if (!ws.match) {
if (this->locale_aware) { if (this->locale_aware) {
if (match_case ? StrNaturalContains(str, ws.start) : StrNaturalContainsIgnoreCase(str, ws.start)) { if (match_case ? StrNaturalContains(str, ws.word) : StrNaturalContainsIgnoreCase(str, ws.word)) {
ws.match = true; ws.match = true;
this->word_matches++; this->word_matches++;
} }
} else { } else {
if ((match_case ? strstr(str, ws.start) : strcasestr(str, ws.start)) != nullptr) { if ((match_case ? strstr(str, ws.word.c_str()) : strcasestr(str, ws.word.c_str())) != nullptr) {
ws.match = true; ws.match = true;
this->word_matches++; this->word_matches++;
} }

View File

@ -31,13 +31,12 @@ struct StringFilter {
private: private:
/** State of a single filter word */ /** State of a single filter word */
struct WordState { struct WordState {
const char *start; ///< Word to filter for. std::string word; ///< Word to filter for.
bool match; ///< Already matched? bool match; ///< Already matched?
}; };
const char *filter_buffer; ///< Parsed filter string. Words separated by 0.
std::vector<WordState> word_index; ///< Word index and filter state. std::vector<WordState> word_index; ///< Word index and filter state.
uint word_matches; ///< Summary of filter state: Number of words matched. uint word_matches = 0; ///< Summary of filter state: Number of words matched.
const bool *case_sensitive; ///< Match case-sensitively (usually a static variable). const bool *case_sensitive; ///< Match case-sensitively (usually a static variable).
bool locale_aware; ///< Match words using the current locale. bool locale_aware; ///< Match words using the current locale.
@ -47,11 +46,9 @@ public:
* Constructor for filter. * Constructor for filter.
* @param case_sensitive Pointer to a (usually static) variable controlling the case-sensitivity. nullptr means always case-insensitive. * @param case_sensitive Pointer to a (usually static) variable controlling the case-sensitivity. nullptr means always case-insensitive.
*/ */
StringFilter(const bool *case_sensitive = nullptr, bool locale_aware = true) : filter_buffer(nullptr), word_matches(0), case_sensitive(case_sensitive), locale_aware(locale_aware) {} StringFilter(const bool *case_sensitive = nullptr, bool locale_aware = true) : case_sensitive(case_sensitive), locale_aware(locale_aware) {}
~StringFilter() { free(this->filter_buffer); }
void SetFilterTerm(const char *str); void SetFilterTerm(std::string_view str);
void SetFilterTerm(const std::string &str);
/** /**
* Check whether any filter words were entered. * Check whether any filter words were entered.