From f8cc7173cae8e2913ca03d33a638f83ddfb84edd Mon Sep 17 00:00:00 2001 From: Siriusuna Date: Fri, 22 May 2026 02:44:50 +0800 Subject: [PATCH] refactor: extract utf8 truncation logic to utility Move the UTF-8 visual-width measurement and truncation helper functions from the mpris module to a common utility (include/util/utf8_string.hpp and src/util/utf8_string.cpp). This decouples string truncation and width measurement from the mpris module, allowing other modules (like mpd) to reuse. The helper function `utf8_truncate` (formerly `truncate` in mpris) and `utf8_width` is exported under the `waybar::util` namespace, while the low-level `measure_and_truncate` (formerly `utf8_truncate`) is encapsulated in an anonymous namespace in src/util/utf8_string.cpp to avoid unnecessary API exposure. No functional changes were made to the mpris module's behavior. --- include/util/utf8_string.hpp | 7 ++++ meson.build | 3 +- src/modules/mpris/mpris.cpp | 62 +++++------------------------------- src/util/utf8_string.cpp | 62 ++++++++++++++++++++++++++++++++++++ 4 files changed, 79 insertions(+), 55 deletions(-) create mode 100644 include/util/utf8_string.hpp create mode 100644 src/util/utf8_string.cpp diff --git a/include/util/utf8_string.hpp b/include/util/utf8_string.hpp new file mode 100644 index 00000000..07a04466 --- /dev/null +++ b/include/util/utf8_string.hpp @@ -0,0 +1,7 @@ +#pragma once +#include + +namespace waybar::util { +size_t utf8_width(const std::string& str); +void utf8_truncate(std::string& s, const std::string& ellipsis, size_t max_len); +} // namespace waybar::util diff --git a/meson.build b/meson.build index 0c494eb2..e528a6da 100644 --- a/meson.build +++ b/meson.build @@ -186,7 +186,8 @@ src_files = files( 'src/util/icon_loader.cpp', 'src/util/regex_collection.cpp', 'src/util/css_reload_helper.cpp', - 'src/util/transform_8bit_to_rgba.cpp' + 'src/util/transform_8bit_to_rgba.cpp', + 'src/util/utf8_string.cpp' ) man_files = files( diff --git a/src/modules/mpris/mpris.cpp b/src/modules/mpris/mpris.cpp index e0a27c92..6ebf34f7 100644 --- a/src/modules/mpris/mpris.cpp +++ b/src/modules/mpris/mpris.cpp @@ -7,6 +7,7 @@ #include #include "util/scope_guard.hpp" +#include "util/utf8_string.hpp" extern "C" { #include @@ -202,69 +203,22 @@ auto Mpris::getIconFromJson(const Json::Value& icons, const std::string& key) -> // Wide characters count as two, zero-width characters count as zero // Modifies str in-place (unless width = std::string::npos) // Returns the total width of the string pre-truncating -size_t utf8_truncate(std::string& str, size_t width = std::string::npos) { - if (str.length() == 0) return 0; - - const gchar* trunc_end = nullptr; - - size_t total_width = 0; - - for (gchar *data = str.data(), *end = data + str.size(); data != nullptr;) { - gunichar c = g_utf8_get_char_validated(data, end - data); - if (c == -1U || c == -2U) { - // invalid unicode, treat string as ascii - if (width != std::string::npos && str.length() > width) str.resize(width); - return str.length(); - } else if (g_unichar_iswide(c)) { - total_width += 2; - } else if (!g_unichar_iszerowidth(c) && c != 0xAD) { // neither zero-width nor soft hyphen - total_width += 1; - } - - data = g_utf8_find_next_char(data, end); - if (width != std::string::npos && total_width <= width && !g_unichar_isspace(c)) - trunc_end = data; - } - - if (trunc_end) str.resize(trunc_end - str.data()); - - return total_width; -} - -size_t utf8_width(const std::string& str) { return utf8_truncate(const_cast(str)); } - -void truncate(std::string& s, const std::string& ellipsis, size_t max_len) { - if (max_len == 0) { - s.resize(0); - return; - } - size_t len = utf8_truncate(s, max_len); - if (len > max_len) { - size_t ellipsis_len = utf8_width(ellipsis); - if (max_len >= ellipsis_len) { - if (ellipsis_len) utf8_truncate(s, max_len - ellipsis_len); - s += ellipsis; - } else { - s.resize(0); - } - } -} auto Mpris::getArtistStr(const PlayerInfo& info, bool truncated) -> std::string { auto artist = info.artist.value_or(std::string()); - if (truncated && artist_len_ >= 0) truncate(artist, ellipsis_, artist_len_); + if (truncated && artist_len_ >= 0) waybar::util::utf8_truncate(artist, ellipsis_, artist_len_); return artist; } auto Mpris::getAlbumStr(const PlayerInfo& info, bool truncated) -> std::string { auto album = info.album.value_or(std::string()); - if (truncated && album_len_ >= 0) truncate(album, ellipsis_, album_len_); + if (truncated && album_len_ >= 0) waybar::util::utf8_truncate(album, ellipsis_, album_len_); return album; } auto Mpris::getTitleStr(const PlayerInfo& info, bool truncated) -> std::string { auto title = info.title.value_or(std::string()); - if (truncated && title_len_ >= 0) truncate(title, ellipsis_, title_len_); + if (truncated && title_len_ >= 0) waybar::util::utf8_truncate(title, ellipsis_, title_len_); return title; } @@ -292,9 +246,9 @@ auto Mpris::getDynamicStr(const PlayerInfo& info, bool truncated, bool html) -> // keep position format same as length format auto position = getPositionStr(info, truncated && truncate_hours_ && length.length() < 6); - size_t artistLen = utf8_width(artist); - size_t albumLen = utf8_width(album); - size_t titleLen = utf8_width(title); + size_t artistLen = waybar::util::utf8_width(artist); + size_t albumLen = waybar::util::utf8_width(album); + size_t titleLen = waybar::util::utf8_width(title); size_t lengthLen = length.length(); size_t posLen = position.length(); @@ -313,7 +267,7 @@ auto Mpris::getDynamicStr(const PlayerInfo& info, bool truncated, bool html) -> // Since the first element doesn't present a separator and we don't know a priori which one // it will be, we add a "virtual separatorLen" to the dynamicLen, since we are adding the // separatorLen to all the other lengths. - size_t separatorLen = utf8_width(dynamic_separator_); + size_t separatorLen = waybar::util::utf8_width(dynamic_separator_); size_t dynamicLen = dynamic_len_ + separatorLen; if (showArtist) artistLen += separatorLen; if (showAlbum) albumLen += separatorLen; diff --git a/src/util/utf8_string.cpp b/src/util/utf8_string.cpp new file mode 100644 index 00000000..f96f2bd1 --- /dev/null +++ b/src/util/utf8_string.cpp @@ -0,0 +1,62 @@ +#include + +#include +#include + +namespace waybar::util { + +namespace { +// Wide characters count as two, zero-width characters count as zero +// Modifies str in-place (unless width = std::string::npos) +// Returns the total width of the string pre-truncating +size_t measure_and_truncate(std::string& str, size_t width = std::string::npos) { + if (str.length() == 0) return 0; + + const gchar* trunc_end = nullptr; + + size_t total_width = 0; + + for (gchar *data = str.data(), *end = data + str.size(); data != nullptr;) { + gunichar c = g_utf8_get_char_validated(data, end - data); + if (c == -1U || c == -2U) { + // invalid unicode, treat string as ascii + if (width != std::string::npos && str.length() > width) str.resize(width); + return str.length(); + } else if (g_unichar_iswide(c)) { + total_width += 2; + } else if (!g_unichar_iszerowidth(c) && c != 0xAD) { // neither zero-width nor soft hyphen + total_width += 1; + } + + data = g_utf8_find_next_char(data, end); + if (width != std::string::npos && total_width <= width && !g_unichar_isspace(c)) + trunc_end = data; + } + + if (trunc_end) str.resize(trunc_end - str.data()); + + return total_width; +} +} // namespace + +size_t utf8_width(const std::string& str) { + return measure_and_truncate(const_cast(str)); +} + +void utf8_truncate(std::string& s, const std::string& ellipsis, size_t max_len) { + if (max_len == 0) { + s.resize(0); + return; + } + size_t len = measure_and_truncate(s, max_len); + if (len > max_len) { + size_t ellipsis_len = utf8_width(ellipsis); + if (max_len >= ellipsis_len) { + if (ellipsis_len) measure_and_truncate(s, max_len - ellipsis_len); + s += ellipsis; + } else { + s.resize(0); + } + } +} +} // namespace waybar::util