v1.5.1 - Fixed Activation ------------------------------------------------------------------------------- v1.5.0 Gloss: - Improved search results - Ignore empty results that are not Japanese - Exported words with empty readings will be substituted by its meaning - Added settings to better filter results, such as for subtitle files - Added Forvo and YouGlish to the External Search drop down menu in results Gloss & Kanji: - Added Google WaveNet Text-To-Speech API Settings: - Improved file and folder validation ------------------------------------------------------------------------------- v1.4.1 General: - Improved speed of scrollbars - Improved explanation of settings - Activation screen now accepts an Invoice ID Gloss: - Added Gloss heatmap - Improved Gloss search results - Removed settings to hide known words - Improved reliability of exporting terms and display of known terms - Added dictionary "PrimaryMeaning" and "PrimaryReading" CSV field names - Renamed dictionary "Value" field name to "SelectedMeaning" - Renamed dictionary "Source" field name to "SourceText" - Renamed dictionary "SourceFg" field name to "SourceTextWithReadings" - When the Determined reading is unknown, the PrimaryReading will be assumed Kanji: - Renamed kanji "Value" field name to "Kanji" Articles: - Added TechCrunch and Gendai news subscriptions ------------------------------------------------------------------------------- v1.4.0 General: - Added setting to control how a Source is saved to file - Added debug command message to console when app is loaded - Improved startup load and transition - Improved resolution scaling of Gloss and Kanji's top menu - Improved performance in various circumstances - Updated dictionary files and dependencies - Updated license - Changed in-app icons to adhere to updated license - Improved GUI descriptions - Scrolling a panel will no longer scroll the main window area Gloss: - Added Gloss & Word Lookup options for search - Removed ``re_m:`` and ``re_r:`` prefixes for regex searches - Improved Gloss' introduction and moved it to a help button window - Added option to ignore likely common particles - Improved hiding of hiragana terms by character count - Added option to ignore katakana by character count - Added display of known & starred entries for regex results - Fixed starred terms not appearing as starred when word is unknown - Fixed starred status not updating for entries that share a meaning with another entry (e.g. 人) - Improved performance and validation of word files when starring or saving terms - Added an option to ignore bracketed hiragana of a character count (e.g. to remove furigana: 食（た）べる) - Improved clipboard monitoring when window has been inactive for a long period - Improved duplicate removal for compounds found via lookahead - Improved lookahead matching & inflection detection removal - Fixed katakana terms of under 3 characters not being searched - Fixed performance problems when large texts are added to the search history menu - Underline the likely readings of words - Option to save sentences with furigana - Greatly increased Gloss speed, but now requires more RAM - Changed short and long text lookahead to just lookahead - Found katakana compound words will now be split - Hiragana and Kanji search accuracy is improved - Button to add all searched terms to the known words file - Readability will now count a word as known even if only one result is found as known - Particles and inflections will no longer be counted as words in the readability score, even if not ignored Articles: - Added ability to create subscription feeds in the Subscription window - Added support of RSS feeds - Added validation of subscription settings - Added Gloss' form of readability scoring for parity - Subscriptions will now preserve the previous ``enabled`` user setting when updating - Improved website localization compatibility (broadcast as Chrome and not Electron; include ``ja-JP`` locale) - Removed ``subfollow`` due to maintainability - Changed display of subscription information - Changed subscription fields - previous custom objects will need to be adapted to the new documentation - Improved furigana / ruby-tag removal for text detection - Simplified export as HTML only ------------------------------------------------------------------------------- v1.3.2 Articles: - Pages will now load with the `ja` language header, coercing websites to display in Japanese if it localizes based on browser language settings - Fixed subfollow not resetting its queue after a previous search - Fixed subfollow not prepending follow pages when host is empty - Improved URL matching - Improved load time of articles by blocking font resources - Added internal error handling for non-200 status codes, etc - Updated web blocking dependency to automatically whitelist article URLs (remote resources loaded from article pages will still be handled normally) ------------------------------------------------------------------------------- v1.3.1 General: - Debug information will now be logged to console once ``debug = true`` has been entered into the console Gloss: - Added a setting to limit clipboard text to a character count - Improved clipboard monitor handling - Improved katakana compound word detection with a supplemental method (e.g. スターバックスコーヒー) - Improved inflection word detection - Removed erroneous marks and English from Gloss results - Fixed Gloss from searching twice at once Articles: - Added `remove_irrelevant`, which is an optional str bool (`true`, `false`) to enable an internal html filter, to remove likely irrelevant text elements (e.g. headers, footers, links, menus...) and regardless of this option, this internal filter will be automatically applied when a body is not defined - Added an internal URL media-extension filter to remove non-text web pages (e.g. images, file types) - Changed display purpose of `id` to `name`, and `id` will remain an internal identifier - Improved the default text finding method, when none is defined by the user - Improved `ignore_urls` options and syntax, allowing either a comma-delimited string with an optional `ignore_urls_pos` value, or an object with a per-url `pos` value which acts as the `ignore_urls_pos` value - Improved display of article text preview (by removing chunks of whitespace) - Articles without Japanese will be removed - Fixed `wait_time` not being acknowledged - Improved Readme ------------------------------------------------------------------------------- v1.3.0 General: - Added single-column multi-word delimiter support: an option to separate imported word files' word columns by a delimiter for importing multiple words from one line or column - Minor UI tweaks (font size, alignments, scrollbar height scaling, context menu...) - Fixed dialog interactions - Improved Settings display - Updated Readme - Updated License - Updated dependencies Gloss and Articles: - Added a Starred Words file, similar to the "known words" file. Starred words can be used for a self-defined goal (e.g. recognizing Anki leeches). Shown in Gloss and Article results Gloss: - Added Clipboard Monitoring, to automatically Gloss copied text if it contains Japanese text, and focus the window afterward - Fixed unknown terms being matched incorrectly as known terms if first character is in a known term - Fixed not being able to match terms by kana reading - Fixed not being able to match terms by English meaning when search contains Japanese - Improved matching of known terms that contain both kana and kanji - Improved search and inflection detection - Improved removal of duplicate results - Improved removal of source when source is a conjugated verb/adjective and same as saved word Kanji: - Improved kana searches, with partial matching from the start of readings rather than exact matching (e.g. た.べる → た or たべ) Articles: - Added a Subscriptions window, and abandoned the wiki. Subscriptions can continue to be added and edited manually, and should be compatible with the previous object format - Added optional ``subfollow`` field, which finds text from pages linked from a ``follow`` page (e.g. ``/page/2``). Page links are matched at from the start, within, or end of a string via an optional ``subfollow_pos`` field, with the values of ``startswith``, ``includes``, ``endswidth`` and ``regex``. Sub-follows should be relative to the same host or page. - Added optional ``follow_pos`` field for the position of where the follow term exists. Same as ``subfollow_pos`` above. - Added optional ``id``, ``description``, and ``difficulty`` fields for subscription objects, and ``id`` and ``description`` will be displayed in Article results as a title - Added optional ``web_block`` field to enable or disable the internal web blocker for subscription objects (e.g. `"web_block": "false"`) (enabled by default if not present) - Improved ignoring of urls from the ``ignore_urls`` property, by internally disregarding starting http syntax, and ending slashes - [BREAKING] Added optional ``wait_time`` field, to wait for a number of milliseconds before processing text of a loaded page (e.g. for dynamic text). Previously, there was an internal wait time of 1200 milliseconds, which has now been reduced to 50. This internal change will break sites with dynamic texts, unless the ``wait_time`` field is used. This will otherwise speed up site processing. ------------------------------------------------------------------------------- v1.2.3 General: - Removed radio buttons from tabs - Updated dependencies Gloss: - Fixed Gloss' readability score (again) - Improved Gloss' lookahead detection - Improved Gloss' inflection detection for hiding - Improved Gloss' honorific term matching - Lowered Gloss' priority of name matches by not matching when using lookahead - Fixed Gloss' katakana terms displaying in hiragana - Saving Glossed words will now add all of that word's meanings to the Known Words output file - Changed Gloss' kanji pane to a togglable view - Gloss now matches names by regex search - Updated dictionary files Kanji: - Kanji tab highlights searched kanji in results as bold - Enlarged kanji font size - Added stroke order font for kanji, where a kanji's information is displayed - Fixed kanji not saving to output file Articles: - Fixed Articles' tab's tag & class removal from websites - Improved Articles' tab processing time with an ad blocker - Simplified Articles' tab explanation - Improved Articles' tab fuzzy matching ------------------------------------------------------------------------------- v1.2.2 - Fixed/Improved known word detection - Improved Gloss' inflection detection for hiding ------------------------------------------------------------------------------- v1.2.1 (Hotfix): - Changed Gloss to display *actual* interpreted searched term, which fixes the saving and detection of Source v1.2.0: - Removed erroneous dictionary data file(s) from installation - Fixed Articles tab processing count - Changed Articles tab to combine unknown hiragana terms found together to improve word detection. - Improved Articles furigana ruby tag removal () - Clarified Gloss and Articles Tab explanations - Improved import of Kanji tab CSS - Improved Gloss' stem word detection (should match stem prior to interpreted matching) (e.g. 禿げ(る)) - Improved Gloss' inflection detection for hiding (e.g. 遊ん(でる), さ(せる)) - Improved Gloss' lookahead to detect verb inflected expressions (e.g. 気にして → 気にする) - Improved Settings tab readability - Ctrl+number now switches between tabs ------------------------------------------------------------------------------- v1.1.9 - Improved Gloss' saved Source format - Improved Gloss' look-ahead detection - Gloss detects し-ending adjectives - Improved Gloss detection of み-ending verb nouns, and verbs - Improved Gloss' known word matching by removing punctuation (e.g. ～を動詞する → 動詞) - Prevented fields from gaining focus when alt and shift modifier keys are pressed - One-click copy of Kanji from Kanji results - Ctrl+Tab and Ctrl+Shift+Tab now switches between tabs - Removed backslash at start of Gloss definitions - Updated dictionary files - Updated dependencies ------------------------------------------------------------------------------- v1.1.8 - Fixed Gloss' searched reading not matching known word meaning - Fixed Gloss' search history menu searching terms when opened ------------------------------------------------------------------------------- v1.1.7 - Fixed glossed hiragana terms from erroneously matching known words (eg. はやる and は) - Fixed "0 to disable" options in settings from not disabling - Made ignoring of hiragana term length configurable in settings, for Gloss and Article - Fixed processing display disappearing if no Article limit - Fixed Gloss displaying duplicate empty verb entries - Changed Gloss' known word match from searched terms to found terms ------------------------------------------------------------------------------- v1.1.6: - removed internal ignored word list (eg. する, いる, ある) and refined results - show a more realistic calculation of article processing, if limits are applied - changed Article's ignored hiragana term length from 3 letters to 2 ------------------------------------------------------------------------------- v1.1.5: - Fixed Gloss' kanji pane displaying duplicate kanji - Removed erratic article timeout logic, and instead improved article load handling - Fixed "NaN" displayed when readability is 0% - Improved Kanji tab mouse highlight - Fixed gloss de-conjugation ------------------------------------------------------------------------------- v1.1.4: - Retry article pages if 'timed-out' - Gracefully continue if processing of an article fails - Remove erroneous meaning and source from saved dictionary terms if they do not exist - Fixed case sensitivity of Article category selection and display - Fixed Gloss displaying duplicate look-ahead terms - Improved Gloss verb de-conjugation ------------------------------------------------------------------------------- v1.1.3 - Updated dependencies - Improved window resizing - Clear cache on startup Hotfix: - Validated integer settings