Merge pull request #456 from siikamiika/parse-text-refactor

Parse text refactor

7 files changed, 148 insertions, 141 deletions

diff --git a/ext/bg/js/backend.js b/ext/bg/js/backend.js
index eb1f3e7b..2265c1a9 100644
--- a/ext/bg/js/backend.js
+++ b/ext/bg/js/backend.js
@@ -84,7 +84,6 @@ class Backend {
             ['kanjiFind', {handler: this._onApiKanjiFind.bind(this), async: true}],
             ['termsFind', {handler: this._onApiTermsFind.bind(this), async: true}],
             ['textParse', {handler: this._onApiTextParse.bind(this), async: true}],
-            ['textParseMecab', {handler: this._onApiTextParseMecab.bind(this), async: true}],
             ['definitionAdd', {handler: this._onApiDefinitionAdd.bind(this), async: true}],
             ['definitionsAddable', {handler: this._onApiDefinitionsAddable.bind(this), async: true}],
             ['noteView', {handler: this._onApiNoteView.bind(this), async: true}],
@@ -314,6 +313,60 @@ class Backend {
         return await this.dictionaryImporter.import(this.database, archiveSource, onProgress, details);
     }
 
+    async _textParseScanning(text, options) {
+        const results = [];
+        while (text.length > 0) {
+            const term = [];
+            const [definitions, sourceLength] = await this.translator.findTerms(
+                'simple',
+                text.substring(0, options.scanning.length),
+                {},
+                options
+            );
+            if (definitions.length > 0 && sourceLength > 0) {
+                dictTermsSort(definitions);
+                const {expression, reading} = definitions[0];
+                const source = text.substring(0, sourceLength);
+                for (const {text: text2, furigana} of jp.distributeFuriganaInflected(expression, reading, source)) {
+                    const reading2 = jp.convertReading(text2, furigana, options.parsing.readingMode);
+                    term.push({text: text2, reading: reading2});
+                }
+                text = text.substring(source.length);
+            } else {
+                const reading = jp.convertReading(text[0], '', options.parsing.readingMode);
+                term.push({text: text[0], reading});
+                text = text.substring(1);
+            }
+            results.push(term);
+        }
+        return results;
+    }
+
+    async _textParseMecab(text, options) {
+        const results = [];
+        const rawResults = await this.mecab.parseText(text);
+        for (const [mecabName, parsedLines] of Object.entries(rawResults)) {
+            const result = [];
+            for (const parsedLine of parsedLines) {
+                for (const {expression, reading, source} of parsedLine) {
+                    const term = [];
+                    for (const {text: text2, furigana} of jp.distributeFuriganaInflected(
+                        expression.length > 0 ? expression : source,
+                        jp.convertKatakanaToHiragana(reading),
+                        source
+                    )) {
+                        const reading2 = jp.convertReading(text2, furigana, options.parsing.readingMode);
+                        term.push({text: text2, reading: reading2});
+                    }
+                    result.push(term);
+                }
+                result.push([{text: '\n', reading: ''}]);
+            }
+            results.push([mecabName, result]);
+        }
+        return results;
+    }
+
     // Message handlers
 
     _onApiYomichanCoreReady(_params, sender) {
@@ -405,61 +458,27 @@ class Backend {
     async _onApiTextParse({text, optionsContext}) {
         const options = this.getOptions(optionsContext);
         const results = [];
-        while (text.length > 0) {
-            const term = [];
-            const [definitions, sourceLength] = await this.translator.findTerms(
-                'simple',
-                text.substring(0, options.scanning.length),
-                {},
-                options
-            );
-            if (definitions.length > 0) {
-                dictTermsSort(definitions);
-                const {expression, reading} = definitions[0];
-                const source = text.substring(0, sourceLength);
-                for (const {text: text2, furigana} of jp.distributeFuriganaInflected(expression, reading, source)) {
-                    const reading2 = jp.convertReading(text2, furigana, options.parsing.readingMode);
-                    term.push({text: text2, reading: reading2});
-                }
-                text = text.substring(source.length);
-            } else {
-                const reading = jp.convertReading(text[0], null, options.parsing.readingMode);
-                term.push({text: text[0], reading});
-                text = text.substring(1);
-            }
-            results.push(term);
+
+        if (options.parsing.enableScanningParser) {
+            results.push({
+                source: 'scanning-parser',
+                id: 'scan',
+                content: await this._textParseScanning(text, options)
+            });
         }
-        return results;
-    }
 
-    async _onApiTextParseMecab({text, optionsContext}) {
-        const options = this.getOptions(optionsContext);
-        const results = [];
-        const rawResults = await this.mecab.parseText(text);
-        for (const [mecabName, parsedLines] of Object.entries(rawResults)) {
-            const result = [];
-            for (const parsedLine of parsedLines) {
-                for (const {expression, reading, source} of parsedLine) {
-                    const term = [];
-                    if (expression !== null && reading !== null) {
-                        for (const {text: text2, furigana} of jp.distributeFuriganaInflected(
-                            expression,
-                            jp.convertKatakanaToHiragana(reading),
-                            source
-                        )) {
-                            const reading2 = jp.convertReading(text2, furigana, options.parsing.readingMode);
-                            term.push({text: text2, reading: reading2});
-                        }
-                    } else {
-                        const reading2 = jp.convertReading(source, null, options.parsing.readingMode);
-                        term.push({text: source, reading: reading2});
-                    }
-                    result.push(term);
-                }
-                result.push([{text: '\n'}]);
+        if (options.parsing.enableMecabParser) {
+            const mecabResults = await this._textParseMecab(text, options);
+            for (const [mecabDictName, mecabDictResults] of mecabResults) {
+                results.push({
+                    source: 'mecab',
+                    dictionary: mecabDictName,
+                    id: `mecab-${mecabDictName}`,
+                    content: mecabDictResults
+                });
             }
-            results.push([mecabName, result]);
         }
+
         return results;
     }
 
diff --git a/ext/bg/js/japanese.js b/ext/bg/js/japanese.js
index 5fef27a7..ac81acb5 100644
--- a/ext/bg/js/japanese.js
+++ b/ext/bg/js/japanese.js
@@ -124,25 +124,25 @@
         return wanakana.toRomaji(text);
     }
 
-    function convertReading(expressionFragment, readingFragment, readingMode) {
+    function convertReading(expression, reading, readingMode) {
         switch (readingMode) {
             case 'hiragana':
-                return convertKatakanaToHiragana(readingFragment || '');
+                return convertKatakanaToHiragana(reading);
             case 'katakana':
-                return convertHiraganaToKatakana(readingFragment || '');
+                return convertHiraganaToKatakana(reading);
             case 'romaji':
-                if (readingFragment) {
-                    return convertToRomaji(readingFragment);
+                if (reading) {
+                    return convertToRomaji(reading);
                 } else {
-                    if (isStringEntirelyKana(expressionFragment)) {
-                        return convertToRomaji(expressionFragment);
+                    if (isStringEntirelyKana(expression)) {
+                        return convertToRomaji(expression);
                     }
                 }
-                return readingFragment;
+                return reading;
             case 'none':
-                return null;
+                return '';
             default:
-                return readingFragment;
+                return reading;
         }
     }
 
@@ -300,7 +300,7 @@
                     const readingLeft = reading2.substring(group.text.length);
                     const segs = segmentize(readingLeft, groups.splice(1));
                     if (segs) {
-                        return [{text: group.text}].concat(segs);
+                        return [{text: group.text, furigana: ''}].concat(segs);
                     }
                 }
             } else {
@@ -368,7 +368,7 @@
         }
 
         if (stemLength !== source.length) {
-            output.push({text: source.substring(stemLength)});
+            output.push({text: source.substring(stemLength), furigana: ''});
         }
 
         return output;
diff --git a/ext/bg/js/mecab.js b/ext/bg/js/mecab.js
index cd6e6c57..597dceae 100644
--- a/ext/bg/js/mecab.js
+++ b/ext/bg/js/mecab.js
@@ -40,7 +40,36 @@ class Mecab {
     }
 
     async parseText(text) {
-        return await this.invoke('parse_text', {text});
+        const rawResults = await this.invoke('parse_text', {text});
+        // {
+        //     'mecab-name': [
+        //         // line1
+        //         [
+        //             {str expression: 'expression', str reading: 'reading', str source: 'source'},
+        //             {str expression: 'expression2', str reading: 'reading2', str source: 'source2'}
+        //         ],
+        //         line2,
+        //         ...
+        //     ],
+        //     'mecab-name2': [...]
+        // }
+        const results = {};
+        for (const [mecabName, parsedLines] of Object.entries(rawResults)) {
+            const result = [];
+            for (const parsedLine of parsedLines) {
+                const line = [];
+                for (const {expression, reading, source} of parsedLine) {
+                    line.push({
+                        expression: expression || '',
+                        reading: reading || '',
+                        source: source || ''
+                    });
+                }
+                result.push(line);
+            }
+            results[mecabName] = result;
+        }
+        return results;
     }
 
     startListener() {
diff --git a/ext/bg/js/search-query-parser-generator.js b/ext/bg/js/search-query-parser-generator.js
index 390841c1..9e7ff8aa 100644
--- a/ext/bg/js/search-query-parser-generator.js
+++ b/ext/bg/js/search-query-parser-generator.js
@@ -36,7 +36,7 @@ class QueryParserGenerator {
             const termContainer = this._templateHandler.instantiate(preview ? 'term-preview' : 'term');
             for (const segment of term) {
                 if (!segment.text.trim()) { continue; }
-                if (!segment.reading || !segment.reading.trim()) {
+                if (!segment.reading.trim()) {
                     termContainer.appendChild(this.createSegmentText(segment.text));
                 } else {
                     termContainer.appendChild(this.createSegment(segment));
@@ -71,7 +71,17 @@ class QueryParserGenerator {
         for (const parseResult of parseResults) {
             const optionContainer = this._templateHandler.instantiate('select-option');
             optionContainer.value = parseResult.id;
-            optionContainer.textContent = parseResult.name;
+            switch (parseResult.source) {
+                case 'scanning-parser':
+                    optionContainer.textContent = 'Scanning parser';
+                    break;
+                case 'mecab':
+                    optionContainer.textContent = `MeCab: ${parseResult.dictionary}`;
+                    break;
+                default:
+                    optionContainer.textContent = 'Unrecognized dictionary';
+                    break;
+            }
             optionContainer.defaultSelected = selectedParser === parseResult.id;
             selectContainer.appendChild(optionContainer);
         }
diff --git a/ext/bg/js/search-query-parser.js b/ext/bg/js/search-query-parser.js
index f90c68ee..eb3b681c 100644
--- a/ext/bg/js/search-query-parser.js
+++ b/ext/bg/js/search-query-parser.js
@@ -21,7 +21,6 @@
  * apiOptionsSet
  * apiTermsFind
  * apiTextParse
- * apiTextParseMecab
  * docSentenceExtract
  */
 
@@ -128,7 +127,7 @@ class QueryParser extends TextScanner {
 
         this.setPreview(text);
 
-        this.parseResults = await this.parseText(text);
+        this.parseResults = await apiTextParse(text, this.getOptionsContext());
         this.refreshSelectedParser();
 
         this.renderParserSelect();
@@ -137,33 +136,11 @@ class QueryParser extends TextScanner {
         this.setSpinnerVisible(false);
     }
 
-    async parseText(text) {
-        const results = [];
-        if (this.options.parsing.enableScanningParser) {
-            results.push({
-                name: 'Scanning parser',
-                id: 'scan',
-                parsedText: await apiTextParse(text, this.getOptionsContext())
-            });
-        }
-        if (this.options.parsing.enableMecabParser) {
-            const mecabResults = await apiTextParseMecab(text, this.getOptionsContext());
-            for (const [mecabDictName, mecabDictResults] of mecabResults) {
-                results.push({
-                    name: `MeCab: ${mecabDictName}`,
-                    id: `mecab-${mecabDictName}`,
-                    parsedText: mecabDictResults
-                });
-            }
-        }
-        return results;
-    }
-
     setPreview(text) {
         const previewTerms = [];
         for (let i = 0, ii = text.length; i < ii; i += 2) {
             const tempText = text.substring(i, i + 2);
-            previewTerms.push([{text: tempText}]);
+            previewTerms.push([{text: tempText, reading: ''}]);
         }
         this.queryParser.textContent = '';
         this.queryParser.appendChild(this.queryParserGenerator.createParseResult(previewTerms, true));
@@ -183,6 +160,6 @@ class QueryParser extends TextScanner {
         const parseResult = this.getParseResult();
         this.queryParser.textContent = '';
         if (!parseResult) { return; }
-        this.queryParser.appendChild(this.queryParserGenerator.createParseResult(parseResult.parsedText));
+        this.queryParser.appendChild(this.queryParserGenerator.createParseResult(parseResult.content));
     }
 }
diff --git a/ext/mixed/js/api.js b/ext/mixed/js/api.js
index 50b285a5..30c08347 100644
--- a/ext/mixed/js/api.js
+++ b/ext/mixed/js/api.js
@@ -44,10 +44,6 @@ function apiTextParse(text, optionsContext) {
     return _apiInvoke('textParse', {text, optionsContext});
 }
 
-function apiTextParseMecab(text, optionsContext) {
-    return _apiInvoke('textParseMecab', {text, optionsContext});
-}
-
 function apiKanjiFind(text, optionsContext) {
     return _apiInvoke('kanjiFind', {text, optionsContext});
 }
diff --git a/test/test-japanese.js b/test/test-japanese.js
index 89e41c36..87efdfad 100644
--- a/test/test-japanese.js
+++ b/test/test-japanese.js
@@ -176,19 +176,19 @@ function testConvertReading() {
         [['アリガトウ', 'アリガトウ', 'hiragana'], 'ありがとう'],
         [['アリガトウ', 'アリガトウ', 'katakana'], 'アリガトウ'],
         [['アリガトウ', 'アリガトウ', 'romaji'], 'arigatou'],
-        [['アリガトウ', 'アリガトウ', 'none'], null],
+        [['アリガトウ', 'アリガトウ', 'none'], ''],
         [['アリガトウ', 'アリガトウ', 'default'], 'アリガトウ'],
 
         [['ありがとう', 'ありがとう', 'hiragana'], 'ありがとう'],
         [['ありがとう', 'ありがとう', 'katakana'], 'アリガトウ'],
         [['ありがとう', 'ありがとう', 'romaji'], 'arigatou'],
-        [['ありがとう', 'ありがとう', 'none'], null],
+        [['ありがとう', 'ありがとう', 'none'], ''],
         [['ありがとう', 'ありがとう', 'default'], 'ありがとう'],
 
         [['有り難う', 'ありがとう', 'hiragana'], 'ありがとう'],
         [['有り難う', 'ありがとう', 'katakana'], 'アリガトウ'],
         [['有り難う', 'ありがとう', 'romaji'], 'arigatou'],
-        [['有り難う', 'ありがとう', 'none'], null],
+        [['有り難う', 'ありがとう', 'none'], ''],
         [['有り難う', 'ありがとう', 'default'], 'ありがとう'],
 
         // Cases with falsy readings
@@ -196,44 +196,20 @@ function testConvertReading() {
         [['ありがとう', '', 'hiragana'], ''],
         [['ありがとう', '', 'katakana'], ''],
         [['ありがとう', '', 'romaji'], 'arigatou'],
-        [['ありがとう', '', 'none'], null],
+        [['ありがとう', '', 'none'], ''],
         [['ありがとう', '', 'default'], ''],
 
-        [['ありがとう', null, 'hiragana'], ''],
-        [['ありがとう', null, 'katakana'], ''],
-        [['ありがとう', null, 'romaji'], 'arigatou'],
-        [['ありがとう', null, 'none'], null],
-        [['ありがとう', null, 'default'], null],
-
-        [['ありがとう', void 0, 'hiragana'], ''],
-        [['ありがとう', void 0, 'katakana'], ''],
-        [['ありがとう', void 0, 'romaji'], 'arigatou'],
-        [['ありがとう', void 0, 'none'], null],
-        [['ありがとう', void 0, 'default'], void 0],
-
         // Cases with falsy readings and kanji expressions
 
         [['有り難う', '', 'hiragana'], ''],
         [['有り難う', '', 'katakana'], ''],
         [['有り難う', '', 'romaji'], ''],
-        [['有り難う', '', 'none'], null],
-        [['有り難う', '', 'default'], ''],
-
-        [['有り難う', null, 'hiragana'], ''],
-        [['有り難う', null, 'katakana'], ''],
-        [['有り難う', null, 'romaji'], null],
-        [['有り難う', null, 'none'], null],
-        [['有り難う', null, 'default'], null],
-
-        [['有り難う', void 0, 'hiragana'], ''],
-        [['有り難う', void 0, 'katakana'], ''],
-        [['有り難う', void 0, 'romaji'], void 0],
-        [['有り難う', void 0, 'none'], null],
-        [['有り難う', void 0, 'default'], void 0]
+        [['有り難う', '', 'none'], ''],
+        [['有り難う', '', 'default'], '']
     ];
 
-    for (const [[expressionFragment, readingFragment, readingMode], expected] of data) {
-        assert.strictEqual(jp.convertReading(expressionFragment, readingFragment, readingMode), expected);
+    for (const [[expression, reading, readingMode], expected] of data) {
+        assert.strictEqual(jp.convertReading(expression, reading, readingMode), expected);
     }
 }
 
@@ -303,9 +279,9 @@ function testDistributeFurigana() {
             ['有り難う', 'ありがとう'],
             [
                 {text: '有', furigana: 'あ'},
-                {text: 'り'},
+                {text: 'り', furigana: ''},
                 {text: '難', furigana: 'がと'},
-                {text: 'う'}
+                {text: 'う', furigana: ''}
             ]
         ],
         [
@@ -317,23 +293,23 @@ function testDistributeFurigana() {
         [
             ['お祝い', 'おいわい'],
             [
-                {text: 'お'},
+                {text: 'お', furigana: ''},
                 {text: '祝', furigana: 'いわ'},
-                {text: 'い'}
+                {text: 'い', furigana: ''}
             ]
         ],
         [
             ['美味しい', 'おいしい'],
             [
                 {text: '美味', furigana: 'おい'},
-                {text: 'しい'}
+                {text: 'しい', furigana: ''}
             ]
         ],
         [
             ['食べ物', 'たべもの'],
             [
                 {text: '食', furigana: 'た'},
-                {text: 'べ'},
+                {text: 'べ', furigana: ''},
                 {text: '物', furigana: 'もの'}
             ]
         ],
@@ -341,9 +317,9 @@ function testDistributeFurigana() {
             ['試し切り', 'ためしぎり'],
             [
                 {text: '試', furigana: 'ため'},
-                {text: 'し'},
+                {text: 'し', furigana: ''},
                 {text: '切', furigana: 'ぎ'},
-                {text: 'り'}
+                {text: 'り', furigana: ''}
             ]
         ],
         // Ambiguous
@@ -373,16 +349,16 @@ function testDistributeFuriganaInflected() {
             ['美味しい', 'おいしい', '美味しかた'],
             [
                 {text: '美味', furigana: 'おい'},
-                {text: 'し'},
-                {text: 'かた'}
+                {text: 'し', furigana: ''},
+                {text: 'かた', furigana: ''}
             ]
         ],
         [
             ['食べる', 'たべる', '食べた'],
             [
                 {text: '食', furigana: 'た'},
-                {text: 'べ'},
-                {text: 'た'}
+                {text: 'べ', furigana: ''},
+                {text: 'た', furigana: ''}
             ]
         ]
     ];