feat(options): separatorCharsオプションの追加 (#27)

* feat(options): separatorCharsオプションの追加 * fix typo * add docs * test: add test case * chore: rename separatorCharacters - 1からアップデートしたときに意味が変わってしまうため * fix
textlint-ja · Dec 7, 2019 · ef6097c · ef6097c
1 parent 0c9d5ec
commit ef6097c
Show file tree

Hide file tree

Showing 5 changed files with 127 additions and 57 deletions.
diff --git a/README.md b/README.md
@@ -13,25 +13,14 @@
 
     npm install textlint-rule-no-doubled-joshi
 
-Require: textlint 5.0 >=
-
-Dependencies
-
-- [azu/kuromojin](https://github.com/azu/kuromojin) a wrapper of [kuromoji.js](https://github.com/takuyaa/kuromoji.js "kuromoji.js")
-- [azu/sentence-splitter](https://github.com/azu/sentence-splitter)
-
 ## Usage
 
 Via `.textlintrc`(推奨)
 
-```js
+```json5
 {
     "rules": {
-        "no-doubled-joshi": {
-            "min_interval" : 1,
-            "strict": false,
-            "allow": []
-        }
+        "no-doubled-joshi": true
     }
 }
 ```
@@ -47,27 +36,43 @@ textlint --rule no-doubled-joshi README.md
 
 `.textlintrc` options.
 
-```js
+```json5
 {
     "rules": {
         "no-doubled-joshi": {
-            // 助詞のtoken同士の距離が2以下ならエラー
+            // 助詞のtoken同士の距離が2以下ならエラーにする
             "min_interval" : 2,
             // 例外を許可するかどうか
             "strict": false,
             // 助詞のうち「も」「や」は複数回の出現を許す
-            "allow": ["も","や"]
+            "allow": ["も","や"],
+            // 文の区切り文字となる配列
+            "separatorCharacters": [
+                ".", // period
+                "．", // (ja) 全角period
+                "。", // (ja) 句点
+                "?", // question mark
+                "!", //  exclamation mark
+                "？", // (ja) 全角 question mark
+                "！" // (ja) 全角 exclamation mark
+            ]
         }
     }
 }
 ```
 
-- `min_interval`(default: 1) : 助詞の最低間隔値
-    - 指定した間隔値以下で同じ助詞が出現した場合エラーが出力されます。
-- `strict`(default: false) :例外もエラーとするかどうか
+- `min_interval`(default: `1`) : 助詞の最低間隔値
+    - 指定した間隔値以下で同じ助詞が出現した場合エラーが出力されます
+- `strict`(default: `false`) : 例外もエラーとするかどうか
     - 下記参照。例外としているものもエラーとするかどうか
-- `allow`(default: []) :複数回の出現を許す助詞
-    - 並立の助詞など、複数回出現しても無視する助詞を指定します。
+- `allow`(default: `[]`) :複数回の出現を許す助詞
+    - 並立の助詞など、複数回出現しても無視する助詞を指定します
+    - 例) `"も"`を許可したい場合は `{ "allow": ["も"] }`
+- `separatorCharacters`(default: `[".", "．", "。", "?", "!", "？", "！"]`) : 文の区切り文字として使用する文字の配列
+    - `separatorCharacters`を設定するとデフォルト値は上書きされます
+    - `。`のみを文の区切り文字にしたい場合は。`{ "separatorCharacters" : ["。"] }`のように指定します
+
+**min_interval**について
 
 > 私**は**彼**は**好き
 
@@ -140,7 +145,7 @@ NG: 文字列**には**そこ**には***問題がある。
 
 オプションで`"allow": ["も"]`を指定することで、**も**を例として扱うことができる。
 
-```js
+```json5
 {
     "rules": {
         "no-doubled-joshi": {
@@ -171,6 +176,11 @@ NG: 文字列**には**そこ**には***問題がある。
 - [作文入門](http://www.slideshare.net/takahi-i/ss-13429892 "作文入門")
 - [形態素解析ツールの品詞体系](http://www.unixuser.org/~euske/doc/postag/index.html#chasen "形態素解析ツールの品詞体系")
 
+## Related Libraries
+
+- [azu/kuromojin](https://github.com/azu/kuromojin) a wrapper of [kuromoji.js](https://github.com/takuyaa/kuromoji.js "kuromoji.js")
+- [azu/sentence-splitter](https://github.com/azu/sentence-splitter)
+
 ## Contributing
 
 1. Fork it!

diff --git a/package.json b/package.json
@@ -33,15 +33,15 @@
     "textlintrule"
   ],
   "devDependencies": {
-    "@textlint/types": "^1.2.2",
-    "@types/node": "^12.11.1",
+    "@textlint/types": "^1.2.3",
+    "@types/node": "^12.12.14",
     "textlint-scripts": "^3.0.0",
-    "ts-node": "^8.4.1",
-    "typescript": "^3.6.4"
+    "ts-node": "^8.5.4",
+    "typescript": "^3.7.3"
   },
   "dependencies": {
     "kuromojin": "^2.0.0",
-    "sentence-splitter": "^3.1.0",
+    "sentence-splitter": "^3.2.0",
     "textlint-rule-helper": "^2.1.1",
     "textlint-util-to-string": "^3.0.0"
   }

diff --git a/src/no-doubled-joshi.ts b/src/no-doubled-joshi.ts
@@ -58,7 +58,15 @@ const defaultOptions = {
     min_interval: 1,
     strict: false,
     allow: [],
-    separatorChars: ["。", "?", "!", "？", "！"]
+    separatorCharacters: [
+        ".", // period
+        "．", // (ja) zenkaku-period
+        "。", // (ja) 句点
+        "?", // question mark
+        "!", //  exclamation mark
+        "？", // (ja) zenkaku question mark
+        "！" // (ja) zenkaku exclamation mark
+    ]
 };
 
 
@@ -83,6 +91,7 @@ const report: TextlintRuleModule<Options> = function (context, options = {}) {
     const minInterval = options.min_interval || defaultOptions.min_interval;
     const isStrict = options.strict || defaultOptions.strict;
     const allow = options.allow || defaultOptions.allow;
+    const separatorCharacters = options.separatorCharacters || defaultOptions.separatorCharacters;
     const {Syntax, report, RuleError} = context;
     return {
         [Syntax.Paragraph](node) {
@@ -92,7 +101,11 @@ const report: TextlintRuleModule<Options> = function (context, options = {}) {
             const isSentenceNode = (node: TxtNode): node is SentenceNode => {
                 return node.type === SentenceSyntax.Sentence;
             };
-            const txtParentNode = splitSentences(node);
+            const txtParentNode = splitSentences(node, {
+                SeparatorParser: {
+                    separatorCharacters
+                }
+            });
             const sentences = txtParentNode.children.filter(isSentenceNode);
             return getTokenizer().then((tokenizer: any) => {
                 const checkSentence = (sentence: SentenceNode) => {

diff --git a/test/no-doubled-joshi-test.ts b/test/no-doubled-joshi-test.ts
@@ -25,11 +25,13 @@ tester.run("no-double-joshi", rule, {
             text: "太字も強調も同じように無視されます。",
             options: {allow: ["も"]}
         },
-        // 全角ピリオドを文区切り文字に含める
+        // 区切り文字をカスタムする
+        // ♪を区切り文字としたので、次の文は2つのセンテンスになる
         {
-            text: "画面に表示されます．それ以外には表示されません．",
-            options: {separatorChars: ["。","．","?","!","？","！"]}
+            text: "これはペンです♪これは鉛筆です♪",
+            options: {separatorCharacters: ["♪"]},
         }
+
     ],
     invalid: [
         // エラー位置は最後の助詞の位置を表示する
@@ -150,7 +152,7 @@ tester.run("no-double-joshi", rule, {
                     column: 6
                 }
             ]
-        },{
+        }, {
             text: `この行にはtextlintによる警告は出ない。
 この行にはtextlintにより警告が発せられる。この行に何かしようとすると起きるという
 この行にはtextlintによる警告は出ない。
@@ -190,6 +192,18 @@ tester.run("no-double-joshi", rule, {
                     index: 21
                 }
             ]
+        },
+        // オプションで、全角ピリオドを読点として認識させなくする
+        // 次のtextは1つのセンテンスとして認識されるので、"は"が重複する
+        {
+            text: "これはペンです．これは鉛筆です．",
+            options: {separatorCharacters: ["。"]},
+            errors: [
+                {
+                    message: `一文に二回以上利用されている助詞 "は" がみつかりました。`,
+                    index: 10
+                }
+            ]
         }
     ]
 });
diff --git a/yarn.lock b/yarn.lock
@@ -713,7 +713,7 @@
     lodash "^4.17.13"
     to-fast-properties "^2.0.0"
 
-"@textlint/ast-node-types@^4.0.2", "@textlint/ast-node-types@^4.2.1", "@textlint/ast-node-types@^4.2.4", "@textlint/ast-node-types@^4.2.5":
+"@textlint/ast-node-types@^4.2.1", "@textlint/ast-node-types@^4.2.4", "@textlint/ast-node-types@^4.2.5":
   version "4.2.5"
   resolved "https://registry.yarnpkg.com/@textlint/ast-node-types/-/ast-node-types-4.2.5.tgz#ae13981bc8711c98313a6ac1c361194d6bf2d39b"
   integrity sha512-+rEx4jLOeZpUcdvll7jEg/7hNbwYvHWFy4IGW/tk2JdbyB3SJVyIP6arAwzTH/sp/pO9jftfyZnRj4//sLbLvQ==
@@ -832,6 +832,13 @@
   dependencies:
     "@textlint/ast-node-types" "^4.2.5"
 
+"@textlint/types@^1.2.3":
+  version "1.2.3"
+  resolved "https://registry.yarnpkg.com/@textlint/types/-/types-1.2.3.tgz#6a27d6a2035873904d45c775b6205f162f169f91"
+  integrity sha512-48wXioiKIKAWPkgTWf/KyvZpjGH5L2ohmrSVgtx/u8q+KcY+q+/dslKy7QN0H+Yeu5Vj4GZRHOzCYWtzHMQleQ==
+  dependencies:
+    "@textlint/ast-node-types" "^4.2.5"
+
 "@textlint/utils@^1.0.3":
   version "1.0.3"
   resolved "https://registry.yarnpkg.com/@textlint/utils/-/utils-1.0.3.tgz#65196f21a5c1884b2dc6088484063f9ad52dfb67"
@@ -842,10 +849,10 @@
   resolved "https://registry.yarnpkg.com/@types/bluebird/-/bluebird-3.5.28.tgz#04c1a520ff076649236bc8ca21198542ce2bdb09"
   integrity sha512-0Vk/kqkukxPKSzP9c8WJgisgGDx5oZDbsLLWIP5t70yThO/YleE+GEm2S1GlRALTaack3O7U5OS5qEm7q2kciA==
 
-"@types/node@^12.11.1":
-  version "12.11.1"
-  resolved "https://registry.yarnpkg.com/@types/node/-/node-12.11.1.tgz#1fd7b821f798b7fa29f667a1be8f3442bb8922a3"
-  integrity sha512-TJtwsqZ39pqcljJpajeoofYRfeZ7/I/OMUQ5pR4q5wOKf2ocrUvBAZUMhWsOvKx3dVc/aaV5GluBivt0sWqA5A==
+"@types/node@^12.12.14":
+  version "12.12.14"
+  resolved "https://registry.yarnpkg.com/@types/node/-/node-12.12.14.tgz#1c1d6e3c75dba466e0326948d56e8bd72a1903d2"
+  integrity sha512-u/SJDyXwuihpwjXy7hOOghagLEV1KdAST6syfnOk6QZAMzZuWZqXy5aYYZbh8Jdpd4escVFP0MvftHNDb9pruA==
 
 "@types/structured-source@^3.0.0":
   version "3.0.0"
@@ -1279,7 +1286,7 @@ [email protected]:
   resolved "https://registry.yarnpkg.com/concat-map/-/concat-map-0.0.1.tgz#d8a96bd77fd68df7793a73036a3ba0d5405d477b"
   integrity sha1-2Klr13/Wjfd5OnMDajug1UBdR3s=
 
-concat-stream@^1.5.1, concat-stream@^1.5.2:
+concat-stream@^1.5.1:
   version "1.6.2"
   resolved "https://registry.yarnpkg.com/concat-stream/-/concat-stream-1.6.2.tgz#904bdf194cd3122fc675c77fc4ac3d4ff0fd1a34"
   integrity sha512-27HBghJxjiZtIk3Ycvn/4kbJk/1uZuJFfuPEns6LaEvpvG1f0hTea8lilrouyo9mVc2GWdcEZ8OLoGmSADlrCw==
@@ -1289,6 +1296,16 @@ concat-stream@^1.5.1, concat-stream@^1.5.2:
     readable-stream "^2.2.2"
     typedarray "^0.0.6"
 
+concat-stream@^2.0.0:
+  version "2.0.0"
+  resolved "https://registry.yarnpkg.com/concat-stream/-/concat-stream-2.0.0.tgz#414cf5af790a48c60ab9be4527d56d5e41133cb1"
+  integrity sha512-MWufYdFw53ccGjCA+Ol7XJYpAlW6/prSMzuPOTRnJGcGzuhLn4Scrz7qf6o8bROZ514ltazcIFJZevcfbo0x7A==
+  dependencies:
+    buffer-from "^1.0.0"
+    inherits "^2.0.3"
+    readable-stream "^3.0.2"
+    typedarray "^0.0.6"
+
 confirmer@^1.1.2:
   version "1.1.2"
   resolved "https://registry.yarnpkg.com/confirmer/-/confirmer-1.1.2.tgz#df36b3eb5ca5992750de8eea9db24781bb4cc254"
@@ -2695,7 +2712,7 @@ object.pick@^1.3.0:
   dependencies:
     isobject "^3.0.1"
 
-object.values@^1.0.4:
+object.values@^1.1.0:
   version "1.1.0"
   resolved "https://registry.yarnpkg.com/object.values/-/object.values-1.1.0.tgz#bf6810ef5da3e5325790eaaa2be213ea84624da9"
   integrity sha512-8mf0nKLAoFX6VlNVdhGj31SVYpaNFtUnuoOXWyFEstsWRgU837AK+JYM0iAxwkSzGRbwn8cbFmgbyxj1j4VbXg==
@@ -3034,6 +3051,15 @@ readable-stream@^2.0.2, readable-stream@^2.0.6, readable-stream@^2.2.2:
     string_decoder "~1.1.1"
     util-deprecate "~1.0.1"
 
+readable-stream@^3.0.2:
+  version "3.4.0"
+  resolved "https://registry.yarnpkg.com/readable-stream/-/readable-stream-3.4.0.tgz#a51c26754658e0a3c21dbf59163bd45ba6f447fc"
+  integrity sha512-jItXPLmrSR8jmTRmRWJXCnGJsfy85mB3Wd/uINMXA65yrnFo0cPClFIUWzo2najVNSl+mx7/4W8ttlLWJe99pQ==
+  dependencies:
+    inherits "^2.0.3"
+    string_decoder "^1.1.1"
+    util-deprecate "^1.0.1"
+
 readdirp@^2.2.1:
   version "2.2.1"
   resolved "https://registry.yarnpkg.com/readdirp/-/readdirp-2.2.1.tgz#0e87622a3325aa33e892285caf8b4e846529a525"
@@ -3225,7 +3251,7 @@ rimraf@^2.6.1:
   dependencies:
     glob "^7.1.3"
 
-safe-buffer@^5.1.2:
+safe-buffer@^5.1.2, safe-buffer@~5.2.0:
   version "5.2.0"
   resolved "https://registry.yarnpkg.com/safe-buffer/-/safe-buffer-5.2.0.tgz#b74daec49b1148f88c64b68d49b1e815c1f2f519"
   integrity sha512-fZEwUGbVl7kouZs1jCdMLdt95hdIv0ZeHg6L7qPeciMZhZ+/gdesW4wgTARkrFWEpspjEATAzUGPG8N2jJiwbg==
@@ -3262,14 +3288,14 @@ semver@^6.3.0:
   resolved "https://registry.yarnpkg.com/semver/-/semver-6.3.0.tgz#ee0a64c8af5e8ceea67687b133761e1becbd1d3d"
   integrity sha512-b39TBaTSfV6yBrapU89p5fKekE2m/NwnDocOVruQFS1/veMgdzuPcnOM34M6CwxW8jH/lxEa5rBoDeUwu5HHTw==
 
-sentence-splitter@^3.1.0:
-  version "3.1.0"
-  resolved "https://registry.yarnpkg.com/sentence-splitter/-/sentence-splitter-3.1.0.tgz#85f5a6c1912a9ddb40e48186ee960d0b834a250c"
-  integrity sha512-S8ye/Ic5ryKynwgjcS0ggncqhVJW3mgmo/gFNIcKKU1cdLjYzqtQ/PCkD7l0rZtSSfcjLqqVjfdMVN6FKkwUuQ==
+sentence-splitter@^3.2.0:
+  version "3.2.0"
+  resolved "https://registry.yarnpkg.com/sentence-splitter/-/sentence-splitter-3.2.0.tgz#fb2cd2f61f40006643ba83d9acf4609233c1c68c"
+  integrity sha512-lKX2tZ1rsA9Tu0gW8vRmMDmIEJoZ1d7cKpzcbFZdUrSpCR6gy/7OPPh7jjT/6Oc6Z79ToUmC2l8tyTEGanVmiA==
   dependencies:
-    "@textlint/ast-node-types" "^4.0.2"
-    concat-stream "^1.5.2"
-    object.values "^1.0.4"
+    "@textlint/ast-node-types" "^4.2.5"
+    concat-stream "^2.0.0"
+    object.values "^1.1.0"
     structured-source "^3.0.2"
 
 set-blocking@^2.0.0, set-blocking@~2.0.0:
@@ -3485,6 +3511,13 @@ string.prototype.trimright@^2.1.0:
     define-properties "^1.1.3"
     function-bind "^1.1.1"
 
+string_decoder@^1.1.1:
+  version "1.3.0"
+  resolved "https://registry.yarnpkg.com/string_decoder/-/string_decoder-1.3.0.tgz#42f114594a46cf1a8e30b0a84f56c78c3edac21e"
+  integrity sha512-hkRX8U1WjJFd8LsDJ2yQ/wWWxaopEsABU1XfkM8A+j0+85JAGppt16cr1Whg6KIbb4okU6Mql6BOj+uup/wKeA==
+  dependencies:
+    safe-buffer "~5.2.0"
+
 string_decoder@~1.1.1:
   version "1.1.1"
   resolved "https://registry.yarnpkg.com/string_decoder/-/string_decoder-1.1.1.tgz#9cf1611ba62685d7030ae9e4ba34149c3af03fc8"
@@ -3737,10 +3770,10 @@ try-resolve@^1.0.1:
   resolved "https://registry.yarnpkg.com/try-resolve/-/try-resolve-1.0.1.tgz#cfde6fabd72d63e5797cfaab873abbe8e700e912"
   integrity sha1-z95vq9ctY+V5fPqrhzq76OcA6RI=
 
-ts-node@^8.4.1:
-  version "8.4.1"
-  resolved "https://registry.yarnpkg.com/ts-node/-/ts-node-8.4.1.tgz#270b0dba16e8723c9fa4f9b4775d3810fd994b4f"
-  integrity sha512-5LpRN+mTiCs7lI5EtbXmF/HfMeCjzt7DH9CZwtkr6SywStrNQC723wG+aOWFiLNn7zT3kD/RnFqi3ZUfr4l5Qw==
+ts-node@^8.5.4:
+  version "8.5.4"
+  resolved "https://registry.yarnpkg.com/ts-node/-/ts-node-8.5.4.tgz#a152add11fa19c221d0b48962c210cf467262ab2"
+  integrity sha512-izbVCRV68EasEPQ8MSIGBNK9dc/4sYJJKYA+IarMQct1RtEot6Xp0bXuClsbUSnKpg50ho+aOAx8en5c+y4OFw==
   dependencies:
     arg "^4.1.0"
     diff "^4.0.1"
@@ -3760,10 +3793,10 @@ typedarray@^0.0.6:
   resolved "https://registry.yarnpkg.com/typedarray/-/typedarray-0.0.6.tgz#867ac74e3864187b1d3d47d996a78ec5c8830777"
   integrity sha1-hnrHTjhkGHsdPUfZlqeOxciDB3c=
 
-typescript@^3.6.4:
-  version "3.6.4"
-  resolved "https://registry.yarnpkg.com/typescript/-/typescript-3.6.4.tgz#b18752bb3792bc1a0281335f7f6ebf1bbfc5b91d"
-  integrity sha512-unoCll1+l+YK4i4F8f22TaNVPRHcD9PA3yCuZ8g5e0qGqlVlJ/8FSateOLLSagn+Yg5+ZwuPkL8LFUc0Jcvksg==
+typescript@^3.7.3:
+  version "3.7.3"
+  resolved "https://registry.yarnpkg.com/typescript/-/typescript-3.7.3.tgz#b36840668a16458a7025b9eabfad11b66ab85c69"
+  integrity sha512-Mcr/Qk7hXqFBXMN7p7Lusj1ktCBydylfQM/FZCk5glCNQJrCUKPkMHdo9R0MTFWsC/4kPFvDS0fDPvukfCkFsw==
 
 unherit@^1.0.4:
   version "1.1.2"
@@ -3895,7 +3928,7 @@ use@^3.1.0:
   resolved "https://registry.yarnpkg.com/use/-/use-3.1.1.tgz#d50c8cac79a19fbc20f2911f56eb973f4e10070f"
   integrity sha512-cwESVXlO3url9YWlFW/TA9cshCEhtu7IKJ/p5soJ/gGpj7vbvFrAY/eIioQ6Dw23KjZhYgiIo8HOs1nQ2vr/oQ==
 
-util-deprecate@~1.0.1:
+util-deprecate@^1.0.1, util-deprecate@~1.0.1:
   version "1.0.2"
   resolved "https://registry.yarnpkg.com/util-deprecate/-/util-deprecate-1.0.2.tgz#450d4dc9fa70de732762fbd2d4a28981419a0ccf"
   integrity sha1-RQ1Nyfpw3nMnYvvS1KKJgUGaDM8=