city - npm Package Compare versions

Comparing version 0.1.1 to 0.1.2

index.js

@@ -15,7 +15,12 @@ #!/usr/bin/env node
 	.option('-s --overseas','包含海外地区')
+	.option('-y --pinyin','输出pinyin')
+	.option('-i --ignore [value]','忽略 \'省|市|区|地区|县\'')
 	.option('-c --code','包含地区编码（身份证前6位）')
-	.option('-u --update','重新抓取原始数据')
-	.option('-k --key [value]','对象键名 name,children,code')
-	// .option('-t --type','输出类型，array/object')
-	// .option('-f --flatten','扁平化输出')
+	.option('-u --update','重新抓取原始数据，或在citydata.json丢失的情况下使用')
+	.option('-k --key [value]','对象键名 name,children,code,pinyin，顺序不能改变，例如n,s,c,p')
+	.option('-f --flat [value]','扁平化输出，可以指定parentId参数名称')
+    .option('--closed','地区树默认为关闭状态')
+    .option('--father [value]','转出指定父节点的所有子节点，不含父节点')
+	.option('-r --reverse','转出以子节点为根的树，便于倒查父节点')
+    // .option('-t --type','输出类型，array/object')
 	// .option('-z --zipcode','包含邮编')
@@ -37,5 +42,11 @@ .parse(process.argv);
 		pretty:commander.pretty,
-		keys:commander.key
+		keys:commander.key,
+		pinyin:commander.pinyin,
+		ignore:commander.ignore,
+        flat:commander.flat,
+        closed:commander.closed,
+        father:commander.father,
+		reverse:commander.reverse
 	};
 	require('./lib/generator.js').generate(options);
 }

lib/generator.js

@@ -0,1 +1,11 @@
+var pinyin = require('pinyin');
+var util = require('util');
+
+var convertToPinyin = function(str) {
+	var pinyins = pinyin(str, { style: pinyin.STYLE_NORMAL });
+	return pinyins.map(function(word) {
+		return word[0];
+	}).join('_').toLowerCase();
+};
+
 exports.generate = function(options) {
@@ -7,32 +17,117 @@
 	if(!options.keys || typeof options.keys !== 'string' || !options.keys.length){
-		options.keys = ['name','children','code'];
+		options.keys = ['name','children','code','pinyin'];
 	}else{
 		options.keys = options.keys.split(',');
 	}
+	options.ignore = options.ignore || '';
+	options.ignoreRegexp = new RegExp(util.format('(%s)$', options.ignore), 'i');
 
+    var ret = null;
 
-	var fillRet = function(arr){
-		currLevel++;
-		var ret = arr.map(function(item){
-			if(item.name === '海外' && !options.overseas){
-				return null;
-			}
-			var tmpObj = {};
-			tmpObj[options.keys[0]] = item.name;
-			if(options.code && item.code){
-				tmpObj[options.keys[2]] = item.code;
-			}
-			if(level > currLevel && item.children){
-				tmpObj[options.keys[1]] = fillRet(item.children);
-			}
-			return tmpObj;
-		});
-		currLevel--;
-		return ret;
-	};
+    var findSon = function (arr) {
+        if(!options.father) return [];
+        for (var i=0; i<arr.length; ++i) {
+            var area = arr[i];
+            if(area.children) {
+                if (area.code == options.father) return area.children;
+                else {
+                    var result = findSon(area.children);
+                    if (result) return result;
+                }
+            }
+        }
+    };
 
-	var ret = fillRet(cityData).filter(function(item){
-		return item;
-	});
+    if(options.father) cityData = findSon(cityData, options.father);
 
+    if(options.reverse) {
+        ret = {};
+        var reverseData = function (arr, parent) {
+            arr = arr || [];
+            parent = parent || null;
+            if(parent) {
+                arr.map(function (item) {
+                    var parentData = {};
+                    parentData[options.keys[2]] = parent.code;
+                    parentData[options.keys[0]] = parent.name.replace(options.ignoreRegexp, '');
+
+                    ret[item.code] = parentData;
+                })
+            }
+            arr.map(function (item) {
+                if(item.children) {
+                    reverseData(item.children, item);
+                }
+            })
+        };
+        reverseData(cityData);
+    } else if(options.flat) {
+        ret = [];
+        var flatData = function (arr, parentCode) {
+            arr = arr || [];
+            parentCode = parentCode || 0;
+
+            arr.map(function (item) {
+                if (item.name === '海外' && !options.overseas) {
+                    return null;
+                }
+                var tmpObj = {};
+                tmpObj[options.keys[0]] = item.name.replace(options.ignoreRegexp, '');
+                if (options.code && item.code) {
+                    tmpObj[options.keys[2]] = item.code;
+                }
+                // 输出拼音
+                if (options.pinyin) {
+                    tmpObj[options.keys[3]] = convertToPinyin(tmpObj[options.keys[0]]);
+                }
+                if(parentCode > 0) {
+                    tmpObj[options.flat] = parentCode;
+                }
+                if(options.closed && item.children) {
+                    tmpObj['state'] = "closed";
+                }
+                ret.push(tmpObj);
+                if (item.children) {
+                    flatData(item.children, item.code);
+                }
+            });
+        };
+
+        flatData(cityData);
+    } else {
+        var fillRet = function (arr) {
+            currLevel++;
+            var ret = arr.map(function (item) {
+                if (item.name === '海外' && !options.overseas) {
+                    return null;
+                }
+                var tmpObj = {};
+                if(item.children && options.father && item.code == options.father) {
+
+                }
+                tmpObj[options.keys[0]] = item.name.replace(options.ignoreRegexp, '');
+                if (options.code && item.code) {
+                    tmpObj[options.keys[2]] = item.code;
+                }
+                // 输出拼音
+                if (options.pinyin) {
+                    tmpObj[options.keys[3]] = convertToPinyin(tmpObj[options.keys[0]]);
+                }
+                if(options.closed && item.children) {
+                    tmpObj['state'] = "closed";
+                }
+                if (level > currLevel && item.children) {
+                    tmpObj[options.keys[1]] = fillRet(item.children);
+                }
+                return tmpObj;
+            });
+            currLevel--;
+            return ret;
+        };
+
+        ret = fillRet(cityData).filter(function (item) {
+            return item;
+        });
+    }
+
 	if(options.amd){
@@ -47,3 +142,2 @@ output(ret,options.pretty,options.output,'amd');
 	}
-
 };
@@ -90,3 +184,2 @@
 	}
-
 }

lib/update.js

 var restler = require('restler');
 var cheerio = require('cheerio');
 
+var xzqh_url = 'http://www.mca.gov.cn/article/sj/xzqh/2020/2020/202003301019.html';
+var overseas_url = 'https://zh.wikipedia.org/zh/%E6%8C%89%E5%A4%A7%E6%B4%B2%E6%8E%92%E5%88%97%E7%9A%84%E5%9B%BD%E5%AE%B6%E5%88%97%E8%A1%A8';
 
-exports.update = function(){
-	var cityDataListUrl = 'http://www.stats.gov.cn/tjsj/tjbz/xzqhdm/';
-	getCityDataUrl(cityDataListUrl,function(err,cityDataUrl){
-		if(err){
-			console.log('loading city data list page failed.');
+exports.update = function () {
+
+	getCityRawData(xzqh_url, function(err, cityRawData) {
+		if (err) {
+			console.log('loading city raw data failed.');
 			console.log(err.message);
@@ -14,19 +16,9 @@ return;
 
-		getCityRawData(cityDataUrl,function(err,cityRawData){
-			if(err){
-				console.log('loading city raw data failed.');
+		var cityData = parseCityData(cityRawData);
+
+		getOverseasRawData(overseas_url, function (err,overseasRawData) {
+			if (err) {
+				console.log('loading overseas raw data failed.');
 				console.log(err.message);
-				return;
-			}
-
-			var cityData = parseCityData(cityRawData);
-
-			var overseasDataListUrl = 'http://zh.wikipedia.org/zh/%E5%9B%BD%E5%AE%B6%E5%88%97%E8%A1%A8_(%E6%8C%89%E6%B4%B2%E6%8E%92%E5%88%97)';
-			getOverseasRawData(overseasDataListUrl,function(err,overseasRawData){
-				if(err){
-					console.log('loading overseas raw data failed.');
-					console.log(err.message);
-					return;
-				}
-
+			} else {
 				console.log('mixing overseas data to city data...');
@@ -41,10 +33,9 @@ var overseasObj = {name:'海外'};
 				console.log('mixing overseas data to city data successed.');
+			}
 
-				console.log('writing data file...');
-				var fs = require('fs');
-				fs.writeFileSync('./lib/citydata.json',JSON.stringify(cityData));
-				console.log('writing data file successed.');
-			});
+			console.log('writing data file...');
+			var fs = require('fs');
+			fs.writeFileSync('./lib/citydata.json',JSON.stringify(cityData));
+			console.log('writing data file successed.');
 		});
-
 	});
@@ -54,104 +45,132 @@
 
+function getCityRawData (url, callback) {
+	console.log('loading city raw data page...');
+	restler.get(url)
+		.on('success',function(data){
+			console.log('loading successed.');
+			console.log('parsing to get city raw data.');
+			var $ = cheerio.load(data);
 
-function getCityDataUrl(cityDataListUrl,callback){
-	console.log('loading citydata list page...');
-	restler.get(cityDataListUrl).on('success',function(data){
-		console.log('loading successed.');
-		console.log('parsing newest city data page url.');
-		var $ = cheerio.load(data);
-		var cityDataUrl = $('.center_list_contlist a').eq(0).attr('href');
-		var url = require('url');
-		console.log('parsing successed, got newest citydata url:'+cityDataUrl);
-		callback(null,url.resolve(cityDataListUrl,cityDataUrl));
-	}).on('fail',function(){
-		callback(new Error('fail'));
-	}).on('error',callback);
-}
+			var $validRows = $('tr').filter(function (_, tr) {
+				return $(tr).attr('height') === '19'
+			});
 
-function getCityRawData(url,callback){
-	console.log('loading city raw data page...');
-	restler.get(url).on('success',function(data){
-		console.log('loading successed.');
-		console.log('parsing to get city raw data.');
-		var $ = cheerio.load(data);
-		var cityRawData = $('.TRS_Editor').last().find('style').remove().end().text();
-		console.log('parsing successed.');
-		callback(null,cityRawData);
-	}).on('fail',function(){
-		callback(new Error('fail'));
-	}).on('error',callback);
+			var rowsContent = [];
+
+			$validRows.each(function (_, tr) {
+				var $validTds = $(tr).find('td').filter(function (_, td) {
+					return $(td).html().trim() !== '';
+				});
+				var colsContent = [];
+				$validTds.each(function (_, td) {
+					colsContent.push($(td).text());
+				});
+				rowsContent.push(colsContent);
+			});
+
+			callback(null, rowsContent);
+		}).on('fail',function () {
+			callback(new Error('fail'));
+		}).on('error',callback);
 }
 
-function parseCityData(rawData){
+function parseCityData (rawData) {
 	console.log('start parsing city data');
-	var cityRegExp = /(\d{6})\s+([\D]+)/g;
-	var matchResult = cityRegExp.exec(rawData);
-	var cityArr = [];
-	while(matchResult){
-		cityArr.push({
-			code:matchResult[1],
-			name:matchResult[2].trim()
+
+	var provinceIndexes = rawData.map(function (row, index) {
+		if (!row[1].match(/\s/)) {
+			return index;
+		} else {
+			return -1;
+		}
+	}).filter(function (idx) {
+		return idx > -1;
+	})
+
+	var groups = provinceIndexes.map(function (idx, index) {
+		var topLevel = {
+			code: rawData[idx][0].trim(),
+			name: rawData[idx][1].trim(),
+			children: []
+		};
+
+		var children = rawData.slice(idx + 1, provinceIndexes[index + 1]);
+		var hasCity = children.some(function (row) {
+			return row[1].length - row[1].trim().length === 1;
 		});
-		matchResult = cityRegExp.exec(rawData);
-	}
-	var cityData = [];
-	var currPrefix = '';
-	var currItem = [];
 
-	cityArr.forEach(function(cityItem){
-		var thisPrefix = cityItem.code.replace(/0*$/,'');
-		var thisLevel = Math.ceil(thisPrefix.length / 2);
+		if (hasCity) {
+			var cityIndexes = children.map(function (row, index) {
+				if (row[1].length - row[1].trim().length === 1) {
+					return index;
+				} else {
+					return -1;
+				}
+			}).filter(function (idx) {
+				return idx > -1;
+			});
 
-		// 直辖市去掉二级（市辖区、县），直接把三级提升
-		var skipCodeArr = ['1101','1102','1201','1202','3101','3102','5001','5002'];
-		if(skipCodeArr.indexOf(thisPrefix) > -1){
-			return;
-		}else if(thisLevel === 3){
-			var shouldLevelUp = false;
-			if(skipCodeArr.some(function(codePrefix){
-				return thisPrefix.indexOf(codePrefix) === 0;
-			})){
-				thisLevel--;
-			}
-		}
+			var cityGroups = cityIndexes.map(function (idx, index) {
+				var cityLevel = {
+					code: children[idx][0].trim(),
+					name: children[idx][1].trim(),
+					children: []
+				};
 
-		if(thisLevel === 1){
-			cityData.push(cityItem);
-			currItem[0] = cityItem;
-		}else{
-			var parent = currItem[thisLevel - 2];
-			if(!parent.children){
-				parent.children = [];
-			}
-			parent.children.push(cityItem);
-			currItem[thisLevel - 1] = cityItem;
+				var countyList = children.slice(idx + 1, cityIndexes[index + 1]).map(function (row) {
+					return {
+						code: row[0].trim(),
+						name: row[1].trim()
+					}
+				});
+
+				cityLevel.children = countyList.slice();
+				return cityLevel;
+			});
+
+			topLevel.children = cityGroups.slice();
+		} else { // 直辖市
+			var _children = children.map(function (child) {
+				return {
+					code: child[0].trim(),
+					name: child[1].trim()
+				}
+			});
+			topLevel.children = _children.slice();
 		}
+		return topLevel;
 	});
+
 	console.log('parsing successed.');
-	return cityData;
+	return groups;
 }
 
-function getOverseasRawData(url,callback){
+function getOverseasRawData (url, callback) {
 	console.log('loading overseas raw data page...');
-	restler.get(url).on('success',function(data){
-		console.log('loading successed.');
-		console.log('parsing to get overseas raw data.');
-		var $ = cheerio.load(data);
-		var overseasRawData = $('.wikitable tr td b a').map(function(){
-			return $(this).text().trim();
-		}).get().filter(function(item){
-			var filterArr = [
-				'中华人民共和国',
-				'香港',
-				'澳门'
-			];
-			return item && filterArr.indexOf(item) === -1;
-		}).sort(function(item1,item2){
-			// todo:按拼音排序
-		});
-		console.log('parsing successed.');
-		callback(null,overseasRawData);
-	}).on('fail',function(){
-		callback(new Error('fail'));
-	}).on('error',callback);
+	restler.get(url)
+		.on('success', function (data) {
+			console.log('loading successed.');
+			console.log('parsing to get overseas raw data.');
+			var $ = cheerio.load(data);
+			var overseasRawData = $('.wikitable tr td b a').map(function () {
+				return $(this).text().trim();
+			}).get().filter(function (item) {
+				var filterArr = [
+					'中华人民共和国',
+					'香港',
+					'澳门',
+					'澳門',
+					'中华民国',
+					'中華民國'
+				];
+				return item && filterArr.indexOf(item) === -1;
+			});
+			// @todo: 拼音支持
+			console.log('parsing successed.');
+			callback(null, overseasRawData);
+		})
+		.on('fail', function () {
+			callback(new Error('fail'));
+		})
+		.on('error', callback);
 }

package.json

 {
   "name": "city",
   "description": "中国国家行政区划数据",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "author": "basecss <[e-mail]>",
@@ -16,2 +16,3 @@ "contributors": [
     "commander": "~2.2.0",
+    "pinyin": "^2.6.2",
     "restler": "^3.2.2"
@@ -18,0 +19,0 @@ },

README.md

@@ -17,4 +17,4 @@ city
     "children":[{
-        "name":"和平区"    
-    }]    
+        "name":"和平区"
+    }]
 }]
@@ -32,2 +32,4 @@ ```
 - 支持输出地区代码（身份证前6位）
+- 支持拼音输出
+- 支持自定义去除 省|市|区|地区|县 等后缀
 
@@ -62,3 +64,26 @@ ### 使用方法
 - `-u` `--update` 重新抓取原始数据
+- `-y` `--pinyin` 输出pinyin
+- `-i` `--ignore '省|市|区|地区|县'` 忽略 省|市|区|地区|县
+- `-k` `--key name,children,code,pinyin` 导出的键名（顺序不能改变，例如n,s,c,p）
+- `-f` `--flat 'parentId'` 扁平化输出，可以指定parentId参数名称
+-      `--closed` 地区树结点默认为关闭状态
+-      `--father` 转出指定父节点的所有子节点，不含父节点
+- `-r` `--reserve` 转出以子节点为根的树，便于倒查父节点
 
+示例：
+1. 嵌套方式列出所有地区，name改为n，children改为s，code改为c，并带上地区编码
+```sh
+city -k n,s,c -o ./area.json -c
+```
+
+2. 平面方式列出所有地区，name改为n，children改为s，code改为c，parentId为p，并带上地区编码
+```sh
+city -k n,s,c -o ./flat-area.json -c -f p
+```
+
+3. 倒树方式列出所有地区，子地区的code为key，val为父地区的{code:code, name:name};
+```sh
+city -k n,s,c -o ./reverse-area.js -j reverse_area -r
+```
+
 ### Todo
@@ -68,1 +93,8 @@
 - [ ] 支持更多格式的输出<https://github.com/TooBug/city/issues/2>
+
+### Changelog
+
+2020.06.12
+
+- 更新内置 `citydata.json` 数据
+- 移除 `市辖区` 数据项

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

New author

Supply chain risk

A new npm collaborator published a version of the package for the first time. New collaborators are usually benign additions to a project, but do indicate a change to the security surface area of a package.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Lines of code

363

increased by44.05%

Number of lines in readme file

96

increased by47.69%

Worsened metrics

Total package byte prevSize

148234

decreased by-3.56%

Dependency count

4

increased by33.33%

Number of medium supply chain risk alerts

1

increased byInfinity%

Dependency changes

• + Addedpinyin@^2.6.2

• + Added@mapbox/node-pre-gyp@1.0.11(transitive)

• + Addedabbrev@1.1.1(transitive)

• + Addedagent-base@6.0.2(transitive)

• + Addedansi-regex@5.0.1(transitive)

• + Addedaproba@2.0.0(transitive)

• + Addedare-we-there-yet@2.0.0(transitive)

• + Addedbalanced-match@1.0.2(transitive)

• + Addedbrace-expansion@1.1.11(transitive)

• + Addedchownr@2.0.0(transitive)

• + Addedcolor-support@1.1.3(transitive)

• + Addedcommander@1.1.1(transitive)

• + Addedconcat-map@0.0.1(transitive)

• + Addedconsole-control-strings@1.1.0(transitive)

• + Addeddebug@4.3.7(transitive)

• + Addeddelegates@1.0.0(transitive)

• + Addeddetect-libc@2.0.3(transitive)

• + Addedemoji-regex@8.0.0(transitive)

• + Addedfs-minipass@2.1.0(transitive)

• + Addedfs.realpath@1.0.0(transitive)

• + Addedgauge@3.0.2(transitive)

• + Addedglob@7.2.3(transitive)

• + Addedhas-unicode@2.0.1(transitive)

• + Addedhttps-proxy-agent@5.0.1(transitive)

• + Addedinflight@1.0.6(transitive)

• + Addedis-fullwidth-code-point@3.0.0(transitive)

• + Addedkeypress@0.1.0(transitive)

• + Addedmake-dir@3.1.0(transitive)

• + Addedminimatch@3.1.2(transitive)

• + Addedminipass@3.3.65.0.0(transitive)

• + Addedminizlib@2.1.2(transitive)

• + Addedmkdirp@1.0.4(transitive)

• + Addedms@2.1.3(transitive)

• + Addednode-addon-api@3.2.1(transitive)

• + Addednode-fetch@2.7.0(transitive)

• + Addednodejieba@2.5.2(transitive)

• + Addednopt@5.0.0(transitive)

• + Addednpmlog@5.0.1(transitive)

• + Addedobject-assign@4.1.1(transitive)

• + Addedonce@1.4.0(transitive)

• + Addedpath-is-absolute@1.0.1(transitive)

• + Addedpinyin@2.11.2(transitive)

• + Addedreadable-stream@3.6.2(transitive)

• + Addedrimraf@3.0.2(transitive)

• + Addedsafe-buffer@5.2.1(transitive)

• + Addedsemver@6.3.17.6.3(transitive)

• + Addedset-blocking@2.0.0(transitive)

• + Addedsignal-exit@3.0.7(transitive)

• + Addedstring-width@4.2.3(transitive)

• + Addedstring_decoder@1.3.0(transitive)

• + Addedstrip-ansi@6.0.1(transitive)

• + Addedtar@6.2.1(transitive)

• + Addedtr46@0.0.3(transitive)

• + Addedutil-deprecate@1.0.2(transitive)

• + Addedwebidl-conversions@3.0.1(transitive)

• + Addedwhatwg-url@5.0.0(transitive)

• + Addedwide-align@1.1.5(transitive)

• + Addedwrappy@1.0.2(transitive)

• + Addedyallist@4.0.0(transitive)