Last active
August 29, 2015 14:01
-
-
Save yhsiang/be11bd02eb0559cfd544 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
require! <[request cheerio fs async prelude-ls]> | |
_ = prelude-ls | |
jar = request.jar! | |
request = request.defaults jar: jar | |
base-form = | |
'__EVENTTARGET':'' | |
'__EVENTARGUMENT':'' | |
'__VIEWSTATE': '' | |
'__EVENTVALIDATION': '' | |
'__VIEWSTATE_ID': '' | |
'WR1_1$btnNext': '下一步>' | |
form-step1 = | |
'WR1_1_Q_DataII': '1' | |
'WR1_1_Q_PointII': 'D' | |
'WR1_1_Q_SexII': '3' | |
'WR1_1$ctl138': 'on' | |
'WR1_1$ctl143':'on' | |
'WR1_1$ctl148':'on' | |
'WR1_1$ctl153':'on' | |
'WR1_1$ctl158':'on' | |
'WR1_1$ctl163':'on' | |
'WR1_1$ctl168':'on' | |
'WR1_1$ctl173':'on' | |
'WR1_1$ctl178':'on' | |
'WR1_1$ctl183':'on' | |
'WR1_1$ctl188':'on' | |
'WR1_1$ctl193':'on' | |
'WR1_1$ctl198':'on' | |
'WR1_1$ctl203':'on' | |
'WR1_1$ctl208':'on' | |
'WR1_1$ctl213':'on' | |
'WR1_1$ctl218':'on' | |
'WR1_1$ctl223':'on' | |
'WR1_1$ctl228':'on' | |
'WR1_1$ctl233':'on' | |
'WR1_1$ctl238':'on' | |
form-step2 = | |
'WR1_1_Q_YearBeginII': '1979' | |
'WR1_1_Q_YearEndII': '2010' | |
form-step3 = | |
'WR1_1$Q_AgeKind': 'Q_AgeKindAll' | |
'WR1_1_Q_AgeFromII': '' | |
'WR1_1_Q_AgeToII': '' | |
form-step4 = | |
'WR1_1_QP_AreaRegionII':'ROOT' | |
'WR1_1$Q_AreaGrp':'Q_AreaTown' | |
#'WR1_1_ctl10_0':'' #CRA_3102' | |
form-step5 = | |
'WR1_1_Q_ReportKindII':'T' | |
'WR1_1_Q_ChartKindII':'L' | |
'WR1_1$cmdQuery':'產生報表' | |
fetch = (options, next) -> | |
url = options.url || 'https://cris.hpa.gov.tw/pagepub/Home.aspx?itemNo=cr.q.10' | |
method = options.method || 'POST' | |
(error, res, body) <- request do | |
url: url | |
method: method | |
headers: | |
'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/34.0.1847.131' | |
'Referer': 'https://cris.hpa.gov.tw/pagepub/Home.aspx?itemNo=cr.q.10' | |
strictSSL: false | |
form: options.form | |
$ = cheerio.load body | |
event-validation = $('#__EVENTVALIDATION').val! | |
viewstate-id = $('#__VIEWSTATE_ID').val! | |
base-form.__EVENTVALIDATION = event-validation | |
base-form.__VIEWSTATE_ID = viewstate-id | |
next error, res, body, form: base-form | |
export-file = (area-code) -> | |
(err, res, body, it)<- fetch do | |
url: 'https://cris.hpa.gov.tw/pagepub/Home.aspx?itemNo=cr.q.10' | |
method: 'GET' | |
form = {} | |
form <<<< it.form <<<< form-step1 | |
(err, res, body, it)<- fetch do | |
form: form | |
form = {} | |
form <<<< it.form <<<< form-step2 | |
(err, res, body, it)<- fetch do | |
form: form | |
form = {} | |
form <<<< it.form <<<< form-step3 | |
(err, res, body, it)<- fetch do | |
form: form | |
form = {} | |
form-step4 <<<< area-code | |
form <<<< it.form <<<< form-step4 | |
(err, res, body, it)<- fetch do | |
form: form | |
form = {} | |
form <<<< it.form <<<< form-step5 | |
(err, res, body, it)<- fetch do | |
form: form | |
#TODO error handle here | |
#console.log res.headers.location | |
report-url = 'https://cris.hpa.gov.tw' + res.headers.location | |
(err, res, body, it)<- fetch do | |
url: report-url | |
method: 'GET' | |
$ = cheerio.load body | |
export-url = 'https://cris.hpa.gov.tw' + $('iframe').attr('src') | |
(err, res, body, it)<- fetch do | |
url: export-url | |
method: 'GET' | |
$ = cheerio.load body | |
paths = $('#report').attr('src') / '&' | |
xls-url = 'https://cris.hpa.gov.tw' + paths.0 + '&' + paths.1 + '&' + paths.2 + '&' + paths.3 + '&' + paths.4 + '&' + paths.5 | |
xls-url += '&OpType=Export&FileName=CrReportN_A01_C&ContentDisposition=AlwaysAttachment&Format=Excel' | |
file-name = _.values area-code | |
console.log "Export to data/#{file-name.0}.xls" | |
file = fs.createWriteStream "data/#{file-name.0}.xls" | |
request do | |
url: xls-url | |
headers: | |
'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/34.0.1847.131' | |
strictSSL: false | |
.pipe file | |
lists = | |
* WR1_1_ctl95_0':'CRA_0201' | |
* WR1_1_ctl95_1':'CRA_0202' | |
* WR1_1_ctl95_2':'CRA_0203' | |
* WR1_1_ctl95_3':'CRA_0204' | |
* WR1_1_ctl95_4':'CRA_0205' | |
* WR1_1_ctl95_5':'CRA_0206' | |
* WR1_1_ctl95_6':'CRA_0207' | |
* WR1_1_ctl95_7':'CRA_0208' | |
* WR1_1_ctl95_8':'CRA_0209' | |
* WR1_1_ctl95_9':'CRA_0210' | |
* WR1_1_ctl95_10':'CRA_0211' | |
* WR1_1_ctl100_0':'CRA_4301' | |
* WR1_1_ctl100_1':'CRA_4302' | |
* WR1_1_ctl100_2':'CRA_4303' | |
* WR1_1_ctl100_3':'CRA_4304' | |
* WR1_1_ctl100_4':'CRA_4305' | |
* WR1_1_ctl100_5':'CRA_4306' | |
* WR1_1_ctl100_6':'CRA_4307' | |
* WR1_1_ctl100_7':'CRA_4308' | |
* WR1_1_ctl100_8':'CRA_4309' | |
* WR1_1_ctl100_9':'CRA_4310' | |
* WR1_1_ctl100_10':'CRA_4311' | |
* WR1_1_ctl100_11':'CRA_4312' | |
* WR1_1_ctl100_12':'CRA_4313' | |
* WR1_1_ctl100_13':'CRA_4314' | |
* WR1_1_ctl100_14':'CRA_4315' | |
* WR1_1_ctl100_15':'CRA_4316' | |
* WR1_1_ctl100_16':'CRA_4317' | |
* WR1_1_ctl100_17':'CRA_4318' | |
* WR1_1_ctl100_18':'CRA_4319' | |
* WR1_1_ctl100_19':'CRA_4320' | |
* WR1_1_ctl100_20':'CRA_4321' | |
* WR1_1_ctl100_21':'CRA_4322' | |
* WR1_1_ctl100_22':'CRA_4323' | |
* WR1_1_ctl100_23':'CRA_4324' | |
* WR1_1_ctl100_24':'CRA_4325' | |
* WR1_1_ctl100_25':'CRA_4326' | |
* WR1_1_ctl100_26':'CRA_4327' | |
* WR1_1_ctl100_27':'CRA_4328' | |
* WR1_1_ctl100_28':'CRA_4329' | |
* WR1_1_ctl100_29':'CRA_4330' | |
* WR1_1_ctl100_30':'CRA_4331' | |
* WR1_1_ctl100_31':'CRA_4332' | |
* WR1_1_ctl100_32':'CRA_4333' | |
* WR1_1_ctl105_0':'CRA_3401' | |
* WR1_1_ctl105_1':'CRA_3402' | |
* WR1_1_ctl105_2':'CRA_3403' | |
* WR1_1_ctl105_3':'CRA_3404' | |
* WR1_1_ctl105_4':'CRA_3405' | |
* WR1_1_ctl105_5':'CRA_3406' | |
* WR1_1_ctl105_6':'CRA_3407' | |
* WR1_1_ctl105_7':'CRA_3408' | |
* WR1_1_ctl105_8':'CRA_3409' | |
* WR1_1_ctl105_9':'CRA_3410' | |
* WR1_1_ctl105_10':'CRA_3411' | |
* WR1_1_ctl105_11':'CRA_3412' | |
* WR1_1_ctl110_0':'CRA_4501' | |
* WR1_1_ctl110_1':'CRA_4502' | |
* WR1_1_ctl110_2':'CRA_4503' | |
* WR1_1_ctl110_3':'CRA_4504' | |
* WR1_1_ctl110_4':'CRA_4505' | |
* WR1_1_ctl110_5':'CRA_4506' | |
* WR1_1_ctl110_6':'CRA_4507' | |
* WR1_1_ctl110_7':'CRA_4508' | |
* WR1_1_ctl110_8':'CRA_4509' | |
* WR1_1_ctl110_9':'CRA_4510' | |
* WR1_1_ctl110_10':'CRA_4511' | |
* WR1_1_ctl110_11':'CRA_4512' | |
* WR1_1_ctl110_12':'CRA_4513' | |
* WR1_1_ctl115_0':'CRA_4601' | |
* WR1_1_ctl115_1':'CRA_4602' | |
* WR1_1_ctl115_2':'CRA_4603' | |
* WR1_1_ctl115_3':'CRA_4604' | |
* WR1_1_ctl115_4':'CRA_4605' | |
* WR1_1_ctl115_5':'CRA_4606' | |
* WR1_1_ctl115_6':'CRA_4607' | |
* WR1_1_ctl115_7':'CRA_4608' | |
* WR1_1_ctl115_8':'CRA_4609' | |
* WR1_1_ctl115_9':'CRA_4610' | |
* WR1_1_ctl115_10':'CRA_4611' | |
* WR1_1_ctl115_11':'CRA_4612' | |
* WR1_1_ctl115_12':'CRA_4613' | |
* WR1_1_ctl115_13':'CRA_4614' | |
* WR1_1_ctl115_14':'CRA_4615' | |
* WR1_1_ctl115_15':'CRA_4616' | |
* WR1_1_ctl120_0':'CRA_4401' | |
* WR1_1_ctl120_1':'CRA_4402' | |
* WR1_1_ctl120_2':'CRA_4403' | |
* WR1_1_ctl120_3':'CRA_4404' | |
* WR1_1_ctl120_4':'CRA_4405' | |
* WR1_1_ctl120_5':'CRA_4406' | |
* WR1_1_ctl125_0':'CRA_9001' | |
* WR1_1_ctl125_1':'CRA_9002' | |
* WR1_1_ctl125_2':'CRA_9003' | |
* WR1_1_ctl125_3':'CRA_9004' | |
* WR1_1_ctl125_4':'CRA_9005' | |
* WR1_1_ctl125_5':'CRA_9006' | |
* WR1_1_ctl130_0':'CRA_9101' | |
* WR1_1_ctl130_1':'CRA_9102' | |
* WR1_1_ctl130_2':'CRA_9103' | |
* WR1_1_ctl130_3':'CRA_9104' | |
fs.mkdir 'data' unless fs.existsSync 'data' | |
code <- lists.forEach | |
export-file code |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment