Python 2.7 从CMS NPI数据查询JSON数据

在这上面敲上几天，可能需要一个叫醒电话！ CMS（医疗保险和医疗补助服务中心）提供了一个API，用于根据个人的NPI（国家提供者标识符）访问医生信息

这里有很多信息，包括megafile每月下载等，但我不需要这些。我只需要对预先限定的单个NPI发出一个查询（低容量），并从检索到的记录中返回一些值

下面是一个随机选择的NPI的示例查询--

如果在浏览器窗口中运行此操作，您将看到结果JSON数据封装在一些页眉/页脚HTML中

我可以转储整个查询结果并以几种不同的方式打印，但无法选择和打印特定的数据元素，如姓名、地址或电话号码。如果在浏览器中运行查询，则可以看到原始输出，下面的代码段将打印结果的净化版本。见下文。想法

import urllib
from bs4 import BeautifulSoup
import json

def main():

url = "https://npiregistry.cms.hhs.gov/api/resultsDemo2/?number=1881761864&pretty=on"
html = urllib.urlopen(url).read()
soup = BeautifulSoup(html,"lxml")

for script in soup(["script", "style"]):
    script.extract()

practitioner_rec = soup.get_text()

# strip out the html to retain the JSON record
lines = (line.strip() for line in practitioner_rec.splitlines())
chunks = (phrase.strip() for line in lines for phrase in line.split("  "))
practitioner_rec = '\n'.join(chunk for chunk in chunks if chunk)

# get a count of lines generated by the query, valid queries are greater than 3 lines long
linect = practitioner_rec.count('\n') +1

if linect == 3:
    VALID_NPI="FALSE"
    VALID_MD="FALSE"
else: # approx. 69 lines of output here
    #   possible issue with JSON formmatting here
    #   In particular, the line   
    #   "result_count":1, "results":[
    #   since result count will always be 1, discard it
    practitioner_rec = practitioner_rec.replace('"result_count":1, ', '')
    print(practitioner_rec)

    practitioner_data = json.loads(practitioner_rec)
    VALID_NPI="TRUE"
    VALID_MD="TRUE"

    '''
    none of these constructs works to print the provider name
    print ['result_count']['results']['basic']['name'],"name"
    print result_count['results']['basic']['name'],"name"
    print practitioner_data['results']['basic']['name'],"name"
    print results['basic']['name'],"name"
    print ['basic']['name'],"name"
    print basic['name'],"name"
    print results[2]['basic']['name'],"name"
    print results['basic']['name'],"name"

    this works, but not useful if I can't pick values out
    print(json.dumps(practitioner_data))      

    print "VALID_NPI is ",VALID_NPI
    print "VALID_MD is  ",VALID_MD
    return [VALID_NPI,VALID_MD]    
    '''


if __name__ == '__main__':
    main()

---

头晕过去了。我是个json新手，现在我是介绍人。下面是一个简短的代码段，以防其他人想要查询CMS NPI JSON数据并从中获取结果。不需要商业API。我看到引用的Bloom one似乎没有激活，其他人需要注册和跟踪数据才能访问公共数据

下面是访问单个字段的代码--

---

在SimpleTalk.com红门网站上有几种方法可以做到这一点：

import urllib
from bs4 import BeautifulSoup
import json

def main():

'''
NOTES:
1.  pretty switch works set to either true OR on
2.  a failed NPI search produces 3-line output like this --
    {
    "result_count":0, "results":[
    ]}
'''

# valid NPI
url = "https://npiregistry.cms.hhs.gov/api/resultsDemo2/?number=1881761864&pretty=on"
html = urllib.urlopen(url).read()
soup = BeautifulSoup(html,"lxml")
# remove HTML from output, producing just a JSON record    
practitioner_rec = soup.text

# count lines generated by the query, valid queries are > than 3 lines long
linect = practitioner_rec.count('\n') +1
#print "there are ", linect," lines in the input file" # only for testing

if linect == 3:
    VALID_NPI="FALSE"
    VALID_MD="FALSE"
else: 
    '''
    query produces a single result, with approx. 60+ lines of output
    JSON data a little squirrelly, so we have to 
    '''
    practitioner_rec = practitioner_rec.replace('"result_count":1, ', '')
    # print(practitioner_rec) # only for testing

    provider_dict = json.loads(practitioner_rec)
    provider_info = provider_dict['results'][0]['basic']
    print("name:", str(provider_info['name'])) # str-strip out unicode tag

    VALID_NPI="TRUE"
    VALID_MD="TRUE"

print "VALID_NPI is ",VALID_NPI
print "VALID_MD is  ",VALID_MD
return [VALID_NPI,VALID_MD]    

if __name__ == '__main__':
    main()