Skip to content

Instantly share code, notes, and snippets.

@dhruvilp

dhruvilp/html_parser.py

Created September 19, 2021 02:57
Show Gist options
  • Save dhruvilp/e1d666767b743771689f626d538eae1e to your computer and use it in GitHub Desktop.
Save dhruvilp/e1d666767b743771689f626d538eae1e to your computer and use it in GitHub Desktop.
Download ZIP
HTML Parser for DB2 Responses
Raw
html_parser.py
from html.parser import HTMLParser
from html.entities import name2codepoint
keys = []
values = []
dataTypes = ["CHAR","VARCHAR","BOOL","BOOLEAN","SMALLINT","INT","INTEGER","DOUBLE","DECIMAL","DATE","DATETIME","TIMESTAMP"]
class MyHTMLParser(HTMLParser):
def handle_starttag(self, tag, attrs):
for attr in attrs:
if "Tip(" in "".join(attr):
fm1 = "".join(attr).replace("Tip(","")
fm2 = fm1.replace(")", "")
MyHTMLParser().feed(fm2)
def handle_data(self, data):
if data[0] == " " and ":" not in data:
if data.replace(" ", "") not in dataTypes:
keys.append(data.replace(" ", ""))
elif data.replace(" ", "") in dataTypes:
if data.replace(" ", "") == "DECIMAL":
values.append("BigDecimal")
elif data.replace(" ", "") == "VARCHAR":
values.append("String")
elif data.replace(" ", "") == "INTEGER":
values.append("int")
else:
values.append(data.replace(" ", "").capitalize())
parser = MyHTMLParser()
parser.feed("""<body><table><tr><td onmouseover="Tip('<b>Label: </b> FPS_PRICE_IND<br/><b>Type: </b> INTEGER<br/><b>Size: </b>1<br/><b>IsNullable: </b>non nullable<br/><b>Precision: </b>1<br/><b>Table: </b>SCHEMA.TABLE_NAME', TITLE, 'FPS_PRICE_IND')" onmouseout="UnTip()">FPS_PRICE_IND</td></tr><tr><td onmouseover="Tip('<b>Label: </b> GOVT_ID<br/><b>Type: </b> DECIMAL<br/><b>Size: </b>1<br/><b>IsNullable: </b>non nullable<br/><b>Precision: </b>1<br/><b>Table: </b>SCHEMA.TABLE_NAME', TITLE, 'FPS_PRICE_IND')" onmouseout="UnTip()">FPS_PRICE_IND</td></tr></table></body>""")
print("\nkeys: ", keys)
print("values: ", values)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment