junte/survey-extraction
1
0
Fork 0
mirror of https://github.com/Hestia-Homes/survey-extraction.git synced 2026-06-08 11:17:29 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

section 8 completed

Browse source
This commit is contained in:
Jun-te Kim 2025-03-12 12:34:55 +00:00
parent 3f51c8f529
commit 2d10ea2f80
2 changed files with 35 additions and 13 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
etl/main.py
View file

@ -11,8 +11,8 @@ DATA_LOC_2 = "/tmp/sharepoint/Abdul Koddus/W.C. 03.03.2025/Southern Housing/16 S
pdfReader = pdfReaderToText(DATA_LOC_1)
doc1 = pdfReader.get_reader()
pdfReader = pdfReaderToText(DATA_LOC_2)
doc2 = pdfReader.get_reader()
pdfReader2 = pdfReaderToText(DATA_LOC_2)
doc2 = pdfReader2.get_reader()
def main():

44
etl/pdfReader/sitenotes.py
View file

@ -30,6 +30,7 @@ class QuidosSiteNotes(SiteNotes):
self.get_section_5()
self.get_section_6()
self.get_section_7()
self.get_section_8()
def get_summary_information(self):
# Summary Information
@ -237,8 +238,6 @@ class QuidosSiteNotes(SiteNotes):
def get_section_7(self):
data = self.raw_data[self.raw_data.index('7.0 Walls'): self.raw_data.index('8.0 Roofs')]
avoid = [
"7.0 Walls",
"8.0 Roofs",
"Construction",
"Insulation",
"Insulation Thickness(mm)",
@ -268,17 +267,40 @@ class QuidosSiteNotes(SiteNotes):
if item in avoid:
setattr(self, f"extensions_{j}_{item.lower().replace(' ', '_').replace('-', '_')}", get_value(item))
def get_section_8(self):
data = self.raw_data[self.raw_data.index('8.0 Roofs'): self.raw_data.index('9.0 Floors')]
avoid = [
"Construction",
"Insulation Type",
"Insulation Thickness",
"U-value Known",
]
titles = [
"Main Property",
"Extension 1",
"Extension 2",
"Extension 3",
"Extension 4",
]
title = titles[0].lower().replace(" ", "_").replace("-","_")
proc_data = data
for items in data:
if items in titles:
title = items.lower().replace(" ", "_").replace("-", "_")
index = titles.index(items)
if titles[index] in data:
proc_data = data[data.index(titles[index]):]
continue
else:
break
get_value = lambda key: None if proc_data[proc_data.index(key) + 1] in avoid else proc_data[proc_data.index(key) + 1]
if items in avoid:
setattr(self, f"section_8_{title}_{items.lower().replace(' ', '_').replace('-','_')}", get_value(items))
def get_section_8(self):
pass
# validatin
# function of object of type this