def cas_pdf_to_text(filename: Union[str, io.IOBase],
password) -> PartialCASData:
"""
Parse CAS pdf and returns line data.
:param filename: CAS pdf file (CAMS or Kfintech)
:param password: CAS pdf password
:return: partial cas data with FileType, InvestorInfo and lines of data
"""
file_type: FileType = FileType.UNKNOWN
if isinstance(filename, str):
fp = open(filename, "rb")
elif isinstance(filename, io.IOBase):
fp = filename
elif hasattr(filename, "read"): # compatibility for Django UploadedFile
fp = filename
else:
raise CASParseError(
"Invalid input. filename should be a string or a file like object")
with fp:
try:
doc = fitz.open(stream=fp.read(), filetype="pdf")
except Exception as e:
raise CASParseError("Unhandled error while opening file :: %s" %
(str(e)))
if doc.needsPass:
rc = doc.authenticate(password)
if not rc:
raise CASParseError("Incorrect PDF password!")
pages = []
investor_info = None
for page in doc:
text_page = page.getTextPage()
page_dict = text_page.extractDICT()
blocks = extract_blocks(page_dict)
if file_type == FileType.UNKNOWN:
file_type = parse_file_type(blocks)
sorted_blocks = sorted(blocks, key=itemgetter(1, 0))
if investor_info is None:
investor_info = parse_investor_info(page_dict)
pages.append(sorted_blocks)
lines = group_similar_rows(pages)
return PartialCASData(file_type=file_type,
investor_info=investor_info,
lines=lines)
if current_folio is None or current_folio != folio:
if curr_scheme_data and current_folio is not None:
folios[current_folio]["schemes"].append(curr_scheme_data)
curr_scheme_data = {}
current_folio = folio
folios[folio] = {
"folio": current_folio,
"amc": current_amc,
"PAN": "",
"KYC": "",
"PANKYC": "",
"schemes": [],
}
elif m := re.search(SCHEME_RE, line, re.DOTALL | re.MULTILINE | re.I):
if current_folio is None:
raise CASParseError(
"Layout Error! Scheme found before folio entry.")
scheme = re.sub(r"\(formerly.+?\)",
"",
m.group(2),
flags=re.I | re.DOTALL).strip()
if curr_scheme_data.get("scheme") != scheme:
if curr_scheme_data:
folios[current_folio]["schemes"].append(curr_scheme_data)
advisor = m.group(3)
if advisor is not None:
advisor = advisor.strip()
rta = m.group(4).strip()
rta_code = m.group(1).strip()
isin, amfi = isin_search(scheme, rta, rta_code)
curr_scheme_data = {
"scheme": scheme,
email_found = True
continue
if name is None:
name = txt
else:
if m := re.search(r"mobile\s*:\s*([+\d]+)(?:s|$)", txt,
re.I):
mobile = m.group(1).strip()
address_lines.append(txt)
if mobile is not None:
return InvestorInfo(email=email,
name=name,
mobile=mobile,
address="\n".join(address_lines))
if email is None or mobile is None:
raise CASParseError("Unable to parse investor data")
def group_similar_rows(elements_list: List[Iterator[Any]]):
"""
Group elements having similar rows, with a tolerance.
:param elements_list: List of elements from each page
"""
lines = []
for elements in elements_list:
sorted_elements = list(sorted(elements, key=itemgetter(3, 0)))
if len(sorted_elements) == 0:
continue
y0, y1 = sorted_elements[0][1], sorted_elements[0][3]
items = []
def cas_pdf_to_text(filename: Union[str, io.IOBase],
password) -> PartialCASData:
"""
Parse CAS pdf and returns line data.
:param filename: CAS pdf file (CAMS or Kfintech)
:param password: CAS pdf password
:return: array of lines from the CAS.
"""
file_type: Optional[FileType] = None
if isinstance(filename, str):
fp = open(filename, "rb")
elif hasattr(filename, "read") and hasattr(filename,
"close"): # file-like object
fp = filename
else:
raise CASParseError(
"Invalid input. filename should be a string or a file like object")
with fp:
pdf_parser = PDFParser(fp)
try:
document = PDFDocument(pdf_parser, password=password)
except PDFPasswordIncorrect:
raise CASParseError("Incorrect PDF password!")
except PDFSyntaxError:
raise CASParseError("Unhandled error while opening file")
line_margin = {
FileType.KFINTECH: 0.1,
FileType.CAMS: 0.2
}.get(detect_pdf_source(document), 0.2)
rsrc_mgr = PDFResourceManager()
laparams = LAParams(line_margin=line_margin, detect_vertical=True)
device = PDFPageAggregator(rsrc_mgr, laparams=laparams)
interpreter = PDFPageInterpreter(rsrc_mgr, device)
pages: List[Iterator[LTTextBoxHorizontal]] = []
investor_info = None
for page in PDFPage.create_pages(document):
interpreter.process_page(page)
layout = device.get_result()
text_elements = filter(
lambda x: isinstance(x, LTTextBoxHorizontal), layout)
if file_type is None:
for el in filter(lambda x: isinstance(x, LTTextBoxVertical),
layout):
if re.search("CAMSCASWS", el.get_text()):
file_type = FileType.CAMS
if re.search("KFINCASWS", el.get_text()):
file_type = FileType.KFINTECH
if investor_info is None:
investor_info = parse_investor_info(layout, *page.mediabox[2:])
pages.append(text_elements)
lines = group_similar_rows(pages)
return PartialCASData(file_type=file_type,
investor_info=investor_info,
lines=lines)