qurator-spk/dinglehopper
1
0
Fork 0
mirror of https://github.com/qurator-spk/dinglehopper.git synced 2025-06-09 11:50:00 +02:00
Code Issues Releases Wiki Activity

🚧 dinglehopper: Re-introduce "substitute_equivalences" as Normalization.NFC_SBB

Browse source
This commit is contained in:
Gerber, Mike 2020-06-12 15:53:15 +02:00
parent 2579e0220c
commit a320d5fd8f
1 changed files with 19 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

22
qurator/dinglehopper/ocr_files.py
View file

@ -5,6 +5,7 @@ from warnings import warn
from lxml import etree as ET from lxml import etree as ET
from lxml.etree import XMLSyntaxError from lxml.etree import XMLSyntaxError
from contextlib import suppress from contextlib import suppress
from .substitute_equivalences import substitute_equivalences
import sys import sys
import attr import attr
import enum import enum
@ -36,16 +37,27 @@ class ExtractedText:
class Normalization(enum.Enum): class Normalization(enum.Enum):
NFC = 1 NFC = 1
NFC_MUFI = 2 NFC_MUFI = 2 # TODO
NFC_SBB = 3
def normalize(text, normalization): def normalize(text, normalization):
if normalization == Normalization.NFC: if normalization == Normalization.NFC:
return unicodedata.normalize('NFC', text) return unicodedata.normalize('NFC', text)
if normalization == Normalization.NFC_MUFI:
raise NotImplementedError()
if normalization == Normalization.NFC_SBB:
# XXX This needs to be redone
# https://github.com/qurator-spk/dinglehopper/issues/11
return substitute_equivalences(text)
else: else:
raise ValueError() raise ValueError()
# XXX hack
normalize_sbb = lambda t: normalize(t, Normalization.NFC_SBB)
@attr.s(frozen=True) @attr.s(frozen=True)
class ExtractedTextSegment: class ExtractedTextSegment:
id = attr.ib(type=str) id = attr.ib(type=str)
@ -54,7 +66,7 @@ class ExtractedTextSegment:
def check(self, attribute, value): def check(self, attribute, value):
if value is not None and normalize(value, self.normalization) != value: if value is not None and normalize(value, self.normalization) != value:
raise ValueError('String "{}" is not normalized.'.format(value)) raise ValueError('String "{}" is not normalized.'.format(value))
normalization = attr.ib(converter=Normalization, default=Normalization.NFC) normalization = attr.ib(converter=Normalization, default=Normalization.NFC_SBB)
@classmethod @classmethod
def from_text_segment(cls, text_segment, nsmap): def from_text_segment(cls, text_segment, nsmap):
@ -64,6 +76,7 @@ class ExtractedTextSegment:
segment_text = None segment_text = None
with suppress(AttributeError): with suppress(AttributeError):
segment_text = text_segment.find('./page:TextEquiv/page:Unicode', namespaces=nsmap).text segment_text = text_segment.find('./page:TextEquiv/page:Unicode', namespaces=nsmap).text
segment_text = normalize_sbb(segment_text)
return cls(segment_id, segment_text) return cls(segment_id, segment_text)
@ -89,7 +102,10 @@ def alto_extract(tree):
' '.join(string.attrib.get('CONTENT') for string in line.iterfind('alto:String', namespaces=nsmap)) ' '.join(string.attrib.get('CONTENT') for string in line.iterfind('alto:String', namespaces=nsmap))
for line in tree.iterfind('.//alto:TextLine', namespaces=nsmap)) for line in tree.iterfind('.//alto:TextLine', namespaces=nsmap))
return ExtractedText((ExtractedTextSegment(None, line_text) for line_text in lines), '\n') return ExtractedText(
(ExtractedTextSegment(None, normalize_sbb(line_text)) for line_text in lines),
'\n'
)
# TODO This currently does not extract any segment id, because we are # TODO This currently does not extract any segment id, because we are
# clueless about the ALTO format. # clueless about the ALTO format.
# FIXME needs to handle normalization # FIXME needs to handle normalization