def test_extent_before_translation() -> None:
with pytest.raises(ValueError):
Text.of_iterable(
[
TextLine.of_iterable(LineNumber(1), [Word.of([Reading.of_name("bu")])]),
TextLine.of_iterable(LineNumber(2), [Word.of([Reading.of_name("bu")])]),
TranslationLine(tuple(), "en", Extent(LineNumber(1))),
]
)
def test_extent_overlapping_languages() -> None:
Text.of_iterable(
[
TextLine.of_iterable(LineNumber(1), [Word.of([Reading.of_name("bu")])]),
TranslationLine(tuple(), "en", Extent(LineNumber(2))),
TextLine.of_iterable(LineNumber(2), [Word.of([Reading.of_name("bu")])]),
TranslationLine(tuple(), "de"),
]
)
def test_exent_overlapping() -> None:
with pytest.raises(ValueError):
Text.of_iterable(
[
TextLine.of_iterable(LineNumber(1), [Word.of([Reading.of_name("bu")])]),
TranslationLine(tuple(), extent=Extent(LineNumber(2))),
TextLine.of_iterable(LineNumber(2), [Word.of([Reading.of_name("bu")])]),
TranslationLine(tuple()),
]
)
def parse_atf_lark(atf_):
def parse_line_(line: str, line_number: int):
try:
parsed_line = parse_line(line) if line else EmptyLine()
validate_line(parsed_line)
return parsed_line, None
except PARSE_ERRORS as ex:
return (None,
create_transliteration_error_data(ex, line, line_number))
def check_errors(pairs):
errors = [error for line, error in pairs if error is not None]
if any(errors):
raise TransliterationError(errors)
lines = atf_.split("\n")
lines = list(dropwhile(lambda line: line == "", reversed(lines)))
lines.reverse()
lines = [parse_line_(line, number) for number, line in enumerate(lines)]
check_errors(lines)
lines = tuple(pair[0] for pair in lines)
text = Text(lines, f"{atf.ATF_PARSER_VERSION}")
if pydash.duplicates(text.labels):
raise DataError("Duplicate labels.")
else:
return text
def test_update_lemmatization() -> None:
tokens = [list(line) for line in TEXT.lemmatization.tokens]
tokens[0][0] = LemmatizationToken(tokens[0][0].value, (WordId("nu I"),))
lemmatization = Lemmatization(tokens)
expected = Text(
(
TextLine(
LineNumber(1),
(
Word.of(
unique_lemma=(WordId("nu I"),),
parts=[
Reading.of_name("ha"),
Joiner.hyphen(),
Reading.of_name("am"),
],
),
),
),
RulingDollarLine(atf.Ruling.SINGLE),
),
TEXT.parser_version,
)
assert TEXT.update_lemmatization(lemmatization) == expected
class TransliterationUpdate:
text: Text = Text()
notes: str = ""
signs: str = attr.ib(default="")
@signs.validator
def _check_signs(self, _attribute, value) -> None:
questionable_lines = self._get_questionable_lines(value)
if questionable_lines:
raise TransliterationError(
[
{"description": "Invalid value", "lineNumber": line_number}
for line_number in questionable_lines
]
)
def _get_questionable_lines(self, value: str) -> List[int]:
lines = [line.atf for line in self.text.lines]
text_lines = [line.atf for line in self.text.text_lines]
signs = value.split("\n")
def get_line_number(text_line_number: int) -> int:
line = text_lines[text_line_number]
return lines.index(line) + 1
return [
get_line_number(index) for index, line in enumerate(signs) if "?" in line
]
def test_dump_line():
text = Text(
(
TextLine.of_iterable(
LineNumber(1),
[
Word.of(
parts=[
Reading.of_name("ha"),
Joiner.hyphen(),
Reading.of_name("am"),
]
)
],
),
EmptyLine(),
ControlLine("#", " comment"),
),
"1.0.0",
)
assert TextSchema().dump(text) == {
"lines": OneOfLineSchema().dump(text.lines, many=True),
"parser_version": text.parser_version,
"numberOfLines": 1,
}
def test_parse_ruling_dollar_line(prefix, ruling, expected_ruling, status,
expected_status, status_space, parenthesis):
ruling = f"{ruling} ruling"
ruling_with_status = (f"{ruling} {status}" if
(status and status_space) else f"{ruling}{status}")
line = f"({ruling_with_status})" if parenthesis else ruling_with_status
assert (parse_atf_lark(f"{prefix}{line}").lines == Text.of_iterable(
[RulingDollarLine(expected_ruling, expected_status)]).lines)
def test_statistics(database, fragment_repository):
database[COLLECTION].insert_many(
[
SCHEMA.dump(
FragmentFactory.build(
text=Text(
(
TextLine(
LineNumber(1),
(
Word.of([Reading.of_name("first")]),
Word.of([Reading.of_name("line")]),
),
),
ControlLine("#", "ignore"),
EmptyLine(),
)
)
)
),
SCHEMA.dump(
FragmentFactory.build(
text=Text(
(
ControlLine("#", "ignore"),
TextLine(
LineNumber(1), (Word.of([Reading.of_name("second")]),)
),
TextLine(
LineNumber(2), (Word.of([Reading.of_name("third")]),)
),
ControlLine("#", "ignore"),
TextLine(
LineNumber(3), (Word.of([Reading.of_name("fourth")]),)
),
)
)
)
),
SCHEMA.dump(FragmentFactory.build(text=Text())),
]
)
assert fragment_repository.count_transliterated_fragments() == 2
assert fragment_repository.count_lines() == 4
def text_with_labels():
return Text.of_iterable([
TextLine.of_iterable(LineNumber(1),
[Word.of([Reading.of_name("bu")])]),
ColumnAtLine(ColumnLabel.from_int(1)),
SurfaceAtLine(SurfaceLabel([], atf.Surface.SURFACE, "Stone wig")),
ObjectAtLine(ObjectLabel([], atf.Object.OBJECT, "Stone wig")),
TextLine.of_iterable(LineNumber(2),
[Word.of([Reading.of_name("bu")])]),
])
def test_load_line(lines):
parser_version = "2.3.1"
serialized_lines = OneOfLineSchema().dump(lines, many=True)
assert TextSchema().load(
{
"lines": serialized_lines,
"parser_version": parser_version,
"numberOfLines": 1,
}
) == Text.of_iterable(lines).set_parser_version(parser_version)
class InterestingFragmentFactory(FragmentFactory):
collection = "Kuyunjik" # pyre-ignore[15]
publication = "" # pyre-ignore[15]
joins: Sequence[str] = tuple()
text = Text()
notes = ""
uncurated_references = (
UncuratedReference("7(0)"),
UncuratedReference("CAD 51", (34, 56)),
UncuratedReference("7(1)"),
)
references = tuple()
class Manuscript:
id: int
siglum_disambiguator: str = ""
museum_number: Optional[MuseumNumber] = None
accession: str = attr.ib(default="")
period_modifier: PeriodModifier = PeriodModifier.NONE
period: Period = Period.NEO_ASSYRIAN
provenance: Provenance = attr.ib(default=Provenance.NINEVEH)
type: ManuscriptType = ManuscriptType.LIBRARY
notes: str = ""
colophon: Text = Text()
unplaced_lines: Text = Text()
references: Sequence[Reference] = tuple()
joins: Joins = Joins()
is_in_fragmentarium: bool = False
@accession.validator
def validate_accession(self, _, value) -> None:
if self.museum_number and value:
raise ValueError("Accession given when museum number present.")
@provenance.validator
def validate_provenance(self, _, value) -> None:
if is_invalid_standard_text(value, self.period, self.type):
raise ValueError(
"Manuscript must not have period and type when provenance is Standard Text."
)
elif is_invalid_non_standard_text(value, self.period, self.type):
raise ValueError(
"Manuscript must have period and type unless provenance is Standard Text."
)
@property
def text_lines(self) -> Sequence[TextLine]:
return [*self.colophon.text_lines, *self.unplaced_lines.text_lines]
@property
def siglum(self) -> Siglum:
return Siglum(self.provenance, self.period, self.type,
self.siglum_disambiguator)
def test_query_lemmas_ignores_in_value(parts, expected, fragment_repository,
lemma_repository):
fragment = FragmentFactory.build(
text=Text.of_iterable([
TextLine.of_iterable(
LineNumber(1),
[Word.of(parts, unique_lemma=(WordId("ana I"), ))])
]),
signs="DIŠ",
)
fragment_repository.create(fragment)
assert lemma_repository.query_lemmas("ana", False) == expected
class ManuscriptFactory(factory.Factory):
class Meta:
model = Manuscript
id = factory.Sequence(lambda n: n + 1)
siglum_disambiguator = factory.Faker("word")
museum_number = factory.Sequence(
lambda n: MuseumNumber("M", str(n)) if pydash.is_odd(n) else None
)
accession = factory.Sequence(lambda n: f"A.{n}" if pydash.is_even(n) else "")
period_modifier = factory.fuzzy.FuzzyChoice(PeriodModifier)
period = factory.fuzzy.FuzzyChoice(set(Period) - {Period.NONE})
provenance = factory.fuzzy.FuzzyChoice(set(Provenance) - {Provenance.STANDARD_TEXT})
type = factory.fuzzy.FuzzyChoice(set(ManuscriptType) - {ManuscriptType.NONE})
notes = factory.Faker("sentence")
colophon = Transliteration.of_iterable(
[TextLine.of_iterable(LineNumber(1, True), (Word.of([Reading.of_name("ku")]),))]
)
unplaced_lines = Transliteration.of_iterable(
[TextLine.of_iterable(LineNumber(1, True), (Word.of([Reading.of_name("nu")]),))]
)
references = factory.List(
[factory.SubFactory(ReferenceFactory, with_document=True)], TupleFactory
)
def test_updating_manuscripts(corpus, text_repository, bibliography, changelog,
signs, sign_repository, user, when) -> None:
uncertain_fragments = (MuseumNumber.of("K.1"), )
updated_chapter = attr.evolve(
CHAPTER,
manuscripts=(attr.evolve(
CHAPTER.manuscripts[0],
colophon=Transliteration.of_iterable([
TextLine.of_iterable(LineNumber(1, True),
(Word.of([Reading.of_name("ba")]), ))
]),
unplaced_lines=Transliteration.of_iterable([
TextLine.of_iterable(LineNumber(1, True),
(Word.of([Reading.of_name("ku")]), ))
]),
notes="Updated manuscript.",
), ),
uncertain_fragments=uncertain_fragments,
signs=("KU ABZ075 ABZ207a\\u002F207b\\u0020X\nBA\nKU", ),
)
expect_find_and_update_chapter(
bibliography,
changelog,
CHAPTER_WITHOUT_DOCUMENTS,
updated_chapter,
signs,
sign_repository,
text_repository,
user,
when,
)
manuscripts = (updated_chapter.manuscripts[0], )
assert (corpus.update_manuscripts(CHAPTER.id_, manuscripts,
uncertain_fragments,
user) == updated_chapter)
def test_parse_normalized_akkadain_shift() -> None:
word = "ha"
line = f"1. {word} %n {word} %sux {word}"
expected = Text((TextLine.of_iterable(
LineNumber(1),
(
Word.of((Reading.of_name(word), ), DEFAULT_LANGUAGE),
LanguageShift.normalized_akkadian(),
AkkadianWord.of((ValueToken.of(word), )),
LanguageShift.of("%sux"),
Word.of((Reading.of_name(word), ), Language.SUMERIAN),
),
), ))
assert parse_atf_lark(line).lines == expected.lines
def test_parse_atf_language_shifts(code: str,
expected_language: Language) -> None:
word = "ha-am"
parts = [Reading.of_name("ha"), Joiner.hyphen(), Reading.of_name("am")]
line = f"1. {word} {code} {word} %sb {word}"
expected = Text((TextLine.of_iterable(
LineNumber(1),
(
Word.of(parts, DEFAULT_LANGUAGE),
LanguageShift.of(code),
Word.of(parts, expected_language),
LanguageShift.of("%sb"),
Word.of(parts, Language.AKKADIAN),
),
), ))
assert parse_atf_lark(line).lines == expected.lines
def test_get(client, text_repository):
chapter = ChapterFactory.build(
lines=tuple(),
manuscripts=(
ManuscriptFactory.build(references=tuple()),
ManuscriptFactory.build(colophon=Text(), references=tuple()),
ManuscriptFactory.build(references=tuple()),
),
)
text_repository.create_chapter(chapter)
result = client.simulate_get(create_chapter_url(chapter, "/colophons"))
assert result.status == falcon.HTTP_OK
assert result.json == [{
"siglum": str(manuscript.siglum),
"text": TextSchema().dump(manuscript.colophon),
} for manuscript in chapter.manuscripts
if not manuscript.colophon.is_empty]
def test_query_by_parallel_line_exists(database, fragment_repository):
parallel_number = MuseumNumber.of("K.1")
fragment = FragmentFactory.build(
text=Text(
(
ParallelFragment(
False, parallel_number, True, Labels(), LineNumber(1), True
),
)
)
)
parallel_fragment = FragmentFactory.build(number=parallel_number)
database[COLLECTION].insert_many(
[
FragmentSchema(exclude=["joins"]).dump(fragment),
FragmentSchema(exclude=["joins"]).dump(parallel_fragment),
]
)
assert fragment_repository.query_by_museum_number(fragment.number) == fragment
def test_parse_dividers() -> None:
line, expected_tokens = (
r'1. :? :#! :# ::? :.@v /@19* :"@20@c ;@v@19!',
[
TextLine.of_iterable(
LineNumber(1),
(
Divider.of(":", tuple(), (atf.Flag.UNCERTAIN, )),
Divider.of(":", tuple(),
(atf.Flag.DAMAGE, atf.Flag.CORRECTION)),
Divider.of(":", tuple(), (atf.Flag.DAMAGE, )),
Divider.of("::", tuple(), (atf.Flag.UNCERTAIN, )),
Divider.of(":.", ("@v", ), tuple()),
Divider.of("/", ("@19", ), (atf.Flag.COLLATION, )),
Divider.of(':"', ("@20", "@c"), tuple()),
Divider.of(";", ("@v", "@19"), (atf.Flag.CORRECTION, )),
),
)
],
)
assert parse_atf_lark(line).lines == Text.of_iterable(
expected_tokens).lines
def test_combinations(
qualification,
extent,
scope,
state,
status,
expected_qualification,
expected_extent,
expected_scope,
expected_state,
expected_status,
):
line = " ".join(["$", qualification, extent, scope, state, status])
expected_line = StateDollarLine(
expected_qualification,
expected_extent,
expected_scope,
expected_state,
expected_status,
)
assert parse_atf_lark(line).lines == Text.of_iterable([expected_line
]).lines
def test_of_iterable() -> None:
assert Text.of_iterable(LINES) == Text(LINES, atf.ATF_PARSER_VERSION)
ANOTHER_LEMMATIZED_FRAGMENT = attr.evolve(
TransliteratedFragmentFactory.build(),
text=Text(
(
TextLine(
LineNumber(1),
(
Word.of(
[Logogram.of_name("GI", 6)], unique_lemma=(WordId("ginâ I"),)
),
Word.of([Reading.of_name("ana")], unique_lemma=(WordId("ana II"),)),
Word.of([Reading.of_name("ana")], unique_lemma=(WordId("ana II"),)),
Word.of(
[
Reading.of_name("u", 4),
Joiner.hyphen(),
Reading.of_name("šu"),
],
unique_lemma=(WordId("ūsu I"),),
),
AkkadianWord.of(
[ValueToken.of("ana")], unique_lemma=(WordId("normalized I"),)
),
),
),
)
),
signs="MI DIŠ DIŠ UD ŠU",
)
def test_set_text():
fragment = FragmentFactory.build()
text = Text((ParallelCompositionFactory.build(), ))
updated_fragment = fragment.set_text(text)
assert updated_fragment.text == text
def test_text():
fragment = FragmentFactory.build()
assert fragment.text == Text()
def test_parse_note_line() -> None:
markup = "this is a note @i{italic text}@akk{kur}@sux{kur}"
atf = f"#note: {markup}"
expected_line = NoteLine(parse_markup(markup))
assert parse_atf_lark(atf).lines == Text.of_iterable([expected_line]).lines
def test_parse_state_dollar_line_surface_ambiguity(line, expected_line):
assert parse_atf_lark(line).lines == Text.of_iterable([expected_line
]).lines
def test_parse_state_dollar_line(prefix, parenthesis, line, expected_line):
atf_ = f"{prefix}({line})" if parenthesis else f"{prefix}{line}"
assert parse_atf_lark(atf_).lines == Text.of_iterable([expected_line
]).lines
class Fragment:
number: MuseumNumber
accession: str = ""
cdli_number: str = ""
bm_id_number: str = ""
publication: str = ""
description: str = ""
collection: str = ""
script: str = ""
museum: str = ""
width: Measure = Measure()
length: Measure = Measure()
thickness: Measure = Measure()
joins: Joins = Joins()
record: Record = Record()
folios: Folios = Folios()
text: Text = Text()
signs: str = ""
notes: str = ""
references: Sequence[Reference] = tuple()
uncurated_references: Optional[Sequence[UncuratedReference]] = None
genres: Sequence[Genre] = tuple()
line_to_vec: Tuple[LineToVecEncodings, ...] = tuple()
@property
def is_lowest_join(self) -> bool:
return (self.joins.lowest or self.number) == self.number
def set_references(self, references: Sequence[Reference]) -> "Fragment":
return attr.evolve(self, references=references)
def set_text(self, text: Text) -> "Fragment":
return attr.evolve(self, text=text)
def update_lowest_join_transliteration(
self, transliteration: TransliterationUpdate,
user: User) -> "Fragment":
if transliteration.text.is_empty or self.is_lowest_join:
return self.update_transliteration(transliteration, user)
else:
raise NotLowestJoinError(
"Transliteration must be empty unless fragment is the lowest in join."
)
def update_transliteration(self, transliteration: TransliterationUpdate,
user: User) -> "Fragment":
record = self.record.add_entry(self.text.atf, transliteration.text.atf,
user)
text = self.text.merge(transliteration.text)
return attr.evolve(
self,
text=text,
notes=transliteration.notes,
signs=transliteration.signs,
record=record,
line_to_vec=create_line_to_vec(text.lines),
)
def set_genres(self, genres_new: Sequence[Genre]) -> "Fragment":
return attr.evolve(self, genres=tuple(genres_new))
def update_lemmatization(self, lemmatization: Lemmatization) -> "Fragment":
text = self.text.update_lemmatization(lemmatization)
return attr.evolve(self, text=text)
def get_matching_lines(self, query: TransliterationQuery) -> Lines:
line_numbers = query.match(self.signs)
lines = [line.atf for line in self.text.text_lines]
return tuple(
tuple(lines[numbers[0]:numbers[1] + 1])
for numbers, _ in groupby(line_numbers))