def cls(obj, context=None, continuous=None, space_formatter=lambda s: sub(r'\s', "_", s), indent=None):
"""Display presumed annotation/factor StreamedAnnotationTable in plaintext CLS format"""
if getattr(obj, "cls_valid", None) is not True:
msg = "Exactly one target assay/study metadata field must be present"
_kw = dict(target_columns=getattr(obj, "metadata_columns", []))
raise GeneFabFormatException(msg, **_kw, format="cls")
else:
target_name = ".".join(obj.metadata_columns[0])
target = obj._column_key_dispatcher[target_name]
def content(continuous=continuous, space_formatter=space_formatter):
if (continuous is None) or (continuous is True):
try:
lines = _list_continuous_cls(obj, target, target_name)
except ValueError:
if continuous is True:
msg = "Cannot represent target annotation as continuous"
_kw = dict(target=target_name, format="cls")
raise GeneFabFormatException(msg, **_kw)
else:
continuous = False
if continuous is False:
space_formatter = space_formatter or as_is
lines = _iter_discrete_cls(obj, target, space_formatter)
yield from lines
return content, "text/plain"
def combine_objects(objects, context, limit=None):
"""Combine objects and post-process"""
if len(objects) == 0:
return None
elif len(objects) == 1:
combined = objects[0]
elif all(isinstance(obj, StreamedDataTableWizard) for obj in objects):
combined = StreamedDataTableWizard.concat(objects, axis=1)
else:
raise NotImplementedError("Merging non-table data objects")
if isinstance(combined, StreamedDataTableWizard):
if context.data_columns and (context.format == "gct"):
msg = "GCT format is disabled for arbitrarily subset tables"
_kw = dict(columns="|".join(context.data_columns))
raise GeneFabFormatException(msg, **_kw)
else:
combined.constrain_columns(context=context)
return combined.get(context=context)
elif context.data_columns or context.data_comparisons:
raise GeneFabFileException(
"Column operations on non-table data objects are not supported",
columns=context.data_columns,
comparisons=context.data_comparisons,
)
else:
return combined
def gct(obj, context=None, indent=None, level_formatter="/".join):
"""Display StreamedDataTable in plaintext GCT format, if supported"""
if (not isinstance(obj, StreamedDataTable)) or (len(obj.datatypes) == 0):
msg = "No datatype information associated with retrieved data"
raise GeneFabConfigurationException(msg)
elif len(obj.datatypes) > 1:
msg = "GCT format does not support mixed datatypes"
raise GeneFabFormatException(msg, datatypes=obj.datatypes)
elif not obj.gct_valid:
msg = "GCT format is not valid for given datatype"
raise GeneFabFormatException(msg, datatype=obj.datatypes.pop())
else:
def content():
obj.na_rep = "" # https://www.genepattern.org/file-formats-guide#GCT
yield "#1.2\n{}\t{}\n".format(*obj.shape)
yield "Name\tDescription"
for level in obj.columns:
yield "\t" + level_formatter(level)
yield "\n"
_iter_value_lines = _iter_xsv_chunks(obj.values, "", "\t", 0)
for (index, *_), value_line in zip(obj.index, _iter_value_lines):
yield f"{index}\t{index}\t{value_line}"
return content, "text/plain"
def content(continuous=continuous, space_formatter=space_formatter):
if (continuous is None) or (continuous is True):
try:
lines = _list_continuous_cls(obj, target, target_name)
except ValueError:
if continuous is True:
msg = "Cannot represent target annotation as continuous"
_kw = dict(target=target_name, format="cls")
raise GeneFabFormatException(msg, **_kw)
else:
continuous = False
if continuous is False:
space_formatter = space_formatter or as_is
lines = _iter_discrete_cls(obj, target, space_formatter)
yield from lines
def _call_and_cache():
obj = method(*args, context=context, **kwargs)
try:
obj = obj.schema if (context.schema == "1") else obj
except AttributeError:
msg = "'schema=1' is not valid for requested data"
_type = type(obj).__name__
raise GeneFabFormatException(msg, type=_type)
default_format = getattr(obj, "default_format", "raw")
content, mimetype = self.dispatch_renderer(
obj,
context=context,
default_format=default_format,
)
response_container.update(content, mimetype, obj)
if getattr(obj, "cacheable", None) is True:
if response_cache is not None:
response_cache.put(response_container, context)
def dispatch_renderer(self, obj, context, default_format, indent=None):
"""Render `obj` according to its type and passed kwargs: pass through content and mimetype"""
for types, fmt_to_renderer in TYPE_RENDERERS.items():
if isinstance(obj, types):
if context.format is None:
renderer = fmt_to_renderer[default_format]
elif context.format in fmt_to_renderer:
renderer = fmt_to_renderer[context.format]
else:
raise GeneFabFormatException(
"Requested format not valid for requested data",
type=type(obj).__name__,
format=context.format,
default_format=default_format,
)
return renderer(obj, context, indent=indent)
else:
msg = "Route returned unsupported object"
raise GeneFabConfigurationException(msg, type=type(obj).__name__)
def speed_up_data_schema(get, self, *, context, limit=None, offset=0):
"""If context.schema == '1', replaces underlying query with quick retrieval of just values informative for schema"""
if context.schema != "1":
return get(self, context=context, limit=limit, offset=offset)
elif context.data_columns or context.data_comparisons:
msg = "Data schema does not support column subsetting / comparisons"
sug = "Remove comparisons and/or column, row slicing from query"
raise GeneFabFormatException(msg, suggestion=sug)
else:
from genefab3.db.sql.streamed_tables import (
SQLiteIndexName,
StreamedDataTableWizard_Single,
StreamedDataTableWizard_OuterJoined,
)
GeneFabLogger.info(f"apply_hack(speed_up_data_schema) for {self.name}")
sub_dfs, sub_indices = OrderedDict(), {}
sub_columns, index_name = [], []
def _extend_parts(obj):
for partname, partcols in obj._inverse_column_dispatcher.items():
if isinstance(partcols[0], SQLiteIndexName):
index_name.clear()
index_name.append(partcols[0])
sub_df = get_sub_df(obj, partname, partcols)
else:
sub_df = get_sub_df(obj, partname,
[*index_name, *partcols])
sub_indices[partname] = get_part_index(obj, partname)
sub_dfs[partname] = sub_df
_ocr2f = obj._columns_raw2full
sub_columns.extend(_ocr2f[c] for c in sub_df.columns)
if isinstance(self, StreamedDataTableWizard_Single):
_extend_parts(self)
elif isinstance(self, StreamedDataTableWizard_OuterJoined):
for obj in self.objs:
_extend_parts(obj)
else:
msg = "Schema speedup applied to unsupported object type"
raise GeneFabConfigurationException(msg, type=type(self))
sub_merged = merge_subs(self, sub_dfs, sub_indices)
return StreamedDataTableSub(sub_merged, sub_columns)