code
string | signature
string | docstring
string | loss_without_docstring
float64 | loss_with_docstring
float64 | factor
float64 |
|---|---|---|---|---|---|
if index_names:
df = df.set_index(index_names)
# Remove names from unnamed indexes
index_names = _denormalize_index_names(index_names)
df.index.names = index_names
return df
|
def _update_index_on_df(df, index_names)
|
Helper function to restore index information after collection. Doesn't
use self so we can serialize this.
| 3.720037
| 3.622205
| 1.027009
|
columns = self._schema_rdd.columns
index_names = self._index_names
def fromRecords(records):
if not records:
return []
else:
loaded_df = pd.DataFrame.from_records([records],
columns=columns)
indexed_df = _update_index_on_df(loaded_df, index_names)
return [indexed_df]
return self._schema_rdd.rdd.flatMap(fromRecords)
|
def _rdd(self)
|
Return an RDD of Panda DataFrame objects. This can be expensive
especially if we don't do a narrow transformation after and get it back
to Spark SQL land quickly.
| 4.910696
| 4.401955
| 1.115571
|
index_names = set(_normalize_index_names(self._index_names))
column_names = [col_name for col_name in self._schema_rdd.columns if
col_name not in index_names]
return column_names
|
def _column_names(self)
|
Return the column names
| 4.663956
| 4.242093
| 1.099447
|
source_rdd = self._rdd()
result_rdd = func(source_rdd)
# By default we don't know what the columns & indexes are so we let
# from_rdd_of_dataframes look at the first partition to determine them.
column_idxs = None
if preserves_cols:
index_names = self._index_names
# Remove indexes from the columns
columns = self._schema_rdd.columns[len(self._index_names):]
column_idxs = (columns, index_names)
return self.from_rdd_of_dataframes(
result_rdd, column_idxs=column_idxs)
|
def _evil_apply_with_dataframes(self, func, preserves_cols=False)
|
Convert the underlying SchmeaRDD to an RDD of DataFrames.
apply the provide function and convert the result back.
This is hella slow.
| 5.634604
| 5.099059
| 1.105028
|
columns = self._schema_rdd.columns
df = pd.DataFrame.from_records(
[self._schema_rdd.first()],
columns=self._schema_rdd.columns)
df = _update_index_on_df(df, self._index_names)
return df
|
def _first_as_df(self)
|
Gets the first row as a Panda's DataFrame. Useful for functions like
dtypes & ftypes
| 4.999991
| 4.421137
| 1.130929
|
def frame_to_spark_sql(frame):
return [r.tolist() for r in frame.to_records()]
def frame_to_schema_and_idx_names(frames):
try:
frame = frames.next()
return [(list(frame.columns), list(frame.index.names))]
except StopIteration:
return []
# Store if the RDD was persisted so we don't uncache an
# explicitly cached input.
was_persisted = rdd.is_cached
# If we haven't been supplied with the schema info cache the RDD
# since we are going to eveluate the first partition and then eveluate
# the entire RDD as part of creating a Spark DataFrame.
(schema, index_names) = ([], [])
if not column_idxs:
rdd.cache()
(schema, index_names) = rdd.mapPartitions(
frame_to_schema_and_idx_names).first()
else:
(schema, index_names) = column_idxs
# Add the index_names to the schema.
index_names = _normalize_index_names(index_names)
schema = index_names + schema
ddf = DataFrame.from_schema_rdd(
self.sql_ctx.createDataFrame(rdd.flatMap(frame_to_spark_sql),
schema=schema))
ddf._index_names = index_names
if not was_persisted:
rdd.unpersist()
return ddf
|
def from_rdd_of_dataframes(self, rdd, column_idxs=None)
|
Take an RDD of Panda's DataFrames and return a Dataframe.
If the columns and indexes are already known (e.g. applyMap)
then supplying them with columnsIndexes will skip eveluating
the first partition to determine index info.
| 4.935381
| 4.656383
| 1.059918
|
result = DataFrame(None, sql_ctx)
return result.from_rdd_of_dataframes(rdd)
|
def fromDataFrameRDD(cls, rdd, sql_ctx)
|
Construct a DataFrame from an RDD of DataFrames.
No checking or validation occurs.
| 8.416659
| 5.72635
| 1.469812
|
def transform_rdd(rdd):
return rdd.map(lambda data: data.applymap(f), **kwargs)
return self._evil_apply_with_dataframes(transform_rdd,
preserves_cols=True)
|
def applymap(self, f, **kwargs)
|
Return a new DataFrame by applying a function to each element of each
Panda DataFrame.
| 9.620307
| 10.306971
| 0.933379
|
from sparklingpandas.groupby import GroupBy
return GroupBy(self, by=by, axis=axis, level=level, as_index=as_index,
sort=sort, group_keys=group_keys, squeeze=squeeze)
|
def groupby(self, by=None, axis=0, level=None, as_index=True, sort=True,
group_keys=True, squeeze=False)
|
Returns a groupby on the schema rdd. This returns a GroupBy object.
Note that grouping by a column name will be faster than most other
options due to implementation.
| 2.033337
| 1.97613
| 1.028949
|
local_df = self._schema_rdd.toPandas()
correct_idx_df = _update_index_on_df(local_df, self._index_names)
return correct_idx_df
|
def collect(self)
|
Collect the elements in an DataFrame
and concatenate the partition.
| 10.640224
| 8.971355
| 1.186022
|
assert (not isinstance(columns, basestring)), "columns should be a " \
"list of strs, " \
"not a str!"
assert isinstance(columns, list), "columns should be a list!"
from pyspark.sql import functions as F
functions = [F.min, F.max, F.avg, F.count]
aggs = list(
self._flatmap(lambda column: map(lambda f: f(column), functions),
columns))
return PStats(self.from_schema_rdd(self._schema_rdd.agg(*aggs)))
|
def stats(self, columns)
|
Compute the stats for each column provided in columns.
Parameters
----------
columns : list of str, contains all columns to compute stats on.
| 4.767192
| 4.860748
| 0.980753
|
return self.from_rdd(
self._rdd.map(lambda data: data.applymap(func), **kwargs))
|
def applymap(self, func, **kwargs)
|
Return a new PRDD by applying a function to each element of each
pandas DataFrame.
| 7.356698
| 6.455985
| 1.139516
|
from sparklingpandas.groupby import GroupBy
return GroupBy(self._rdd, *args, **kwargs)
|
def groupby(self, *args, **kwargs)
|
Takes the same parameters as groupby on DataFrame.
Like with groupby on DataFrame disabling sorting will result in an
even larger performance improvement. This returns a Sparkling Pandas
L{GroupBy} object which supports many of the same operations as regular
GroupBy but not all.
| 5.831481
| 3.931386
| 1.483314
|
# The order of the frame order appends is based on the implementation
# of reduce which calls our function with
# f(valueToBeAdded, accumulator) so we do our reduce implementation.
def append_frames(frame_a, frame_b):
return frame_a.append(frame_b)
return self._custom_rdd_reduce(append_frames)
|
def collect(self)
|
Collect the elements in an PRDD and concatenate the partition.
| 17.561262
| 14.992385
| 1.171345
|
def accumulating_iter(iterator):
acc = None
for obj in iterator:
if acc is None:
acc = obj
else:
acc = reduce_func(acc, obj)
if acc is not None:
yield acc
vals = self._rdd.mapPartitions(accumulating_iter).collect()
return reduce(accumulating_iter, vals)
|
def _custom_rdd_reduce(self, reduce_func)
|
Provides a custom RDD reduce which preserves ordering if the RDD has
been sorted. This is useful for us because we need this functionality
as many pandas operations support sorting the results. The standard
reduce in PySpark does not have this property. Note that when PySpark
no longer does partition reduces locally this code will also need to
be updated.
| 2.948543
| 2.842265
| 1.037392
|
def reduce_func(sc1, sc2):
return sc1.merge_pstats(sc2)
return self._rdd.mapPartitions(lambda partition: [
PStatCounter(dataframes=partition, columns=columns)])\
.reduce(reduce_func)
|
def stats(self, columns)
|
Compute the stats for each column provided in columns.
Parameters
----------
columns : list of str, contains all columns to compute stats on.
| 9.696252
| 10.604432
| 0.914358
|
def csv_file(partition_number, files):
# pylint: disable=unexpected-keyword-arg
file_count = 0
for _, contents in files:
# Only skip lines on the first file
if partition_number == 0 and file_count == 0 and _skiprows > 0:
yield pandas.read_csv(
sio(contents), *args,
header=None,
names=mynames,
skiprows=_skiprows,
**kwargs)
else:
file_count += 1
yield pandas.read_csv(
sio(contents), *args,
header=None,
names=mynames,
**kwargs)
def csv_rows(partition_number, rows):
# pylint: disable=unexpected-keyword-arg
in_str = "\n".join(rows)
if partition_number == 0:
return iter([
pandas.read_csv(
sio(in_str), *args, header=None,
names=mynames,
skiprows=_skiprows,
**kwargs)])
else:
# could use .iterows instead?
return iter([pandas.read_csv(sio(in_str), *args, header=None,
names=mynames, **kwargs)])
# If we need to peak at the first partition and determine the column
# names
mynames = None
_skiprows = skiprows
if names:
mynames = names
else:
# In the future we could avoid this expensive call.
first_line = self.spark_ctx.textFile(file_path).first()
frame = pandas.read_csv(sio(first_line), **kwargs)
# pylint sees frame as a tuple despite it being a DataFrame
mynames = list(frame.columns)
_skiprows += 1
# Do the actual load
if use_whole_file:
return self.from_pandas_rdd(
self.spark_ctx.wholeTextFiles(file_path)
.mapPartitionsWithIndex(csv_file))
else:
return self.from_pandas_rdd(
self.spark_ctx.textFile(file_path)
.mapPartitionsWithIndex(csv_rows))
|
def read_csv(self, file_path, use_whole_file=False, names=None, skiprows=0,
*args, **kwargs)
|
Read a CSV file in and parse it into Pandas DataFrames. By default,
the first row from the first partition of that data is parsed and used
as the column names for the data from. If no 'names' param is
provided we parse the first row of the first partition of data and
use it for column names.
Parameters
----------
file_path: string
Path to input. Any valid file path in Spark works here, eg:
'file:///my/path/in/local/file/system' or 'hdfs:/user/juliet/'
use_whole_file: boolean
Whether of not to use the whole file.
names: list of strings, optional
skiprows: integer, optional
indicates how many rows of input to skip. This will
only be applied to the first partition of the data (so if
#skiprows > #row in first partition this will not work). Generally
this shouldn't be an issue for small values of skiprows.
No other value of header is supported.
All additional parameters available in pandas.read_csv() are usable
here.
Returns
-------
A SparklingPandas DataFrame that contains the data from the
specified file.
| 3.289908
| 3.2115
| 1.024415
|
schema_rdd = self.sql_ctx.jsonFile(path, schema, sampling_ratio)
return self.from_spark_rdd(schema_rdd)
|
def jsonFile(self, path, schema=None, sampling_ratio=1.0)
|
Loads a text file storing one JSON object per line as a
L{DataFrame}.
Parameters
----------
path: string
The path of the json files to load. Should be Hadoop style
paths (e.g. hdfs://..., file://... etc.).
schema: StructType, optional
If you know the schema of your input data you can specify it. The
schema is specified using Spark SQL's schema format. If not
specified will sample the json records to determine the schema.
Spark SQL's schema format is documented (somewhat) in the
"Programmatically Specifying the Schema" of the Spark SQL
programming guide at: http://bit.ly/sparkSQLprogrammingGuide
sampling_ratio: int, default=1.0
Percentage of the records to sample when infering schema.
Defaults to all records for safety, but you may be able to set to
a lower ratio if the same fields are present accross records or
your input is of sufficient size.
Returns
-------
A L{DataFrame} of the contents of the json files.
| 5.252553
| 6.841201
| 0.767782
|
def frame_to_rows(frame):
# TODO: Convert to row objects directly?
return [r.tolist() for r in frame.to_records()]
schema = list(local_df.columns)
index_names = list(local_df.index.names)
index_names = _normalize_index_names(index_names)
schema = index_names + schema
rows = self.spark_ctx.parallelize(frame_to_rows(local_df))
sp_df = DataFrame.from_schema_rdd(
self.sql_ctx.createDataFrame(
rows,
schema=schema,
# Look at all the rows, should be ok since coming from
# a local dataset
samplingRatio=1))
sp_df._index_names = index_names
return sp_df
|
def from_pd_data_frame(self, local_df)
|
Make a Sparkling Pandas dataframe from a local Pandas DataFrame.
The intend use is for testing or joining distributed data with local
data.
The types are re-infered, so they may not match.
Parameters
----------
local_df: Pandas DataFrame
The data to turn into a distributed Sparkling Pandas DataFrame.
See http://bit.ly/pandasDataFrame for docs.
Returns
-------
A Sparkling Pandas DataFrame.
| 5.555088
| 5.592452
| 0.993319
|
return DataFrame.from_spark_rdd(self.sql_ctx.sql(query), self.sql_ctx)
|
def sql(self, query)
|
Perform a SQL query and create a L{DataFrame} of the result.
The SQL query is run using Spark SQL. This is not intended for
querying arbitrary databases, but rather querying Spark SQL tables.
Parameters
----------
query: string
The SQL query to pass to Spark SQL to execute.
Returns
-------
Sparkling Pandas DataFrame.
| 7.285645
| 7.331191
| 0.993787
|
return DataFrame.from_spark_rdd(self.sql_ctx.table(table),
self.sql_ctx)
|
def table(self, table)
|
Returns the provided Spark SQL table as a L{DataFrame}
Parameters
----------
table: string
The name of the Spark SQL table to turn into a L{DataFrame}
Returns
-------
Sparkling Pandas DataFrame.
| 8.718139
| 8.362267
| 1.042557
|
return self.from_pd_data_frame(pandas.DataFrame(
elements,
*args,
**kwargs))
|
def DataFrame(self, elements, *args, **kwargs)
|
Create a Sparkling Pandas DataFrame for the provided
elements, following the same API as constructing a Panda's DataFrame.
Note: since elements is local this is only useful for distributing
dataframes which are small enough to fit on a single machine anyways.
Parameters
----------
elements: numpy ndarray (structured or homogeneous), dict, or
Pandas DataFrame.
Input elements to use with the DataFrame.
Additional parameters as defined by L{pandas.DataFrame}.
Returns
-------
Sparkling Pandas DataFrame.
| 5.98359
| 8.262319
| 0.724202
|
def json_file_to_df(files):
for _, contents in files:
yield pandas.read_json(sio(contents), *args, **kwargs)
return self.from_pandas_rdd(self.spark_ctx.wholeTextFiles(file_path)
.mapPartitions(json_file_to_df))
|
def read_json(self, file_path,
*args, **kwargs)
|
Read a json file in and parse it into Pandas DataFrames.
If no names is provided we use the first row for the names.
Currently, it is not possible to skip the first n rows of a file.
Headers are provided in the json file and not specified separately.
Parameters
----------
file_path: string
Path to input. Any valid file path in Spark works here, eg:
'my/path/in/local/file/system' or 'hdfs:/user/juliet/'
Other than skipRows, all additional parameters available in
pandas.read_csv() are usable here.
Returns
-------
A SparklingPandas DataFrame that contains the data from the
specified file.
| 5.874828
| 6.605205
| 0.889424
|
# Load generator module and create instance if we haven't already
if (not self.srcfile):
raise IIIFError(text=("No generator specified"))
if (not self.gen):
try:
(name, ext) = os.path.splitext(self.srcfile)
(pack, mod) = os.path.split(name)
module_name = 'iiif.generators.' + mod
try:
module = sys.modules[module_name]
except KeyError:
self.logger.debug(
"Loading generator module %s" % (module_name))
# Would be nice to use importlib but this is available only
# in python 2.7 and higher
pack = __import__(module_name) # returns iiif package
module = getattr(pack.generators, mod)
self.gen = module.PixelGen()
except ImportError:
raise IIIFError(
text=("Failed to load generator %s" % (str(self.srcfile))))
(self.width, self.height) = self.gen.size
|
def do_first(self)
|
Load generator, set size.
We take the generator module name from self.srcfile so that
this manipulator will work with different generators in a
similar way to how the ordinary generators work with
different images
| 4.575465
| 4.123003
| 1.109741
|
if (x is None):
self.rx = 0
self.ry = 0
self.rw = self.width
self.rh = self.height
else:
self.rx = x
self.ry = y
self.rw = w
self.rh = h
|
def do_region(self, x, y, w, h)
|
Record region.
| 2.034848
| 1.887265
| 1.078199
|
if (w is None):
self.sw = self.rw
self.sh = self.rh
else:
self.sw = w
self.sh = h
# Now we have region and size, generate the image
image = Image.new("RGB", (self.sw, self.sh), self.gen.background_color)
for y in range(0, self.sh):
for x in range(0, self.sw):
ix = int((x * self.rw) // self.sw + self.rx)
iy = int((y * self.rh) // self.sh + self.ry)
color = self.gen.pixel(ix, iy)
if (color is not None):
image.putpixel((x, y), color)
self.image = image
|
def do_size(self, w, h)
|
Record size.
| 2.528916
| 2.482242
| 1.018803
|
headers = {}
headers['Access-control-allow-origin'] = '*'
headers['Content-type'] = 'text/html'
auth = request.authorization
if (auth and auth.username == auth.password):
return self.set_cookie_close_window_response(
"valid-http-basic-login")
else:
headers['WWW-Authenticate'] = (
'Basic realm="HTTP-Basic-Auth at %s (u=p to login)"' %
(self.name))
return make_response("", 401, headers)
|
def login_handler(self, config=None, prefix=None, **args)
|
HTTP Basic login handler.
Respond with 401 and WWW-Authenticate header if there are no
credentials or bad credentials. If there are credentials then
simply check for username equal to password for validity.
| 6.389725
| 5.724393
| 1.116227
|
desc = super(IIIFAuthClickthrough, self).login_service_description()
desc['confirmLabel'] = self.confirm_label
return desc
|
def login_service_description(self)
|
Clickthrough login service description.
The login service description _MUST_ include the token service
description. Additionally, for a clickthroudh loginThe authentication pattern is indicated via the
profile URI which is built using self.auth_pattern.
| 13.914603
| 12.781752
| 1.08863
|
# API parameters
self.identifier = None
self.region = None
self.size = None
self.rotation = None
self.quality = None
self.format = None
self.info = None
# Derived data and flags
self.region_full = False
self.region_square = False
self.region_pct = False
self.region_xywh = None # (x,y,w,h)
self.size_full = False
self.size_max = False # new in 2.1
self.size_pct = None
self.size_bang = None
self.size_wh = None # (w,h)
self.rotation_mirror = False
self.rotation_deg = 0.0
|
def clear(self)
|
Clear all data that might pertain to an individual IIIF URL.
Does not change/reset the baseurl or API version which might be
useful in a sequence of calls.
| 4.294648
| 3.868928
| 1.110036
|
self._api_version = v
if (self._api_version >= '2.0'):
self.default_quality = 'default'
self.allowed_qualities = ['default', 'color', 'bitonal', 'gray']
else: # versions 1.0 and 1.1
self.default_quality = 'native'
self.allowed_qualities = ['native', 'color', 'bitonal', 'grey']
|
def api_version(self, v)
|
Set the api_version and associated configurations.
| 3.440366
| 3.353125
| 1.026018
|
self._setattrs(**params)
path = self.baseurl + self.quote(self.identifier) + "/"
if (self.info):
# info request
path += "info"
format = self.format if self.format else "json"
else:
# region
if self.region:
region = self.region
elif self.region_xywh:
region = "%d,%d,%d,%d" % tuple(self.region_xywh)
else:
region = "full"
# size
if self.size:
size = self.size
elif self.size_wh:
if (self.size_wh[0] is None):
size = ",%d" % (self.size_wh[1])
elif (self.size_wh[1] is None):
size = "%d," % (self.size_wh[0])
else:
size = "%d,%d" % (self.size_wh[0], self.size_wh[1])
elif (self.size_max and self.api_version >= '2.1'):
size = 'max'
else:
size = "full"
# rotation and quality
rotation = self.rotation if self.rotation else "0"
quality = self.quality if self.quality else self.default_quality
# parameterized form
path += self.quote(region) + "/" +\
self.quote(size) + "/" +\
self.quote(rotation) + "/" +\
self.quote(quality)
format = self.format
if (format):
path += "." + format
return(path)
|
def url(self, **params)
|
Build a URL path for image or info request.
An IIIF Image request with parameterized form is assumed unless
the info parameter is specified, in which case an Image Information
request URI is constructred.
| 2.626194
| 2.306211
| 1.138749
|
self.split_url(url)
if (not self.info):
self.parse_parameters()
return(self)
|
def parse_url(self, url)
|
Parse an IIIF API URL path and each component.
Will parse a URL or URL path that accords with either the
parametrized or info request forms. Will raise an
IIIFRequestError on failure. A wrapper for the split_url()
and parse_parameters() methods.
Note that behavior of split_url() depends on whether
self.identifier is set.
| 8.657424
| 8.36296
| 1.035211
|
# clear data first
identifier = self.identifier
self.clear()
# url must start with baseurl if set (including slash)
if (self.baseurl is not None):
(path, num) = re.subn('^' + self.baseurl, '', url, 1)
if (num != 1):
raise IIIFRequestError(
text="Request URL does not start with base URL")
url = path
# Break up by path segments, count to decide format
segs = url.split('/')
if (identifier is not None):
segs.insert(0, identifier)
elif (self.allow_slashes_in_identifier):
segs = self._allow_slashes_in_identifier_munger(segs)
# Now have segments with identifier as first
if (len(segs) > 5):
raise IIIFRequestPathError(
text="Request URL (%s) has too many path segments" % url)
elif (len(segs) == 5):
self.identifier = urlunquote(segs[0])
self.region = urlunquote(segs[1])
self.size = urlunquote(segs[2])
self.rotation = urlunquote(segs[3])
self.quality = self.strip_format(urlunquote(segs[4]))
self.info = False
elif (len(segs) == 2):
self.identifier = urlunquote(segs[0])
info_name = self.strip_format(urlunquote(segs[1]))
if (info_name != "info"):
raise IIIFRequestError(
text="Bad name for Image Information")
if (self.api_version == '1.0'):
if (self.format not in ['json', 'xml']):
raise IIIFRequestError(
text="Invalid format for Image Information (json and xml allowed)")
elif (self.format != 'json'):
raise IIIFRequestError(
text="Invalid format for Image Information (only json allowed)")
self.info = True
elif (len(segs) == 1):
self.identifier = urlunquote(segs[0])
raise IIIFRequestBaseURI()
else:
raise IIIFRequestPathError(
text="Bad number of path segments in request")
return(self)
|
def split_url(self, url)
|
Parse an IIIF API URL path into components.
Will parse a URL or URL path that accords with either the
parametrized or info API forms. Will raise an IIIFRequestError on
failure.
If self.identifier is set then url is assumed not to include the
identifier.
| 3.30303
| 3.192314
| 1.034682
|
m = re.match("(.+)\.([a-zA-Z]\w*)$", str_and_format)
if (m):
# There is a format string at end, chop off and store
str_and_format = m.group(1)
self.format = (m.group(2) if (m.group(2) is not None) else '')
return(str_and_format)
|
def strip_format(self, str_and_format)
|
Look for optional .fmt at end of URI.
The format must start with letter. Note that we want to catch
the case of a dot and no format (format='') which is different
from no dot (format=None)
Sets self.format as side effect, returns possibly modified string
| 3.720695
| 2.905644
| 1.280506
|
self.parse_region()
self.parse_size()
self.parse_rotation()
self.parse_quality()
self.parse_format()
|
def parse_parameters(self)
|
Parse the parameters of an Image Information request.
Will throw an IIIFRequestError on failure, set attributes on
success. Care is taken not to change any of the artibutes
which store path components. All parsed values are stored
in new attributes.
| 3.986018
| 3.117526
| 1.278584
|
self.region_full = False
self.region_square = False
self.region_pct = False
if (self.region is None or self.region == 'full'):
self.region_full = True
return
if (self.api_version >= '2.1' and self.region == 'square'):
self.region_square = True
return
xywh = self.region
pct_match = re.match('pct:(.*)$', self.region)
if (pct_match):
xywh = pct_match.group(1)
self.region_pct = True
# Now whether this was pct: or now, we expect 4 values...
str_values = xywh.split(',', 5)
if (len(str_values) != 4):
raise IIIFRequestError(
code=400, parameter="region",
text="Bad number of values in region specification, "
"must be x,y,w,h but got %d value(s) from '%s'" %
(len(str_values), xywh))
values = []
for str_value in str_values:
# Must be either integer (not pct) or interger/float (pct)
if (pct_match):
try:
# This is rather more permissive that the iiif spec
value = float(str_value)
except ValueError:
raise IIIFRequestError(
parameter="region",
text="Bad floating point value for percentage in "
"region (%s)." % str_value)
if (value > 100.0):
raise IIIFRequestError(
parameter="region",
text="Percentage over value over 100.0 in region "
"(%s)." % str_value)
else:
try:
value = int(str_value)
except ValueError:
raise IIIFRequestError(
parameter="region",
text="Bad integer value in region (%s)." % str_value)
if (value < 0):
raise IIIFRequestError(
parameter="region",
text="Negative values not allowed in region (%s)." %
str_value)
values.append(value)
# Zero size region is w or h are zero (careful that they may be float)
if (values[2] == 0.0 or values[3] == 0.0):
raise IIIFZeroSizeError(
code=400, parameter="region",
text="Zero size region specified (%s))." % xywh)
self.region_xywh = values
|
def parse_region(self)
|
Parse the region component of the path.
/full/ -> self.region_full = True (test this first)
/square/ -> self.region_square = True (test this second)
/x,y,w,h/ -> self.region_xywh = (x,y,w,h)
/pct:x,y,w,h/ -> self.region_xywh and self.region_pct = True
Will throw errors if the parameters are illegal according to the
specification but does not know about and thus cannot do any tests
against any image being manipulated.
| 3.297532
| 2.955945
| 1.115559
|
if (size is not None):
self.size = size
self.size_pct = None
self.size_bang = False
self.size_full = False
self.size_wh = (None, None)
if (self.size is None or self.size == 'full'):
self.size_full = True
return
elif (self.size == 'max' and self.api_version >= '2.1'):
self.size_max = True
return
pct_match = re.match('pct:(.*)$', self.size)
if (pct_match is not None):
pct_str = pct_match.group(1)
try:
self.size_pct = float(pct_str)
except ValueError:
raise IIIFRequestError(
code=400, parameter="size",
text="Percentage size value must be a number, got "
"'%s'." % (pct_str))
# Note that Image API specificaton places no upper limit on
# size so none is implemented here.
if (self.size_pct < 0.0):
raise IIIFRequestError(
code=400, parameter="size",
text="Base size percentage, must be > 0.0, got %f." %
(self.size_pct))
else:
if (self.size[0] == '!'):
# Have "!w,h" form
size_no_bang = self.size[1:]
(mw, mh) = self._parse_w_comma_h(size_no_bang, 'size')
if (mw is None or mh is None):
raise IIIFRequestError(
code=400, parameter="size",
text="Illegal size requested: both w,h must be "
"specified in !w,h requests.")
self.size_wh = (mw, mh)
self.size_bang = True
else:
# Must now be "w,h", "w," or ",h"
self.size_wh = self._parse_w_comma_h(self.size, 'size')
# Sanity check w,h
(w, h) = self.size_wh
if ((w is not None and w <= 0) or
(h is not None and h <= 0)):
raise IIIFZeroSizeError(
code=400, parameter='size',
text="Size parameters request zero size result image.")
|
def parse_size(self, size=None)
|
Parse the size component of the path.
/full/ -> self.size_full = True
/max/ -> self.size_mac = True (2.1 and up)
/w,/ -> self.size_wh = (w,None)
/,h/ -> self.size_wh = (None,h)
/w,h/ -> self.size_wh = (w,h)
/pct:p/ -> self.size_pct = p
/!w,h/ -> self.size_wh = (w,h), self.size_bang = True
Expected use:
(w,h) = iiif.size_to_apply(region_w,region_h)
if (q is None):
# full image
else:
# scale to w by h
Returns (None,None) if no scaling is required.
| 3.326919
| 2.837242
| 1.172589
|
try:
(wstr, hstr) = whstr.split(',', 2)
w = self._parse_non_negative_int(wstr, 'w')
h = self._parse_non_negative_int(hstr, 'h')
except ValueError as e:
raise IIIFRequestError(
code=400, parameter=param,
text="Illegal %s value (%s)." % (param, str(e)))
if (w is None and h is None):
raise IIIFRequestError(
code=400, parameter=param,
text="Must specify at least one of w,h for %s." % (param))
return(w, h)
|
def _parse_w_comma_h(self, whstr, param)
|
Utility to parse "w,h" "w," or ",h" values.
Returns (w,h) where w,h are either None or ineteger. Will
throw a ValueError if there is a problem with one or both.
| 2.637373
| 2.561009
| 1.029818
|
if (istr == ''):
return(None)
try:
i = int(istr)
except ValueError:
raise ValueError("Failed to extract integer value for %s" % (name))
if (i < 0):
raise ValueError("Illegal negative value for %s" % (name))
return(i)
|
def _parse_non_negative_int(self, istr, name)
|
Parse integer from string (istr).
The (name) parameter is used just for IIIFRequestError message
generation to indicate what the error is in.
| 2.72689
| 2.78893
| 0.977755
|
if (rotation is not None):
self.rotation = rotation
self.rotation_deg = 0.0
self.rotation_mirror = False
if (self.rotation is None):
return
# Look for ! prefix first
if (self.rotation[0] == '!'):
self.rotation_mirror = True
self.rotation = self.rotation[1:]
# Interpret value now
try:
self.rotation_deg = float(self.rotation)
except ValueError:
raise IIIFRequestError(
code=400, parameter="rotation",
text="Bad rotation value, must be a number, got '%s'." %
(self.rotation))
if (self.rotation_deg < 0.0 or self.rotation_deg > 360.0):
raise IIIFRequestError(
code=400, parameter="rotation",
text="Illegal rotation value, must be 0 <= rotation "
"<= 360, got %f." % (self.rotation_deg))
elif (self.rotation_deg == 360.0):
# The spec admits 360 as valid, but change to 0
self.rotation_deg = 0.0
|
def parse_rotation(self, rotation=None)
|
Check and interpret rotation.
Uses value of self.rotation as starting point unless rotation
parameter is specified in the call. Sets self.rotation_deg to a
floating point number 0 <= angle < 360. Includes translation of
360 to 0. If there is a prefix bang (!) then self.rotation_mirror
will be set True, otherwise it will be False.
| 3.037367
| 2.569567
| 1.182054
|
if (self.quality is None):
self.quality_val = self.default_quality
elif (self.quality not in self.allowed_qualities):
raise IIIFRequestError(
code=400, parameter="quality",
text="The quality parameter must be '%s', got '%s'." %
("', '".join(self.allowed_qualities), self.quality))
else:
self.quality_val = self.quality
|
def parse_quality(self)
|
Check quality paramater.
Sets self.quality_val based on simple substitution of
'native' for default. Checks for the three valid values
else throws an IIIFRequestError.
| 3.644445
| 2.479827
| 1.469637
|
if (self.format is not None and
not re.match(r'''\w{1,20}$''', self.format)):
raise IIIFRequestError(
parameter='format',
text='Bad format parameter')
|
def parse_format(self)
|
Check format parameter.
All formats values listed in the specification are lowercase
alphanumeric value commonly used as file extensions. To leave
opportunity for extension here just do a limited sanity check
on characters and length.
| 8.649801
| 7.053654
| 1.226287
|
return(self.region_full and
self.size_wh[0] is not None and
self.size_wh[1] is not None and
not self.size_bang and
self.rotation_deg == 0.0 and
self.quality == self.default_quality and
self.format == 'jpg')
|
def is_scaled_full_image(self)
|
True if this request is for a scaled full image.
To be used to determine whether this request should be used
in the set of `sizes` specificed in the Image Information.
| 5.742511
| 5.211882
| 1.101812
|
p = configargparse.ArgParser(description='IIIF Image API Reference Service',
default_config_files=[os.path.join(base_dir, 'iiif_reference_server.cfg')],
formatter_class=configargparse.ArgumentDefaultsHelpFormatter)
add_shared_configs(p, base_dir)
p.add('--scale-factors', default='auto',
help="Set of tile scale factors or 'auto' to calculate for each image "
"such that there are tiles up to the full image")
p.add('--api-versions', default='1.0,1.1,2.0,2.1',
help="Set of API versions to support")
args = p.parse_args()
if (args.debug):
args.verbose = True
elif (args.verbose):
args.quiet = False
# Split list arguments
args.scale_factors = split_comma_argument(args.scale_factors)
args.api_versions = split_comma_argument(args.api_versions)
logging_level = logging.WARNING
if args.verbose:
logging_level = logging.INFO
elif args.quiet:
logging_level = logging.ERROR
logging.basicConfig(format='%(name)s: %(message)s', level=logging_level)
return(args)
|
def get_config(base_dir='')
|
Get config from defaults, config file and/or parse arguments.
Uses configargparse to allow argments to be set from a config file
or via command line arguments.
base_dir - set a specific base directory for file/path defaults.
| 3.308564
| 3.369086
| 0.982036
|
# Create Flask app
app = Flask(__name__)
Flask.secret_key = "SECRET_HERE"
app.debug = cfg.debug
# Install request handlers
client_prefixes = dict()
for api_version in cfg.api_versions:
handler_config = Config(cfg)
handler_config.api_version = api_version
handler_config.klass_name = 'pil'
handler_config.auth_type = 'none'
# Set same prefix on local server as expected on iiif.io
handler_config.prefix = "api/image/%s/example/reference" % (api_version)
handler_config.client_prefix = handler_config.prefix
add_handler(app, handler_config)
return app
|
def create_reference_server_flask_app(cfg)
|
Create referece server Flask application with one or more IIIF handlers.
| 6.887955
| 6.312732
| 1.091121
|
if (size is None):
size = self.sz
# Have we go to the smallest element?
if (size <= 3):
if (_middle(x, y)):
return (None)
else:
return (0, 0, 0)
divisor = size // 3
if (_middle(x // divisor, y // divisor)):
return None
return self.pixel(x % divisor, y % divisor, divisor)
|
def pixel(self, x, y, size=None)
|
Return color for a pixel.
| 6.219067
| 6.046792
| 1.02849
|
id = ''
if (self.server_and_prefix is not None and
self.server_and_prefix != ''):
id += self.server_and_prefix + '/'
if (self.identifier is not None):
id += self.identifier
return id
|
def id(self)
|
id property based on server_and_prefix and identifier.
| 3.734461
| 2.037337
| 1.833011
|
i = value.rfind('/')
if (i > 0):
self.server_and_prefix = value[:i]
self.identifier = value[(i + 1):]
elif (i == 0):
self.server_and_prefix = ''
self.identifier = value[(i + 1):]
else:
self.server_and_prefix = ''
self.identifier = value
|
def id(self, value)
|
Split into server_and_prefix and identifier.
| 2.58285
| 1.736085
| 1.487744
|
if (api_version is None):
api_version = self.api_version
if (api_version not in CONF):
raise IIIFInfoError("Unknown API version %s" % (api_version))
self.params = CONF[api_version]['params']
self.array_params = CONF[api_version]['array_params']
self.complex_params = CONF[api_version]['complex_params']
for a in ('context', 'compliance_prefix', 'compliance_suffix',
'protocol', 'required_params'):
if (a in CONF[api_version]):
self._setattr(a, CONF[api_version][a])
|
def set_version_info(self, api_version=None)
|
Set up normal values for given api_version.
Will use current value of self.api_version if a version number
is not specified in the call. Will raise an IIIFInfoError
| 3.794533
| 3.327275
| 1.140433
|
if (self.api_version < '2.0'):
self.profile = value
else:
try:
self.profile[0] = value
except AttributeError:
# handle case where profile not initialized as array
self.profile = [value]
|
def compliance(self, value)
|
Set the compliance profile URI.
| 5.931497
| 5.235283
| 1.132985
|
m = re.match(
self.compliance_prefix +
r'(\d)' +
self.compliance_suffix +
r'$',
self.compliance)
if (m):
return int(m.group(1))
raise IIIFInfoError(
"Bad compliance profile URI, failed to extract level number")
|
def level(self)
|
Extract level number from compliance profile URI.
Returns integer level number or raises IIIFInfoError
| 7.166693
| 3.861184
| 1.856087
|
self.compliance = self.compliance_prefix + \
("%d" % value) + self.compliance_suffix
|
def level(self, value)
|
Build profile URI from level.
Level should be an integer 0,1,2
| 10.372285
| 11.07996
| 0.93613
|
if (self.service is None):
self.service = service
elif (isinstance(self.service, dict)):
self.service = [self.service, service]
else:
self.service.append(service)
|
def add_service(self, service)
|
Add a service description.
Handles transition from self.service=None, self.service=dict for a
single service, and then self.service=[dict,dict,...] for multiple
| 2.601396
| 2.323399
| 1.119651
|
errors = []
for param in self.required_params:
if (not hasattr(self, param) or getattr(self, param) is None):
errors.append("missing %s parameter" % (param))
if (len(errors) > 0):
raise IIIFInfoError("Bad data for info.json: " + ", ".join(errors))
return True
|
def validate(self)
|
Validate this object as Image API data.
Raise IIIFInfoError with helpful message if not valid.
| 4.000926
| 2.88183
| 1.388328
|
if (validate):
self.validate()
json_dict = {}
if (self.api_version > '1.0'):
json_dict['@context'] = self.context
params_to_write = set(self.params)
params_to_write.discard('identifier')
if (self.identifier):
if (self.api_version == '1.0'):
json_dict['identifier'] = self.identifier # local id
else:
json_dict['@id'] = self.id # URI
params_to_write.discard('profile')
if (self.compliance):
if (self.api_version < '2.0'):
json_dict['profile'] = self.compliance
else:
# FIXME - need to support extra profile features
json_dict['profile'] = [self.compliance]
d = {}
if (self.formats is not None):
d['formats'] = self.formats
if (self.qualities is not None):
d['qualities'] = self.qualities
if (self.supports is not None):
d['supports'] = self.supports
if (len(d) > 0):
json_dict['profile'].append(d)
params_to_write.discard('formats')
params_to_write.discard('qualities')
params_to_write.discard('supports')
for param in params_to_write:
if (hasattr(self, param) and
getattr(self, param) is not None):
json_dict[param] = getattr(self, param)
return(json.dumps(json_dict, sort_keys=True, indent=2))
|
def as_json(self, validate=True)
|
Return JSON serialization.
Will raise IIIFInfoError if insufficient parameters are present to
have a valid info.json response (unless validate is False).
| 2.467594
| 2.300153
| 1.072796
|
j = json.load(fh)
#
# @context and API version
self.context = None
if (api_version == '1.0'):
# v1.0 did not have a @context so we simply take the version
# passed in
self.api_version = api_version
elif ('@context' in j):
# determine API version from context
self.context = j['@context']
api_version_read = None
for v in CONF:
if (v > '1.0' and self.context == CONF[v]['context']):
api_version_read = v
break
if (api_version_read is None):
raise IIIFInfoError(
"Unknown @context, cannot determine API version (%s)" %
(self.context))
else:
if (api_version is not None and
api_version != api_version_read):
raise IIIFInfoError(
"Expected API version '%s' but got @context for API version '%s'" %
(api_version, api_version_read))
else:
self.api_version = api_version_read
else: # no @context and not 1.0
if (api_version is None):
raise IIIFInfoError("No @context (and no default given)")
self.api_version = api_version
self.set_version_info()
#
# @id or identifier
if (self.api_version == '1.0'):
if ('identifier' in j):
self.id = j['identifier']
else:
raise IIIFInfoError("Missing identifier in info.json")
else:
if ('@id' in j):
self.id = j['@id']
else:
raise IIIFInfoError("Missing @id in info.json")
#
# other params
for param in self.params:
if (param == 'identifier'):
continue # dealt with above
if (param in j):
if (param in self.complex_params):
# use function ref in complex_params to parse, optional
# dst to map to a different property name
self._setattr(param, self.complex_params[
param](self, j[param]))
else:
self._setattr(param, j[param])
return True
|
def read(self, fh, api_version=None)
|
Read info.json from file like object.
Parameters:
fh -- file like object supporting fh.read()
api_version -- IIIF Image API version expected
If api_version is set then the parsing will assume this API version,
else the version will be determined from the incoming data. NOTE that
the value of self.api_version is NOT used in this routine.
If an api_version is specified and there is a @context specified then
an IIIFInfoError will be raised unless these match. If no known
@context is present and no api_version set then an IIIFInfoError
will be raised.
| 3.323804
| 3.063668
| 1.08491
|
if (pixels):
Image.MAX_IMAGE_PIXELS = pixels
Image.warnings.simplefilter(
'error', Image.DecompressionBombWarning)
|
def set_max_image_pixels(self, pixels)
|
Set PIL limit on pixel size of images to load if non-zero.
WARNING: This is a global setting in PIL, it is
not local to this manipulator instance!
Setting a value here will not only set the given limit but
also convert the PIL "DecompressionBombWarning" into an
error. Thus setting a moderate limit sets a hard limit on
image size loaded, setting a very large limit will have the
effect of disabling the warning.
| 5.533839
| 4.245174
| 1.30356
|
self.logger.debug("do_first: src=%s" % (self.srcfile))
try:
self.image = Image.open(self.srcfile)
except Image.DecompressionBombWarning as e:
# This exception will be raised only if PIL has been
# configured to raise an error in the case of images
# that exceeed Image.MAX_IMAGE_PIXELS, with
# Image.warnings.simplefilter('error', Image.DecompressionBombWarning)
raise IIIFError(text=("Image size limit exceeded (PIL: %s)" % (str(e))))
except Exception as e:
raise IIIFError(text=("Failed to read image (PIL: %s)" % (str(e))))
(self.width, self.height) = self.image.size
|
def do_first(self)
|
Create PIL object from input image file.
Image location must be in self.srcfile. Will result in
self.width and self.height being set to the image dimensions.
Will raise an IIIFError on failure to load the image
| 4.066747
| 3.365524
| 1.208355
|
if (x is None):
self.logger.debug("region: full (nop)")
else:
self.logger.debug("region: (%d,%d,%d,%d)" % (x, y, w, h))
self.image = self.image.crop((x, y, x + w, y + h))
self.width = w
self.height = h
|
def do_region(self, x, y, w, h)
|
Apply region selection.
| 2.968461
| 2.974004
| 0.998136
|
if (w is None):
self.logger.debug("size: no scaling (nop)")
else:
self.logger.debug("size: scaling to (%d,%d)" % (w, h))
self.image = self.image.resize((w, h))
self.width = w
self.height = h
|
def do_size(self, w, h)
|
Apply size scaling.
| 3.516372
| 3.137564
| 1.120733
|
if (not mirror and rot == 0.0):
self.logger.debug("rotation: no rotation (nop)")
else:
# FIXME - with PIL one can use the transpose() method to do 90deg
# FIXME - rotations as well as mirroring. This would be more efficient
# FIXME - for these cases than mirror _then_ rotate.
if (mirror):
self.logger.debug("rotation: mirror (about vertical axis)")
self.image = self.image.transpose(Image.FLIP_LEFT_RIGHT)
if (rot != 0.0):
self.logger.debug("rotation: by %f degrees clockwise" % (rot))
self.image = self.image.rotate(-rot, expand=True)
|
def do_rotation(self, mirror, rot)
|
Apply rotation and/or mirroring.
| 5.007245
| 4.808348
| 1.041365
|
if (quality == 'grey' or quality == 'gray'):
# Checking for 1.1 gray or 20.0 grey elsewhere
self.logger.debug("quality: converting to gray")
self.image = self.image.convert('L')
elif (quality == 'bitonal'):
self.logger.debug("quality: converting to bitonal")
self.image = self.image.convert('1')
else: # color or default/native (which we take as color)
# Deal first with conversions from I;16* formats which Pillow
# appears not to handle properly, resulting in mostly white images
# if we convert directly. See:
# <http://stackoverflow.com/questions/7247371/python-and-16-bit-tiff>
if (self.image.mode.startswith('I;16')):
self.logger.debug("quality: fudged conversion from mode %s to I"
% (self.image.mode))
self.image = self.image.convert('I')
self.image = self.image.point(lambda i: i * (1.0 / 256.0))
if (self.image.mode not in ('1', 'L', 'RGB', 'RGBA')):
# Need to convert from palette etc. in order to write out
self.logger.debug("quality: converting from mode %s to RGB"
% (self.image.mode))
self.image = self.image.convert('RGB')
else:
self.logger.debug("quality: quality (nop)")
|
def do_quality(self, quality)
|
Apply value of quality parameter.
For PIL docs see
<http://pillow.readthedocs.org/en/latest/reference/Image.html#PIL.Image.Image.convert>
| 4.838577
| 4.639401
| 1.042931
|
fmt = ('jpg' if (format is None) else format)
if (fmt == 'png'):
self.logger.debug("format: png")
self.mime_type = "image/png"
self.output_format = fmt
format = 'png'
elif (fmt == 'jpg'):
self.logger.debug("format: jpg")
self.mime_type = "image/jpeg"
self.output_format = fmt
format = 'jpeg'
elif (fmt == 'webp'):
self.logger.debug("format: webp")
self.mime_type = "image/webp"
self.output_format = fmt
format = 'webp'
else:
raise IIIFError(code=415, parameter='format',
text="Unsupported output file format (%s), only png,jpg,webp are supported." % (fmt))
if (self.outfile is None):
# Create temp
f = tempfile.NamedTemporaryFile(delete=False)
self.outfile = f.name
self.outtmp = f.name
self.image.save(f, format=format)
else:
# Save to specified location
self.image.save(self.outfile, format=format)
|
def do_format(self, format)
|
Apply format selection.
Assume that for tiling applications we want jpg so return
that unless an explicit format is requested.
| 2.617742
| 2.523846
| 1.037203
|
if self.image:
try:
self.image.close()
except Exception:
pass
if (self.outtmp is not None):
try:
os.unlink(self.outtmp)
except OSError as e:
self.logger.warning("Failed to cleanup tmp output file %s"
% (self.outtmp))
|
def cleanup(self)
|
Cleanup: ensure image closed and remove temporary output file.
| 3.543516
| 2.958002
| 1.197942
|
if (size is None):
size = self.sz
# Have we go to the smallest element?
if (size <= 3):
if (_num(x, y) % 2):
return (red, 0, 0)
else:
return None
divisor = size // 3
n = _num(x // divisor, y // divisor)
if (n == 5):
# Middle square further divided
return self.pixel(x % divisor, y % divisor,
divisor, min(red + 25, 255))
elif (n % 2):
return (red, 0, 0)
else:
return None
|
def pixel(self, x, y, size=None, red=0)
|
Return color for a pixel.
The paremeter size is used to recursively follow down to smaller
and smaller middle squares (number 5). The paremeter red is used
to shade from black to red going toward the middle of the figure.
| 4.861413
| 4.302616
| 1.129874
|
html = "<html>\n<head><title>%s</title></head>\n<body>\n" % (title)
html += "<h1>%s</h1>\n" % (title)
html += body
html += "</body>\n</html>\n"
return html
|
def html_page(title="Page Title", body="")
|
Create HTML page as string.
| 1.860575
| 1.769295
| 1.051591
|
title = "IIIF Test Server on %s" % (config.host)
body = "<ul>\n"
for prefix in sorted(config.prefixes.keys()):
body += '<li><a href="/%s">%s</a></li>\n' % (prefix, prefix)
body += "</ul>\n"
return html_page(title, body)
|
def top_level_index_page(config)
|
HTML top-level index page which provides a link to each handler.
| 3.40313
| 3.314818
| 1.026641
|
ids = []
if (config.klass_name == 'gen'):
for generator in os.listdir(config.generator_dir):
if (generator == '__init__.py'):
continue
(gid, ext) = os.path.splitext(generator)
if (ext == '.py' and
os.path.isfile(os.path.join(config.generator_dir, generator))):
ids.append(gid)
else:
for image_file in os.listdir(config.image_dir):
(iid, ext) = os.path.splitext(image_file)
if (ext in ['.jpg', '.png', '.tif'] and
os.path.isfile(os.path.join(config.image_dir, image_file))):
ids.append(iid)
return ids
|
def identifiers(config)
|
Show list of identifiers for this prefix.
Handles both the case of local file based identifiers and
also image generators.
Arguments:
config - configuration object in which:
config.klass_name - 'gen' if a generator function
config.generator_dir - directory for generator code
config.image_dir - directory for images
Returns:
ids - a list of ids
| 2.399394
| 1.861269
| 1.289117
|
title = "IIIF Image API services under %s" % (config.client_prefix)
# details of this prefix handler
body = '<p>\n'
body += 'host = %s<br/>\n' % (config.host)
body += 'api_version = %s<br/>\n' % (config.api_version)
body += 'manipulator = %s<br/>\n' % (config.klass_name)
body += 'auth_type = %s\n</p>\n' % (config.auth_type)
# table of identifiers and example requests
ids = identifiers(config)
api_version = config.api_version
default = 'native' if api_version < '2.0' else 'default'
body += '<table border="1">\n<tr><th align="left">Image identifier</th>'
body += '<th> </th><th>full</th>'
if (config.klass_name != 'dummy'):
body += '<th>256,256</th>'
body += '<th>30deg</th>'
if (config.include_osd):
body += '<th> </th>'
body += "</tr>\n"
for identifier in sorted(ids):
base = urljoin('/', config.client_prefix + '/' + identifier)
body += '<tr><th align="left">%s</th>' % (identifier)
info = base + "/info.json"
body += '<td><a href="%s">%s</a></td>' % (info, 'info')
suffix = "full/full/0/%s" % (default)
body += '<td><a href="%s">%s</a></td>' % (base + '/' + suffix, suffix)
if (config.klass_name != 'dummy'):
suffix = "full/256,256/0/%s" % (default)
body += '<td><a href="%s">%s</a></td>' % (base + '/' + suffix, suffix)
suffix = "full/100,/30/%s" % (default)
body += '<td><a href="%s">%s</a></td>' % (base + '/' + suffix, suffix)
if (config.include_osd):
body += '<td><a href="%s/osd.html">OSD</a></td>' % (base)
body += "</tr>\n"
body += "</table<\n"
return html_page(title, body)
|
def prefix_index_page(config)
|
HTML index page for a specific prefix.
The prefix seen by the client is obtained from config.client_prefix
as opposed to the local server prefix in config.prefix. Also uses
the identifiers(config) function to get identifiers available.
Arguments:
config - configuration object in which:
config.client_prefix - URI path prefix seen by client
config.host - URI host seen by client
config.api_version - string for api_version
config.manipulator - string manipulator type
config.auth_type - string for auth type
config.include_osd - whether OSD is included
| 2.973941
| 2.724424
| 1.091585
|
uri = "http://" + host
if (port != 80):
uri += ':' + str(port)
if (prefix):
uri += '/' + prefix
return uri
|
def host_port_prefix(host, port, prefix)
|
Return URI composed of scheme, server, port, and prefix.
| 3.254084
| 2.485016
| 1.309482
|
template_dir = os.path.join(os.path.dirname(__file__), 'templates')
with open(os.path.join(template_dir, 'testserver_osd.html'), 'r') as f:
template = f.read()
d = dict(prefix=prefix,
identifier=identifier,
api_version=config.api_version,
osd_version='2.0.0',
osd_uri='/openseadragon200/openseadragon.min.js',
osd_images_prefix='/openseadragon200/images',
osd_height=500,
osd_width=500,
info_json_uri='info.json')
return make_response(Template(template).safe_substitute(d))
|
def osd_page_handler(config=None, identifier=None, prefix=None, **args)
|
Flask handler to produce HTML response for OpenSeadragon view of identifier.
Arguments:
config - Config object for this IIIF handler
identifier - identifier of image/generator
prefix - path prefix
**args - other aguments ignored
| 3.376614
| 3.171299
| 1.064742
|
if (not auth or degraded_request(identifier) or auth.info_authz()):
# go ahead with request as made
if (auth):
logging.debug("Authorized for image %s" % identifier)
i = IIIFHandler(prefix, identifier, config, klass, auth)
try:
return i.image_information_response()
except IIIFError as e:
return i.error_response(e)
elif (auth.info_authn()):
# authn but not authz -> 401
abort(401)
else:
# redirect to degraded
response = redirect(host_port_prefix(
config.host, config.port, prefix) + '/' + identifier + '-deg/info.json')
response.headers['Access-control-allow-origin'] = '*'
return response
|
def iiif_info_handler(prefix=None, identifier=None,
config=None, klass=None, auth=None, **args)
|
Handler for IIIF Image Information requests.
| 6.669762
| 6.519824
| 1.022997
|
if (not auth or degraded_request(identifier) or auth.image_authz()):
# serve image
if (auth):
logging.debug("Authorized for image %s" % identifier)
i = IIIFHandler(prefix, identifier, config, klass, auth)
try:
return i.image_request_response(path)
except IIIFError as e:
return i.error_response(e)
else:
# redirect to degraded (for not authz and for authn but not authz too)
degraded_uri = host_port_prefix(
config.host, config.port, prefix) + '/' + identifier + '-deg/' + path
logging.info("Redirection to degraded: %s" % degraded_uri)
response = redirect(degraded_uri)
response.headers['Access-control-allow-origin'] = '*'
return response
|
def iiif_image_handler(prefix=None, identifier=None,
path=None, config=None, klass=None, auth=None, **args)
|
Handler for IIIF Image Requests.
Behaviour for case of a non-authn or non-authz case is to
return 403.
| 6.07547
| 5.618177
| 1.081395
|
result = []
for media_range in accept.split(","):
parts = media_range.split(";")
media_type = parts.pop(0).strip()
media_params = []
q = 1.0
for part in parts:
(key, value) = part.lstrip().split("=", 1)
if key == "q":
q = float(value)
else:
media_params.append((key, value))
result.append((media_type, tuple(media_params), q))
result.sort(key=lambda x: -x[2])
return result
|
def parse_accept_header(accept)
|
Parse an HTTP Accept header.
Parses *accept*, returning a list with pairs of
(media_type, q_value), ordered by q values.
Adapted from <https://djangosnippets.org/snippets/1042/>
| 2.008865
| 2.03261
| 0.988318
|
try:
(auth_type, auth_info) = value.split(' ', 1)
auth_type = auth_type.lower()
except ValueError:
return
if (auth_type == 'basic'):
try:
decoded = base64.b64decode(auth_info).decode(
'utf-8') # b64decode gives bytes in python3
(username, password) = decoded.split(':', 1)
except (ValueError, TypeError): # py3, py2
return
return {'type': 'basic', 'username': username, 'password': password}
elif (auth_type == 'digest'):
try:
auth_map = parse_keqv_list(parse_http_list(auth_info))
except ValueError:
return
logging.debug(auth_map)
for key in 'username', 'realm', 'nonce', 'uri', 'response':
if key not in auth_map:
return
if 'qop' in auth_map and ('nc' not in auth_map or 'cnonce' not in auth_map):
return
auth_map['type'] = 'digest'
return auth_map
else:
# unknown auth type
return
|
def parse_authorization_header(value)
|
Parse the Authenticate header.
Returns nothing on failure, opts hash on success with type='basic' or 'digest'
and other params.
<http://nullege.com/codes/search/werkzeug.http.parse_authorization_header>
<http://stackoverflow.com/questions/1349367/parse-an-http-request-authorization-header-with-python>
<http://bugs.python.org/file34041/0001-Add-an-authorization-header-to-the-initial-request.patch>
| 2.300683
| 2.232641
| 1.030476
|
for result in parse_accept_header(accept):
mime_type = result[0]
if (mime_type in supported):
return mime_type
return None
|
def do_conneg(accept, supported)
|
Parse accept header and look for preferred type in supported list.
Arguments:
accept - HTTP Accept header
supported - list of MIME type supported by the server
Returns:
supported MIME type with highest q value in request, else None.
FIXME - Should replace this with negotiator2
| 3.914128
| 3.451855
| 1.13392
|
base = urljoin('/', prefix + '/') # Must end in slash
app.add_url_rule(base + 'login', prefix + 'login_handler',
auth.login_handler, defaults=params)
app.add_url_rule(base + 'logout', prefix + 'logout_handler',
auth.logout_handler, defaults=params)
if (auth.client_id_handler):
app.add_url_rule(base + 'client', prefix + 'client_id_handler',
auth.client_id_handler, defaults=params)
app.add_url_rule(base + 'token', prefix + 'access_token_handler',
auth.access_token_handler, defaults=params)
if (auth.home_handler):
app.add_url_rule(base + 'home', prefix + 'home_handler',
auth.home_handler, defaults=params)
|
def setup_auth_paths(app, auth, prefix, params)
|
Add URL rules for auth paths.
| 2.024377
| 1.979847
| 1.022492
|
prefix = "%s_%s" % (api_version, manipulator)
if (auth_type and auth_type != 'none'):
prefix += '_' + auth_type
return prefix
|
def make_prefix(api_version, manipulator, auth_type)
|
Make prefix string based on configuration parameters.
| 2.800062
| 2.679949
| 1.044819
|
terms = []
for term in comma_sep_str.split(','):
if term:
terms.append(term)
return terms
|
def split_comma_argument(comma_sep_str)
|
Split a comma separated option into a list.
| 3.072287
| 2.992942
| 1.02651
|
p.add('--host', default='localhost',
help="Service host")
p.add('--port', '-p', type=int, default=8000,
help="Service port")
p.add('--app-host', default=None,
help="Local application host for reverse proxy deployment, "
"as opposed to service --host (must also specify --app-port)")
p.add('--app-port', type=int, default=None,
help="Local application port for reverse proxy deployment, "
"as opposed to service --port (must also specify --app-host)")
p.add('--image-dir', '-d', default=os.path.join(base_dir, 'testimages'),
help="Image file directory")
p.add('--generator-dir', default=os.path.join(base_dir, 'iiif/generators'),
help="Generator directory for manipulator='gen'")
p.add('--tile-height', type=int, default=512,
help="Tile height")
p.add('--tile-width', type=int, default=512,
help="Tile width")
p.add('--gauth-client-secret', default=os.path.join(base_dir, 'client_secret.json'),
help="Name of file with Google auth client secret")
p.add('--include-osd', action='store_true',
help="Include a page with OpenSeadragon for each source")
p.add('--access-cookie-lifetime', type=int, default=3600,
help="Set access cookie lifetime for authenticated access in seconds")
p.add('--access-token-lifetime', type=int, default=10,
help="Set access token lifetime for authenticated access in seconds")
p.add('--config', is_config_file=True, default=None,
help='Read config from given file path')
p.add('--debug', action='store_true',
help="Set debug mode for web application. INSECURE!")
p.add('--verbose', '-v', action='store_true',
help="Be verbose")
p.add('--quiet', '-q', action='store_true',
help="Minimal output only")
|
def add_shared_configs(p, base_dir='')
|
Add configargparser/argparse configs for shared argument.
Arguments:
p - configargparse.ArgParser object
base_dir - base directory for file/path defaults.
| 2.869283
| 2.882346
| 0.995468
|
if not basedir:
basedir = os.path.join(os.path.dirname(__file__), 'third_party')
return send_from_directory(os.path.join(basedir, prefix), filename)
|
def serve_static(filename=None, prefix='', basedir=None)
|
Handler for static files: server filename in basedir/prefix.
If not specified then basedir defaults to the local third_party
directory.
| 2.468393
| 2.211659
| 1.116082
|
pidfile = os.path.basename(sys.argv[0])[:-3] + '.pid' # strip .py, add .pid
with open(pidfile, 'w') as fh:
fh.write("%d\n" % os.getpid())
fh.close()
|
def write_pid_file()
|
Write a file with the PID of this server instance.
Call when setting up a command line testserver.
| 2.736414
| 2.826676
| 0.968068
|
# Set up app_host and app_port in case that we are running
# under reverse proxy setup, otherwise they default to
# config.host and config.port.
if (cfg.app_host and cfg.app_port):
logging.warning("Reverse proxy for service at http://%s:%d/ ..." % (cfg.host, cfg.port))
app.wsgi_app = ReverseProxied(app.wsgi_app, cfg.host)
elif (cfg.app_host or cfg.app_port):
logging.critical("Must specify both app-host and app-port for reverse proxy configuration, aborting")
sys.exit(1)
else:
cfg.app_host = cfg.host
cfg.app_port = cfg.port
logging.warning("Setup server on http://%s:%d/ ..." % (cfg.app_host, cfg.app_port))
return(app)
|
def setup_app(app, cfg)
|
Setup Flask app and handle reverse proxy setup if configured.
Arguments:
app - Flask application
cfg - configuration data
| 3.422948
| 3.399592
| 1.00687
|
return(host_port_prefix(self.config.host, self.config.port, self.prefix))
|
def server_and_prefix(self)
|
Server and prefix from config.
| 8.266179
| 6.694118
| 1.234842
|
mime_type = "application/json"
if (self.api_version >= '1.1' and 'Accept' in request.headers):
mime_type = do_conneg(request.headers['Accept'], [
'application/ld+json']) or mime_type
return mime_type
|
def json_mime_type(self)
|
Return the MIME type for a JSON response.
For version 2.0+ the server must return json-ld MIME type if that
format is requested. Implement for 1.1 also.
http://iiif.io/api/image/2.1/#information-request
| 5.070995
| 4.34782
| 1.166331
|
file = None
if (self.config.klass_name == 'gen'):
for ext in ['.py']:
file = os.path.join(
self.config.generator_dir, self.identifier + ext)
if (os.path.isfile(file)):
return file
else:
for ext in ['.jpg', '.png', '.tif']:
file = os.path.join(self.config.image_dir,
self.identifier + ext)
if (os.path.isfile(file)):
return file
# failed, show list of available identifiers as error
available = "\n ".join(identifiers(self.config))
raise IIIFError(code=404, parameter="identifier",
text="Image resource '" + self.identifier + "' not found. Local resources available:" + available + "\n")
|
def file(self)
|
Filename property for the source image for the current identifier.
| 4.630033
| 4.26775
| 1.084889
|
if (self.manipulator.compliance_uri is not None):
self.headers['Link'] = '<' + \
self.manipulator.compliance_uri + '>;rel="profile"'
|
def add_compliance_header(self)
|
Add IIIF Compliance level header to response.
| 6.78472
| 4.889988
| 1.387472
|
if headers:
for header in headers:
self.headers[header] = headers[header]
return make_response(content, code, self.headers)
|
def make_response(self, content, code=200, headers=None)
|
Wrapper around Flask.make_response which also adds any local headers.
| 2.861079
| 2.512166
| 1.138889
|
dr = degraded_request(self.identifier)
if (dr):
self.logger.info("image_information: degraded %s -> %s" %
(self.identifier, dr))
self.degraded = self.identifier
self.identifier = dr
else:
self.logger.info("image_information: %s" % (self.identifier))
# get size
self.manipulator.srcfile = self.file
self.manipulator.do_first()
# most of info.json comes from config, a few things specific to image
info = {'tile_height': self.config.tile_height,
'tile_width': self.config.tile_width,
'scale_factors': self.config.scale_factors
}
# calculate scale factors if not hard-coded
if ('auto' in self.config.scale_factors):
info['scale_factors'] = self.manipulator.scale_factors(
self.config.tile_width, self.config.tile_height)
i = IIIFInfo(conf=info, api_version=self.api_version)
i.server_and_prefix = self.server_and_prefix
i.identifier = self.iiif.identifier
i.width = self.manipulator.width
i.height = self.manipulator.height
if (self.api_version >= '2.0'):
# FIXME - should come from manipulator
i.qualities = ["default", "color", "gray"]
else:
# FIXME - should come from manipulator
i.qualities = ["native", "color", "gray"]
i.formats = ["jpg", "png"] # FIXME - should come from manipulator
if (self.auth):
self.auth.add_services(i)
return self.make_response(i.as_json(),
headers={"Content-Type": self.json_mime_type})
|
def image_information_response(self)
|
Parse image information request and create response.
| 4.169374
| 4.148865
| 1.004943
|
# Parse the request in path
if (len(path) > 1024):
raise IIIFError(code=414,
text="URI Too Long: Max 1024 chars, got %d\n" % len(path))
try:
self.iiif.identifier = self.identifier
self.iiif.parse_url(path)
except IIIFRequestPathError as e:
# Reraise as IIIFError with code=404 because we can't tell
# whether there was an encoded slash in the identifier or
# whether there was a bad number of path segments.
raise IIIFError(code=404, text=e.text)
except IIIFError as e:
# Pass through
raise e
except Exception as e:
# Something completely unexpected => 500
raise IIIFError(code=500,
text="Internal Server Error: unexpected exception parsing request (" + str(e) + ")")
dr = degraded_request(self.identifier)
if (dr):
self.logger.info("image_request: degraded %s -> %s" %
(self.identifier, dr))
self.degraded = self.identifier
self.identifier = dr
self.iiif.quality = 'gray'
else:
# Parsed request OK, attempt to fulfill
self.logger.info("image_request: %s" % (self.identifier))
file = self.file
self.manipulator.srcfile = file
self.manipulator.do_first()
if (self.api_version < '2.0' and
self.iiif.format is None and
'Accept' in request.headers):
# In 1.0 and 1.1 conneg was specified as an alternative to format, see:
# http://iiif.io/api/image/1.0/#format
# http://iiif.io/api/image/1.1/#parameters-format
formats = {'image/jpeg': 'jpg', 'image/tiff': 'tif',
'image/png': 'png', 'image/gif': 'gif',
'image/jp2': 'jps', 'application/pdf': 'pdf'}
accept = do_conneg(request.headers['Accept'], list(formats.keys()))
# Ignore Accept header if not recognized, should this be an error
# instead?
if (accept in formats):
self.iiif.format = formats[accept]
(outfile, mime_type) = self.manipulator.derive(file, self.iiif)
# FIXME - find efficient way to serve file with headers
self.add_compliance_header()
return send_file(outfile, mimetype=mime_type)
|
def image_request_response(self, path)
|
Parse image request and create response.
| 4.631508
| 4.626367
| 1.001111
|
self.add_compliance_header()
return self.make_response(*e.image_server_response(self.api_version))
|
def error_response(self, e)
|
Make response for an IIIFError e.
Also add compliance header.
| 16.194099
| 8.310901
| 1.948537
|
self.send_response(code)
self.send_header('Content-Type', 'text/xml')
self.add_compliance_header()
self.end_headers()
self.wfile.write(content)
|
def error_response(self, code, content='')
|
Construct and send error response.
| 2.62417
| 2.395551
| 1.095435
|
self.compliance_uri = None
self.iiif = IIIFRequest(baseurl='/')
try:
(of, mime_type) = self.do_GET_body()
if (not of):
raise IIIFError("Unexpected failure to open result image")
self.send_response(200, 'OK')
if (mime_type is not None):
self.send_header('Content-Type', mime_type)
self.add_compliance_header()
self.end_headers()
while (1):
buffer = of.read(8192)
if (not buffer):
break
self.wfile.write(buffer)
# Now cleanup
self.manipulator.cleanup()
except IIIFError as e:
if (self.debug):
e.text += "\nRequest parameters:\n" + str(self.iiif)
self.error_response(e.code, str(e))
except Exception as ue:
# Anything else becomes a 500 Internal Server Error
e = IIIFError(code=500, text="Something went wrong... %s ---- %s.\n" %
(str(ue), traceback.format_exc()))
if (self.debug):
e.text += "\nRequest parameters:\n" + str(self.iiif)
self.error_response(e.code, str(e))
|
def do_GET(self)
|
Implement the HTTP GET method.
The bulk of this code is wrapped in a big try block and anywhere
within the code may raise an IIIFError which then results in an
IIIF error response (section 5 of spec).
| 3.8448
| 3.531605
| 1.088683
|
iiif = self.iiif
if (len(self.path) > 1024):
raise IIIFError(code=414,
text="URI Too Long: Max 1024 chars, got %d\n" % len(self.path))
try:
# self.path has leading / then identifier/params...
self.path = self.path.lstrip('/')
sys.stderr.write("path = %s" % (self.path))
iiif.parse_url(self.path)
except Exception as e:
# Something completely unexpected => 500
raise IIIFError(code=500,
text="Internal Server Error: unexpected exception parsing request (" + str(e) + ")")
# Now we have a full iiif request
if (re.match('[\w\.\-]+$', iiif.identifier)):
file = os.path.join(TESTIMAGE_DIR, iiif.identifier)
if (not os.path.isfile(file)):
images_available = ""
for image_file in os.listdir(TESTIMAGE_DIR):
if (os.path.isfile(os.path.join(TESTIMAGE_DIR, image_file))):
images_available += " " + image_file + "\n"
raise IIIFError(code=404, parameter="identifier",
text="Image resource '" + iiif.identifier + "' not found. Local image files available:\n" + images_available)
else:
raise IIIFError(code=404, parameter="identifier",
text="Image resource '" + iiif.identifier + "' not found. Only local test images and http: URIs for images are supported.\n")
# Now know image is OK
manipulator = IIIFRequestHandler.manipulator_class()
# Stash manipulator object so we can cleanup after reading file
self.manipulator = manipulator
self.compliance_uri = manipulator.compliance_uri
if (iiif.info):
# get size
manipulator.srcfile = file
manipulator.do_first()
# most of info.json comes from config, a few things
# specific to image
i = IIIFInfo()
i.identifier = self.iiif.identifier
i.width = manipulator.width
i.height = manipulator.height
import io
return(io.StringIO(i.as_json()), "application/json")
else:
(outfile, mime_type) = manipulator.derive(file, iiif)
return(open(outfile, 'r'), mime_type)
|
def do_GET_body(self)
|
Create body of GET.
| 4.666954
| 4.639129
| 1.005998
|
if (cookie_prefix is None):
self.cookie_prefix = "%06d_" % int(random.random() * 1000000)
else:
self.cookie_prefix = cookie_prefix
|
def set_cookie_prefix(self, cookie_prefix=None)
|
Set a random cookie prefix unless one is specified.
In order to run multiple demonstration auth services on the
same server we need to have different cookie names for each
auth domain. Unless cookie_prefix is set, generate a random
one.
| 2.843912
| 2.53841
| 1.120352
|
if (self.login_uri):
svc = self.login_service_description()
svcs = []
if (self.logout_uri):
svcs.append(self.logout_service_description())
if (self.client_id_uri):
svcs.append(self.client_id_service_description())
if (self.access_token_uri):
svcs.append(self.access_token_service_description())
# Add one as direct child of service property, else array for >1
if (len(svcs) == 1):
svc['service'] = svcs[0]
elif (len(svcs) > 1):
svc['service'] = svcs
info.add_service(svc)
|
def add_services(self, info)
|
Add auth service descriptions to an IIIFInfo object.
Login service description is the wrapper for all other
auth service descriptions so we have nothing unless
self.login_uri is specified. If we do then add any other
auth services at children.
Exactly the same structure is used in the authorized
and unauthorized cases (although in the data could be
different).
| 3.173129
| 2.750813
| 1.153524
|
label = 'Login to ' + self.name
if (self.auth_type):
label = label + ' (' + self.auth_type + ')'
desc = {"@id": self.login_uri,
"profile": self.profile_base + self.auth_pattern,
"label": label}
if (self.header):
desc['header'] = self.header
if (self.description):
desc['description'] = self.description
return desc
|
def login_service_description(self)
|
Login service description.
The login service description _MUST_ include the token service
description. The authentication pattern is indicated via the
profile URI which is built using self.auth_pattern.
| 4.11698
| 3.40516
| 1.209041
|
label = 'Logout from ' + self.name
if (self.auth_type):
label = label + ' (' + self.auth_type + ')'
return({"@id": self.logout_uri,
"profile": self.profile_base + 'logout',
"label": label})
|
def logout_service_description(self)
|
Logout service description.
| 6.07914
| 6.239717
| 0.974265
|
if (token):
data = {"accessToken": token,
"expiresIn": self.access_token_lifetime}
if (message_id):
data['messageId'] = message_id
else:
data = {"error": "client_unauthorized",
"description": "No authorization details received"}
return data
|
def access_token_response(self, token, message_id=None)
|
Access token response structure.
Success if token is set, otherwise (None, empty string) give error
response. If message_id is set then an extra messageId attribute is
set in the response to handle postMessage() responses.
| 3.7221
| 3.907502
| 0.952552
|
uri = scheme + '://' + host
if (port and not ((scheme == 'http' and port == 80) or
(scheme == 'https' and port == 443))):
uri += ':' + str(port)
if (prefix):
uri += '/' + prefix
return uri
|
def scheme_host_port_prefix(self, scheme='http', host='host',
port=None, prefix=None)
|
Return URI composed of scheme, server, port, and prefix.
| 2.150217
| 2.065581
| 1.040974
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.