MCPcopy Create free account
hub / github.com/OpenDriveLab/ReSim / url_opener

Function url_opener

SwissArmyTransformer/sat/data_utils/webds.py:192–225  ·  view source on GitHub ↗

Open URLs and yield a stream of url+stream pairs. Args: data: iterator over dict(url=...) handler: exception handler. kw: keyword arguments for gopen.gopen. Yields: a stream of url+stream pairs.

(
    data,
    handler,
    **kw,
)

Source from the content-addressed store, hash-verified

190 break
191
192def url_opener(
193 data,
194 handler,
195 **kw,
196):
197 """Open URLs and yield a stream of url+stream pairs.
198
199 Args:
200 data: iterator over dict(url=...)
201 handler: exception handler.
202 kw: keyword arguments for gopen.gopen.
203
204 Yields:
205 a stream of url+stream pairs.
206 """
207 for sample in data:
208 assert isinstance(sample, dict), sample
209 assert "url" in sample
210 url = sample["url"]
211 try:
212 stream = gopen(url, **kw)
213 if hasattr(stream, 'meta_stream'):
214 meta_stream = stream.meta_stream
215 del stream.meta_stream
216 else:
217 meta_stream = None
218 sample.update(stream=stream, meta_stream=meta_stream)
219 yield sample
220 except Exception as exn:
221 exn.args = exn.args + (url,)
222 if handler(exn):
223 continue
224 else:
225 break
226
227def tarfile_samples_with_meta(src, meta_names, handler=reraise_exception):
228 streams = url_opener(src, handler=handler)

Callers 2

jsonl_samplesFunction · 0.70

Calls 1

updateMethod · 0.45

Tested by

no test coverage detected