Merge pull request #183 from lsst/tickets/DM-25965

DM-25965: Switch ingest to using ButlerURI

python/lsst/obs/decam/ingest.py

@@ -23,39 +23,52 @@
 import re
 
 
+from lsst.daf.butler import ButlerURI, Formatter
 from astro_metadata_translator import fix_header, DecamTranslator
 from lsst.afw.fits import readMetadata
 from lsst.pipe.tasks.ingest import ParseTask, IngestTask, IngestArgumentParser
 from lsst.obs.base.ingest import RawFileData
 import lsst.obs.base
-from ._instrument import DarkEnergyCamera
 
 __all__ = ["DecamRawIngestTask", "DecamIngestArgumentParser", "DecamIngestTask", "DecamParseTask"]
 
 
 class DecamRawIngestTask(lsst.obs.base.RawIngestTask):
     """Task for ingesting raw DECam data into a Gen3 Butler repository.
     """
-    def extractMetadata(self, filename: str) -> RawFileData:
+    def extractMetadata(self, filename: ButlerURI) -> RawFileData:
         datasets = []
-        fitsData = lsst.afw.fits.Fits(filename, 'r')
-        # NOTE: The primary header (HDU=0) does not contain detector data.
-        for i in range(1, fitsData.countHdus()):
-            fitsData.setHdu(i)
-            header = fitsData.readMetadata()
-            if header['CCDNUM'] > 62:  # ignore the guide CCDs
-                continue
-            datasets.append(self._calculate_dataset_info(header, filename))
-
-        # The data model currently assumes that whilst multiple datasets
-        # can be associated with a single file, they must all share the
-        # same formatter.
-        instrument = DarkEnergyCamera()
-        FormatterClass = instrument.getRawFormatter(datasets[0].dataId)
+        try:
+            with filename.as_local() as local_file:
+                fitsData = lsst.afw.fits.Fits(local_file.ospath, 'r')
+                # NOTE: The primary header (HDU=0) does not contain detector data.
+                for i in range(1, fitsData.countHdus()):
+                    fitsData.setHdu(i)
+                    header = fitsData.readMetadata()
+                    if header['CCDNUM'] > 62:  # ignore the guide CCDs
+                        continue
+                    datasets.append(self._calculate_dataset_info(header, filename))
+        except Exception as e:
+            self.log.debug("Problem extracting metadata from %s: %s", filename, e)
+            # Indicate to the caller that we failed to read.
+            # Do not try to ingest partial contents of file.
+            datasets = []
+            formatterClass = Formatter
+            instrument = None
+            self._on_metadata_failure(filename, e)
+            if self.config.failFast:
+                raise RuntimeError(f"Problem extracting metadata for file {filename}") from e
+        else:
+            # The data model currently assumes that whilst multiple datasets
+            # can be associated with a single file, they must all share the
+            # same formatter.
+            instrument, formatterClass = self._determine_instrument_formatter(datasets[0].dataId, filename)
+            if instrument is None:
+                datasets = []
 
         self.log.debug(f"Found images for {len(datasets)} detectors in {filename}")
         return RawFileData(datasets=datasets, filename=filename,
-                           FormatterClass=FormatterClass,
+                           FormatterClass=formatterClass,
                            instrumentClass=type(instrument))