Skip to content
This repository has been archived by the owner on Oct 16, 2024. It is now read-only.

Commit

Permalink
Merge pull request #2208 from jcushman/export-fixes
Browse files Browse the repository at this point in the history
Add volume_folder to export json
  • Loading branch information
bensteinberg authored Mar 6, 2024
2 parents b9abdf8 + 90768a9 commit f8db93b
Show file tree
Hide file tree
Showing 9 changed files with 23 additions and 8 deletions.
11 changes: 7 additions & 4 deletions capstone/scripts/export_cap_static.py
Original file line number Diff line number Diff line change
Expand Up @@ -46,7 +46,7 @@ def finalize_reporters_dir(dest_dir: Path) -> None:
if reporter_dir.name in reporter_slug_dict_reverse:
reporter = Reporter.objects.get(pk=reporter_slug_dict_reverse[reporter_dir.name])
else:
reporter = Reporter.objects.get(short_name_slug=reporter_dir.name)
reporter = Reporter.objects.filter(short_name_slug=reporter_dir.name).exclude(start_year=None).get()

# export reporter metadata
reporter_dict = call_serializer(ReporterSerializer, reporter)
Expand All @@ -57,9 +57,12 @@ def finalize_reporters_dir(dest_dir: Path) -> None:

# write reporter-level VolumesMetadata.json
print("Writing VolumesMetadata.json")
volumes_metadata = [json.loads(f.read_text()) for f in natsorted(reporter_dir.glob("*/VolumeMetadata.json"))]
for volume in volumes_metadata:
volumes_metadata = []
for f in natsorted(reporter_dir.glob("*/VolumeMetadata.json")):
volume = json.loads(f.read_text())
volume["volume_folder"] = f.parent.name
volume["reporter_slug"] = reporter_dir.name
volumes_metadata.append(volume)
write_json(reporter_dir / "VolumesMetadata.json", volumes_metadata)

# write ReportersMetadata.json
Expand Down Expand Up @@ -114,7 +117,6 @@ def crosscheck_reporters_dir(dest_dir: Path) -> None:
case_count=Count('case_metadatas', filter=Q(case_metadatas__in_scope=True))
).select_related("reporter")
reporter_dirs = set()
# suspicious_volumes = []
for volume in tqdm(volumes):
if not volume.case_count:
continue
Expand Down Expand Up @@ -223,6 +225,7 @@ def export_volume(volume: VolumeMetadata, dest_dir: Path) -> None:
html_dir = temp_volume_dir / "html"
html_dir.mkdir()
volume_metadata = volume_to_dict(volume)
volume_metadata["volume_folder"] = volume_prefix
write_json(temp_volume_dir / "VolumeMetadata.json", volume_metadata)

# variables for case export loop
Expand Down
2 changes: 2 additions & 0 deletions capstone/test_data/cap_static/redacted/VolumesMetadata.json
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null,
"volume_folder": "1",
"reporter_slug": "us"
},
{
Expand Down Expand Up @@ -57,6 +58,7 @@
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null,
"volume_folder": "2",
"reporter_slug": "us"
}
]
Original file line number Diff line number Diff line change
Expand Up @@ -25,5 +25,6 @@
"publication_city": null,
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null
"nominative_reporter": null,
"volume_folder": "1"
}
Original file line number Diff line number Diff line change
Expand Up @@ -25,5 +25,6 @@
"publication_city": null,
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null
"nominative_reporter": null,
"volume_folder": "2"
}
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null,
"volume_folder": "1",
"reporter_slug": "us"
},
{
Expand Down Expand Up @@ -57,6 +58,7 @@
"second_part_of_id": null,
"redacted": true,
"nominative_reporter": null,
"volume_folder": "2",
"reporter_slug": "us"
}
]
2 changes: 2 additions & 0 deletions capstone/test_data/cap_static/unredacted/VolumesMetadata.json
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null,
"volume_folder": "1",
"reporter_slug": "us"
},
{
Expand Down Expand Up @@ -57,6 +58,7 @@
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null,
"volume_folder": "2",
"reporter_slug": "us"
}
]
Original file line number Diff line number Diff line change
Expand Up @@ -25,5 +25,6 @@
"publication_city": null,
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null
"nominative_reporter": null,
"volume_folder": "1"
}
Original file line number Diff line number Diff line change
Expand Up @@ -25,5 +25,6 @@
"publication_city": null,
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null
"nominative_reporter": null,
"volume_folder": "2"
}
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null,
"volume_folder": "1",
"reporter_slug": "us"
},
{
Expand Down Expand Up @@ -57,6 +58,7 @@
"second_part_of_id": null,
"redacted": false,
"nominative_reporter": null,
"volume_folder": "2",
"reporter_slug": "us"
}
]

0 comments on commit f8db93b

Please sign in to comment.