def get_html_body_str(self):
"""HTML for browser track
This HTML section provides:
1) A plain text for UCSC genome browser &hgct_customText=
- This is written to a text file (.txt) on the output directory
2) An encoded URL for UCSC genome browser &hgct_customText=
3) Clickable href link
- Full URL is written to a text file (.url) on the output directory
"""
if self._ucsc_genome_db is None:
return ''
txt = self.__make_ucsc_track_txt()
if txt is None or txt == '':
return ''
if self._ucsc_genome_pos is not None:
extra_param = (
CrooHtmlReportUCSCTracks.UCSC_BROWSER_QUERY_POS_PARAM +
self._ucsc_genome_pos)
else:
extra_param = ''
# save to TXT
uri_txt = os.path.join(
self._out_dir,
CrooHtmlReportUCSCTracks.UCSC_TRACKS_TXT.format(
workflow_id=self._workflow_id),
)
# localize TXT
# long URL doesn't work
u = AutoURI(uri_txt)
u.write(txt, no_lock=True)
url_trackhub_txt_file = None
if isinstance(u, GCSURI):
if self._public_gcs:
url_trackhub_txt_file = u.get_public_url()
elif self._use_presigned_url_gcs:
url_trackhub_txt_file = u.get_presigned_url(
duration=self._duration_presigned_url_gcs,
private_key_file=self._gcp_private_key,
)
elif isinstance(u, S3URI):
if self._use_presigned_url_s3:
url_trackhub_txt_file = u.get_presigned_url(
duration=self._duration_presigned_url_s3)
elif isinstance(u, AbsPath):
if self._map_path_to_url:
url_trackhub_txt_file = u.get_mapped_url(
map_path_to_url=self._map_path_to_url)
html = ''
url = CrooHtmlReportUCSCTracks.UCSC_BROWSER_QUERY_URL.format(
db=self._ucsc_genome_db,
extra_param=extra_param,
encoded=urllib.parse.quote(txt),
)
html += CrooHtmlReportUCSCTracks.HTML_TRACK_HUB_LINK.format(
title='UCSC browser tracks', url=url)
if url_trackhub_txt_file is not None:
url = CrooHtmlReportUCSCTracks.UCSC_BROWSER_QUERY_URL.format(
db=self._ucsc_genome_db,
extra_param=extra_param,
encoded=urllib.parse.quote(url_trackhub_txt_file),
)
html += CrooHtmlReportUCSCTracks.HTML_TRACK_HUB_LINK.format(
title='UCSC browser tracks (if the above link does not work)',
url=url)
html += CrooHtmlReportUCSCTracks.HTML_TRACK_HUB_TEXT.format(
title=
'UCSC track hub plain text. Paste it directly to custom track edit box '
'on UCSC genome browser.',
txt=txt,
)
return html
def organize_output(self):
"""Organize outputs
"""
report = CrooHtmlReport(
out_dir=self._out_dir,
workflow_id=self._cm.get_workflow_id(),
dag=self._task_graph,
task_graph_template=self._task_graph_template,
public_gcs=self._public_gcs,
gcp_private_key=self._gcp_private_key,
use_presigned_url_gcs=self._use_presigned_url_gcs,
use_presigned_url_s3=self._use_presigned_url_s3,
duration_presigned_url_s3=self._duration_presigned_url_s3,
duration_presigned_url_gcs=self._duration_presigned_url_gcs,
map_path_to_url=self._map_path_to_url,
ucsc_genome_db=self._ucsc_genome_db,
ucsc_genome_pos=self._ucsc_genome_pos,
)
if self._input_def_json is not None:
for input_name, input_obj in self._input_def_json.items():
node_format = input_obj.get('node')
subgraph = input_obj.get('subgraph')
for _, node in self._task_graph.get_nodes():
# if node is pipeline's input
if (node.type != 'output' or node.task_name is not None
or node.output_name != input_name):
continue
full_path = node.output_path
shard_idx = node.shard_idx
if node_format is not None:
interpreted_node_format = Croo.__interpret_inline_exp(
node_format, full_path, shard_idx)
if subgraph is not None:
interpreted_subgraph = Croo.__interpret_inline_exp(
subgraph, full_path, shard_idx)
else:
interpreted_subgraph = None
report.add_to_task_graph(
node.output_name,
None,
shard_idx,
full_path,
interpreted_node_format,
interpreted_subgraph,
)
for task_name, out_vars in self._out_def_json.items():
for output_name, output_obj in out_vars.items():
path = output_obj.get('path')
table_item = output_obj.get('table')
ucsc_track = output_obj.get('ucsc_track')
node_format = output_obj.get('node')
subgraph = output_obj.get('subgraph')
for _, node in self._task_graph.get_nodes():
# look at output nodes only (not a task node)
if task_name != node.task_name or node.type != 'task':
continue
all_outputs = node.all_outputs
shard_idx = node.shard_idx
if not all_outputs:
continue
for k, full_path, _ in all_outputs:
if k != output_name:
continue
target_uri = full_path
if path is not None:
interpreted_path = Croo.__interpret_inline_exp(
path, full_path, shard_idx)
au = AutoURI(full_path)
target_path = os.path.join(self._out_dir,
interpreted_path)
if self._soft_link:
au_target = AutoURI(target_path)
if isinstance(au, AbsPath) and isinstance(
au_target, AbsPath):
au.soft_link(target_path, force=True)
target_uri = target_path
else:
target_uri = full_path
else:
target_uri = au.cp(
target_path,
no_checksum=self._no_checksum,
make_md5_file=True,
no_lock=True,
)
# get presigned URLs if possible
target_url = None
if (path is not None or table_item is not None
or ucsc_track is not None
or node_format is not None):
u = AutoURI(target_uri)
if isinstance(u, GCSURI):
if self._public_gcs:
target_url = u.get_public_url()
elif self._use_presigned_url_gcs:
target_url = u.get_presigned_url(
duration=self.
_duration_presigned_url_gcs,
private_key_file=self._gcp_private_key,
)
elif isinstance(u, S3URI):
if self._use_presigned_url_s3:
target_url = u.get_presigned_url(
duration=self.
_duration_presigned_url_s3)
elif isinstance(u, AbsPath):
if self._map_path_to_url:
target_url = u.get_mapped_url(
map_path_to_url=self._map_path_to_url)
if table_item is not None:
interpreted_table_item = Croo.__interpret_inline_exp(
table_item, full_path, shard_idx)
# add to file table
report.add_to_file_table(target_uri, target_url,
interpreted_table_item)
if ucsc_track is not None and target_url is not None:
interpreted_ucsc_track = Croo.__interpret_inline_exp(
ucsc_track, full_path, shard_idx)
report.add_to_ucsc_track(target_url,
interpreted_ucsc_track)
if node_format is not None:
interpreted_node_format = Croo.__interpret_inline_exp(
node_format, full_path, shard_idx)
if subgraph is not None:
interpreted_subgraph = Croo.__interpret_inline_exp(
subgraph, full_path, shard_idx)
else:
interpreted_subgraph = None
report.add_to_task_graph(
output_name,
task_name,
shard_idx,
full_path
if target_url is None else target_url,
interpreted_node_format,
interpreted_subgraph,
)
# write to html report
report.save_to_file()