PDF path mismatch between scan monitor and CHX

[hyperrealist]

Currently, scan monitor is looking for PDF files like

${output_dir}/*.pdf

sirepo/sirepo/raydata/analysis_driver/__init__.py

Lines 32 to 33 in 59b71d0

	def get_analysis_pdf_paths(self):
	return pkio.walk_tree(self.get_output_dir(), r".*\.pdf$")

However, CHX pdf files are written to the parent directory of this path, with a file name tagged with the uid and a revision number. I believe this is to avoid overwriting the PDF files between multiple notebook runs.

Semantically, CHX PDF path is something like

${output_dir}/../*${uid}*.pdf    # note: this will include revisions

This path mismatch is preventing raydata from offering generated PDF files for download (manually copying to ${output_dir} makes raydata recognize them).

Edit:
To clarify further, CHX ${output_dir} itself is structured like ${PARENT_OUTPUT_DIR}/${uid} with the same ${PARENT_OUTPUT_DIR} for all scans for a given user (and cycle). It stores all PDF revisions from all runs, tagged with uid and a revision number. Everything in ${PARENT_OUTPUT_DIR}/${uid} seems to get overwritten upon rerun.

[e-carlin]

Thanks for opening this @hyperrealist. I'll address.