From 5ba858fc608fa629fa0b8c2fd0528861b3f15095 Mon Sep 17 00:00:00 2001
From: amc-corey-cox <69321580+amc-corey-cox@users.noreply.github.com>
Date: Thu, 10 Sep 2026 10:37:21 -0500
Subject: [PATCH 1/2] Prefix synthetic raw tables with SYNTHETIC so they cannot
 be mistaken for dbGaP exports

---
 synthetic/README.md   |  9 +++++++--
 synthetic/generate.py | 13 ++++++++++++-
 2 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/synthetic/README.md b/synthetic/README.md
index 86114be..511e01e 100644
--- a/synthetic/README.md
+++ b/synthetic/README.md
@@ -16,10 +16,15 @@ second implementation of the transformation, free to drift from the real one in
 ways nobody would notice until a portal built on it met real data.
 
 ```
-generate.py  ->  data/raw/*.txt.gz  ->  dm-bip map-data  ->  harmonized BDCHM
-specs.py     ->  specs/*/*.yaml     ->  ^
+generate.py  ->  data/raw/SYNTHETIC.*.txt.gz  ->  dm-bip map-data  ->  harmonized BDCHM
+specs.py     ->  specs/*/*.yaml               ->  ^
 ```
 
+The raw tables follow the dbGaP naming convention exactly apart from a
+`SYNTHETIC.` prefix. Without it a file lifted out of its directory is
+indistinguishable from a controlled-access export, and a file is identified by
+its name rather than by the citation line in its header.
+
 ## Running it
 
 ```bash
diff --git a/synthetic/generate.py b/synthetic/generate.py
index ee0c136..d872c7e 100644
--- a/synthetic/generate.py
+++ b/synthetic/generate.py
@@ -18,6 +18,17 @@
 
 CITATION = "Synthetic corpus for portal development. Not derived from participant data."
 
+# Every raw table is named so it cannot be mistaken for a controlled-access
+# dbGaP export. The header already says so, but a file that escapes its
+# directory is identified by its name, not by its contents, and these otherwise
+# follow the dbGaP naming convention exactly.
+#
+# A prefix rather than a suffix: it is what a directory listing sorts on and
+# shows first, and it survives the truncation that hides the middle of a long
+# name. dm-bip finds the table accession with an unanchored search for
+# `pht[0-9]+`, so the prefix does not disturb the pipeline.
+SYNTHETIC_MARKER = "SYNTHETIC"
+
 # Column layouts. The phv accessions are fictional but well-formed, and are
 # what the transformation specs reference.
 LAYOUTS = {
@@ -93,7 +104,7 @@ def write_table(out_dir, study, table, rows):
     """Write one table in dbGaP raw format, returning its path."""
     columns, phv_count = LAYOUTS[table]
     pht = study.tables[table]
-    path = out_dir / f"{study.phs}.v1.{pht}.v1.p1.c1.ex0_1s.HMB.txt.gz"
+    path = out_dir / f"{SYNTHETIC_MARKER}.{study.phs}.v1.{pht}.v1.p1.c1.ex0_1s.HMB.txt.gz"
 
     with gzip.open(path, "wt", newline="") as fh:
         fh.write(f"# Study accession: {study.phs}.v1.p1\n")

From 4d22e20fb4bda02de781a86f0e8e66bd05ec004b Mon Sep 17 00:00:00 2001
From: amc-corey-cox <69321580+amc-corey-cox@users.noreply.github.com>
Date: Thu, 10 Sep 2026 10:48:49 -0500
Subject: [PATCH 2/2] Address PR review feedback

---
 synthetic/README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/synthetic/README.md b/synthetic/README.md
index 511e01e..0fb8e35 100644
--- a/synthetic/README.md
+++ b/synthetic/README.md
@@ -16,8 +16,8 @@ second implementation of the transformation, free to drift from the real one in
 ways nobody would notice until a portal built on it met real data.
 
 ```
-generate.py  ->  data/raw/SYNTHETIC.*.txt.gz  ->  dm-bip map-data  ->  harmonized BDCHM
-specs.py     ->  specs/*/*.yaml               ->  ^
+generate.py  ->  data/raw/*/SYNTHETIC.*.txt.gz  ->  dm-bip map-data  ->  harmonized BDCHM
+specs.py     ->  specs/*/*.yaml                 ->  ^
 ```
 
 The raw tables follow the dbGaP naming convention exactly apart from a