From 6e8437172dafe4139049d4e7c8707943ea9c57ac Mon Sep 17 00:00:00 2001
From: Shlok Natarajan <shlok.natarajan@gmail.com>
Date: Mon, 14 Jul 2025 12:27:19 -0700
Subject: [PATCH 1/2] feat: copy markdown helper

---
 pixi.toml                              |  1 +
 src/components/annotations_pipeline.py |  2 +-
 src/copy_markdown.py                   | 39 ++++++++++++++++++++++++++
 3 files changed, 41 insertions(+), 1 deletion(-)
 create mode 100644 src/copy_markdown.py

diff --git a/pixi.toml b/pixi.toml
index b9ec1b2..c9cb954 100644
--- a/pixi.toml
+++ b/pixi.toml
@@ -17,6 +17,7 @@ update-download-map = "python -c 'from src.fetch_articles.article_downloader imp
 download-articles = "python -m src.fetch_articles.article_downloader"
 download-data = "gdown --fuzzy https://drive.google.com/file/d/1qtQWvi0x_k5_JofgrfsgkWzlIdb6isr9/view && unzip autogkb-data.zip && rm autogkb-data.zip"
 setup-repo = "pixi install && pixi run download-data"
+copy-markdown = "python -m src.copy_markdown"
 
 [dependencies]
 seaborn = ">=0.13.2,<0.14"
diff --git a/src/components/annotations_pipeline.py b/src/components/annotations_pipeline.py
index 8e6ba13..1111929 100644
--- a/src/components/annotations_pipeline.py
+++ b/src/components/annotations_pipeline.py
@@ -40,7 +40,7 @@ def generate_final_structure(self):
             "functional_annotations": self.functional_annotations,
         }
 
-    def run(self, save_path: str = "data/extractions"):
+    def run(self, save_path: str = "data/annotations"):
         logger.info("Getting Study Parameters")
         self.study_parameters = get_study_parameters(self.article_text)
 
diff --git a/src/copy_markdown.py b/src/copy_markdown.py
new file mode 100644
index 0000000..cc9aa6e
--- /dev/null
+++ b/src/copy_markdown.py
@@ -0,0 +1,39 @@
+from typing import List
+import os
+import shutil
+from loguru import logger
+from pathlib import Path
+"""
+Pass in a list of PMCIDs and have those files copied to a new folder
+"""
+
+def copy_markdown(pmcids: List[str]) -> None:
+    succesful = 0
+    for pmcid in pmcids:
+        try:
+            source_file = Path("data") / "articles" / f"{pmcid}.md" 
+            destination = Path("data") / "extractions" / "markdown" 
+            os.makedirs(destination, exist_ok=True)
+            shutil.copy2(source_file, destination / f"{pmcid}.md")
+            succesful += 1
+        except Exception as e:
+            logger.error(e)
+    logger.info(f"Copied {succesful}/{len(pmcids)} markdown to data/extractions/markdown")
+
+def main():
+    pmcids = [
+        "PMC4737107",
+        "PMC5712579",
+        "PMC5728534",
+        "PMC5749368",
+        "PMC11730665"
+    ]
+    copy_markdown(pmcids=pmcids)
+
+if __name__ == "__main__":
+    main()
+
+
+
+
+

From a7a4c73c511db92e36d55a7266be5d0a4f517a22 Mon Sep 17 00:00:00 2001
From: Shlok Natarajan <shlok.natarajan@gmail.com>
Date: Tue, 15 Jul 2025 10:48:44 -0700
Subject: [PATCH 2/2] chore: black formatting

---
 src/copy_markdown.py | 25 ++++++++++---------------
 1 file changed, 10 insertions(+), 15 deletions(-)

diff --git a/src/copy_markdown.py b/src/copy_markdown.py
index cc9aa6e..e432cd1 100644
--- a/src/copy_markdown.py
+++ b/src/copy_markdown.py
@@ -3,37 +3,32 @@
 import shutil
 from loguru import logger
 from pathlib import Path
+
 """
 Pass in a list of PMCIDs and have those files copied to a new folder
 """
 
+
 def copy_markdown(pmcids: List[str]) -> None:
     succesful = 0
     for pmcid in pmcids:
         try:
-            source_file = Path("data") / "articles" / f"{pmcid}.md" 
-            destination = Path("data") / "extractions" / "markdown" 
+            source_file = Path("data") / "articles" / f"{pmcid}.md"
+            destination = Path("data") / "extractions" / "markdown"
             os.makedirs(destination, exist_ok=True)
             shutil.copy2(source_file, destination / f"{pmcid}.md")
             succesful += 1
         except Exception as e:
             logger.error(e)
-    logger.info(f"Copied {succesful}/{len(pmcids)} markdown to data/extractions/markdown")
+    logger.info(
+        f"Copied {succesful}/{len(pmcids)} markdown to data/extractions/markdown"
+    )
+
 
 def main():
-    pmcids = [
-        "PMC4737107",
-        "PMC5712579",
-        "PMC5728534",
-        "PMC5749368",
-        "PMC11730665"
-    ]
+    pmcids = ["PMC4737107", "PMC5712579", "PMC5728534", "PMC5749368", "PMC11730665"]
     copy_markdown(pmcids=pmcids)
 
+
 if __name__ == "__main__":
     main()
-
-
-
-
-