Add ProHarvest Seeds: 119 varieties + 161 cross-vendor plot reports (#16)

Co-authored-by: claude <claude@jpaul.io>
Co-committed-by: claude <claude@jpaul.io>

rag/chunk.py

@@ -509,6 +509,28 @@ def _render_trial_chunk(sidecar: dict, md_text: str | None = None) -> str:
     # for each (Golden Harvest / LG Seeds / AgriGold).
     if source in ("gh_plot_reports", "lg_plot_reports", "agrigold_plot_reports"):
         return _render_gh_plot_chunk(sidecar)
+    if source == "proharvest_plots":
+        # Structured rows → shared cross-vendor renderer (publisher brand
+        # read from the sidecar). Foreign-format third-party PDFs that
+        # couldn't be parsed into rows carry raw_text=True and the verbatim
+        # table text in the .md body — embed that so they're still found.
+        if sidecar.get("results"):
+            return _render_gh_plot_chunk(sidecar)
+        crop = (sidecar.get("crop") or "").lower()
+        crop_label = {"corn": "Corn", "soybeans": "Soybean"}.get(crop, crop.title())
+        coop = sidecar.get("cooperator") or ""
+        state = sidecar.get("state") or ""
+        year = sidecar.get("year") or ""
+        head = [
+            f"# {crop_label} yield trial — {coop} ({state}, {year})", "",
+            "ProHarvest Seeds plot report (cross-vendor, verbatim from PDF).", "",
+        ]
+        body = md_text or ""
+        sep = "## Trial data (verbatim from PDF)"
+        if sep in body:
+            body = body.split(sep, 1)[1].strip()
+            body = re.sub(r"```", "", body).strip()
+        return "\n".join(head) + "\n" + body + "\n"
     if source == "agripro_trials":
         header = _render_agripro_trial_chunk(sidecar)
         if md_text: