From 56e8c439992aead39e920cf88ff47e3d62aa1a48 Mon Sep 17 00:00:00 2001
From: almac2022 <al@newgraphenvironment.com>
Date: Sun, 5 Apr 2026 08:53:05 -0700
Subject: [PATCH] Add id_segment for unique sub-segment identity after breaking
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

.frs_add_id_segment() initializes id_segment from linear_feature_id.
When frs_break_apply(segment_id = "id_segment") splits segments,
sub-segments get new unique id_segment values while linear_feature_id
is carried from the parent (FWA provenance preserved).

Tested: 510 segments → 532 after breaking. All id_segment unique.
Sub-segments share parent linear_feature_id correctly (e.g. one parent
split into 7 sub-segments).

Also adds naming conventions to CLAUDE.md (id_* prefix, col_* params,
frs_noun_verb functions, long-format habitat table structure).

Fixes #82
Relates to #83

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---
 R/frs_break.R |  8 ++++----
 R/utils.R     | 20 ++++++++++++++++++++
 2 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/R/frs_break.R b/R/frs_break.R
index a68d3c9..766d448 100644
--- a/R/frs_break.R
+++ b/R/frs_break.R
@@ -582,20 +582,20 @@ frs_break_apply <- function(conn, table, breaks,
   # Get writable columns only (excludes GENERATED ALWAYS columns)
   cols_writable <- .frs_table_columns(conn, table, exclude_generated = TRUE)
 
-  # Split columns: these come from the temp table, not the parent
+  # Split columns: segment_id, measures, and geom come from the split
   cols_split_all <- c(sid, "downstream_route_measure",
                       "upstream_route_measure", "geom")
-  # Only include split columns that are actually writable
   cols_split <- intersect(cols_split_all, cols_writable)
 
   # Carry columns: everything writable that isn't a split column
+  # This preserves linear_feature_id (FWA provenance) from parent
   cols_carry <- setdiff(cols_writable, cols_split_all)
 
   # Build INSERT column list and SELECT expressions
   cols_insert_parts <- character(0)
   select_parts <- character(0)
 
-  # segment_id — new ID from max + row_number
+  # segment_id — new unique ID from max + row_number
   if (sid %in% cols_split) {
     cols_insert_parts <- c(cols_insert_parts, sid)
     select_parts <- c(select_parts, sprintf(
@@ -604,7 +604,7 @@ frs_break_apply <- function(conn, table, breaks,
        )", sid, table))
   }
 
-  # Carried columns from parent
+  # Carried columns from parent (includes linear_feature_id for provenance)
   if (length(cols_carry) > 0) {
     cols_insert_parts <- c(cols_insert_parts, cols_carry)
     select_parts <- c(select_parts, paste0("s.", cols_carry))
diff --git a/R/utils.R b/R/utils.R
index da2354d..af2734a 100644
--- a/R/utils.R
+++ b/R/utils.R
@@ -60,6 +60,26 @@
 }
 
 
+#' Add id_segment column to a working table
+#'
+#' Assigns a unique integer ID to every row. Uses `linear_feature_id`
+#' as the starting value (so original FWA segments keep their ID),
+#' then generates new IDs for any rows added later by
+#' [frs_break_apply()].
+#'
+#' @param conn DBI connection.
+#' @param table Schema-qualified table name.
+#' @noRd
+.frs_add_id_segment <- function(conn, table) {
+  if (!inherits(conn, "DBIConnection")) return(invisible(NULL))
+  .frs_db_execute(conn, sprintf(
+    "ALTER TABLE %s ADD COLUMN IF NOT EXISTS id_segment integer", table))
+  .frs_db_execute(conn, sprintf(
+    "UPDATE %s SET id_segment = linear_feature_id
+     WHERE id_segment IS NULL", table))
+}
+
+
 #' Add indexes to a working table based on available columns
 #'
 #' Checks which index-worthy columns exist and creates appropriate indexes.