Thin db export revert

Author: abdullahajam

commands/web/thin-db-export

@@ -45,6 +45,8 @@ DB_NAME=""
 
 # Patterns to empty on the copied DB (only applied to tables that exist)
 TRUNCATE_PATTERNS=(
+  "^node"                # nodes and revisions/fields
+  "^media"               # media entities
   "^taxonomy"            # taxonomy terms and indexes
   "^comment"             # comments
   "^path_alias"          # path aliases
@@ -64,12 +66,6 @@ TRUNCATE_PATTERNS=(
   "^history$"            # content read markers
 )
 
-# Patterns for selective cleanup (keep 1 of each type)
-SELECTIVE_CLEANUP_PATTERNS=(
-  "^node"                # nodes and revisions/fields
-  "^media"               # media entities
-)
-
 # --------------------------
 # Create copy database
 # --------------------------
@@ -126,7 +122,7 @@ echo "✅ Database copy created."
 # --------------------------
 # Thin the copied DB
 # --------------------------
-echo "🗑️  Removing content and rebuildable data from ${DB_NAME} (keeping 1 node of each content type and 1 media of each type)..."
+echo "🗑️  Removing content and rebuildable data from ${DB_NAME}..."
 
 # Build and execute truncate statements in batches for better performance
 echo "Building truncate statements..."
@@ -154,140 +150,6 @@ echo "Executing batch truncate..."
 drush sql:cli $DRUSH_ARGS < "$TRUNCATE_SQL" || true
 rm -f "$TRUNCATE_SQL"
 
-# Selective cleanup: keep 1 of each content type/bundle
-for pattern in "${SELECTIVE_CLEANUP_PATTERNS[@]}"; do
-  echo "Selective cleanup for tables matching: ${pattern}"
-
-  if [[ "$pattern" == "^node" ]]; then
-    echo "Keeping 1 node of each content type..."
-
-    # Create batch SQL file for node cleanup
-    NODE_CLEANUP_SQL="/tmp/node_cleanup.sql"
-    echo "SET FOREIGN_KEY_CHECKS=0;" > "$NODE_CLEANUP_SQL"
-
-    # Get nodes to keep (one per type) in a single query
-    NODES_TO_KEEP=$(drush sql:query $DRUSH_ARGS "
-      SELECT nfd1.type, MIN(nfd1.nid) as keep_nid
-      FROM node_field_data nfd1
-      GROUP BY nfd1.type;
-    " 2>/dev/null | grep -v '^type' || true)
-
-    if [ -n "$NODES_TO_KEEP" ]; then
-      # Build list of NIDs to keep for IN clause
-      KEEP_NIDS=""
-      while IFS=$'\t' read -r content_type keep_nid; do
-        [ -z "$content_type" ] || [ -z "$keep_nid" ] && continue
-        echo "Keeping node ID ${keep_nid} for type ${content_type}"
-        if [ -z "$KEEP_NIDS" ]; then
-          KEEP_NIDS="$keep_nid"
-        else
-          KEEP_NIDS="$KEEP_NIDS,$keep_nid"
-        fi
-      done <<< "$NODES_TO_KEEP"
-
-      if [ -n "$KEEP_NIDS" ]; then
-        # Delete nodes not in keep list - batch operation
-        echo "DELETE FROM node WHERE nid NOT IN ($KEEP_NIDS);" >> "$NODE_CLEANUP_SQL"
-        echo "DELETE FROM node_field_data WHERE nid NOT IN ($KEEP_NIDS);" >> "$NODE_CLEANUP_SQL"
-        echo "DELETE FROM node_field_revision WHERE nid NOT IN ($KEEP_NIDS);" >> "$NODE_CLEANUP_SQL"
-
-        # Get all node field tables once
-        NODE_FIELD_TABLES=$(drush sql:query $DRUSH_ARGS "
-          SELECT TABLE_NAME FROM information_schema.TABLES
-          WHERE TABLE_SCHEMA='${DB_NAME}' AND TABLE_NAME LIKE 'node\_\_field\_%';
-        " 2>/dev/null || true)
-
-        # Add field table cleanup to batch
-        for table in $NODE_FIELD_TABLES; do
-          [ -z "$table" ] && continue
-          echo "DELETE FROM \`${table}\` WHERE entity_id NOT IN ($KEEP_NIDS);" >> "$NODE_CLEANUP_SQL"
-        done
-
-        # Get all node revision field tables once
-        NODE_REVISION_TABLES=$(drush sql:query $DRUSH_ARGS "
-          SELECT TABLE_NAME FROM information_schema.TABLES
-          WHERE TABLE_SCHEMA='${DB_NAME}' AND TABLE_NAME LIKE 'node\_revision\_\_field\_%';
-        " 2>/dev/null || true)
-
-        # Add revision field table cleanup to batch
-        for table in $NODE_REVISION_TABLES; do
-          [ -z "$table" ] && continue
-          echo "DELETE FROM \`${table}\` WHERE entity_id NOT IN ($KEEP_NIDS);" >> "$NODE_CLEANUP_SQL"
-        done
-      fi
-    fi
-
-    echo "SET FOREIGN_KEY_CHECKS=1;" >> "$NODE_CLEANUP_SQL"
-    echo "Executing batch node cleanup..."
-    drush sql:cli $DRUSH_ARGS < "$NODE_CLEANUP_SQL" || true
-    rm -f "$NODE_CLEANUP_SQL"
-
-  elif [[ "$pattern" == "^media" ]]; then
-    echo "Keeping 1 media of each media type..."
-
-    # Create batch SQL file for media cleanup
-    MEDIA_CLEANUP_SQL="/tmp/media_cleanup.sql"
-    echo "SET FOREIGN_KEY_CHECKS=0;" > "$MEDIA_CLEANUP_SQL"
-
-    # Get media to keep (one per type) in a single query
-    MEDIA_TO_KEEP=$(drush sql:query $DRUSH_ARGS "
-      SELECT mfd1.bundle, MIN(mfd1.mid) as keep_mid
-      FROM media_field_data mfd1
-      GROUP BY mfd1.bundle;
-    " 2>/dev/null | grep -v '^bundle' || true)
-
-    if [ -n "$MEDIA_TO_KEEP" ]; then
-      # Build list of MIDs to keep for IN clause
-      KEEP_MIDS=""
-      while IFS=$'\t' read -r media_type keep_mid; do
-        [ -z "$media_type" ] || [ -z "$keep_mid" ] && continue
-        echo "Keeping media ID ${keep_mid} for type ${media_type}"
-        if [ -z "$KEEP_MIDS" ]; then
-          KEEP_MIDS="$keep_mid"
-        else
-          KEEP_MIDS="$KEEP_MIDS,$keep_mid"
-        fi
-      done <<< "$MEDIA_TO_KEEP"
-
-      if [ -n "$KEEP_MIDS" ]; then
-        # Delete media not in keep list - batch operation
-        echo "DELETE FROM media WHERE mid NOT IN ($KEEP_MIDS);" >> "$MEDIA_CLEANUP_SQL"
-        echo "DELETE FROM media_field_data WHERE mid NOT IN ($KEEP_MIDS);" >> "$MEDIA_CLEANUP_SQL"
-        echo "DELETE FROM media_field_revision WHERE mid NOT IN ($KEEP_MIDS);" >> "$MEDIA_CLEANUP_SQL"
-
-        # Get all media field tables once
-        MEDIA_FIELD_TABLES=$(drush sql:query $DRUSH_ARGS "
-          SELECT TABLE_NAME FROM information_schema.TABLES
-          WHERE TABLE_SCHEMA='${DB_NAME}' AND TABLE_NAME LIKE 'media\_\_field\_%';
-        " 2>/dev/null || true)
-
-        # Add field table cleanup to batch
-        for table in $MEDIA_FIELD_TABLES; do
-          [ -z "$table" ] && continue
-          echo "DELETE FROM \`${table}\` WHERE entity_id NOT IN ($KEEP_MIDS);" >> "$MEDIA_CLEANUP_SQL"
-        done
-
-        # Get all media revision field tables once
-        MEDIA_REVISION_TABLES=$(drush sql:query $DRUSH_ARGS "
-          SELECT TABLE_NAME FROM information_schema.TABLES
-          WHERE TABLE_SCHEMA='${DB_NAME}' AND TABLE_NAME LIKE 'media\_revision\_\_field\_%';
-        " 2>/dev/null || true)
-
-        # Add revision field table cleanup to batch
-        for table in $MEDIA_REVISION_TABLES; do
-          [ -z "$table" ] && continue
-          echo "DELETE FROM \`${table}\` WHERE entity_id NOT IN ($KEEP_MIDS);" >> "$MEDIA_CLEANUP_SQL"
-        done
-      fi
-    fi
-
-    echo "SET FOREIGN_KEY_CHECKS=1;" >> "$MEDIA_CLEANUP_SQL"
-    echo "Executing batch media cleanup..."
-    drush sql:cli $DRUSH_ARGS < "$MEDIA_CLEANUP_SQL" || true
-    rm -f "$MEDIA_CLEANUP_SQL"
-  fi
-done
-
 # Remove pathauto state keys if table exists (safe on all sites with key_value)
 drush sql:query $DRUSH_ARGS "
   DELETE FROM key_value