There's a pretty straightforward way that you could do this with Python, and you could convert it into a script tool as well to use going forward. Although, in my version I work mostly in a GDB.
import arcpy
# Gathering layer and field info
project = arcpy.mp.ArcGISProject("CURRENT")
mv = project.activeMap
lyr = mv.listLayers("POLY_LAYER_NAME")
desc = arcpy.Describe(lyr)
shape_name = desc.shapeFieldName
oid_name = desc.OIDFieldName
flds = [fld.name.replace(shape_name, "SHAPE@").replace(oid_name, "OID@") for fld in arcpy.ListFields(lyr) if "Shape_" not in fld.name]
shape_index = flds.index("SHAPE@")
#Create Empty Line Feature Class
workspace = lyr.connectionProperties["connection_info"]["database"]
result = arcpy.management.CreateFeatureclass(workspace, f"{lyr.name}_line", 'POLYLINE', lyr)
line_path = result.getOutput(0)
# Get the polygon features and convert to lines
features = [list(row) for row in arcpy.da.SearchCursor(lyr, flds)]
for ftr in features:
geom = ftr[shape_index]
ftr[shape_index] = arcpy.AsShape(geom.JSON.replace("rings", "paths").replace("Rings", "Paths"), True)
# Insert new line features into new Line Feature Class
with arcpy.da.InsertCursor(line_path, flds) as cursor:
for ftr in features:
cursor.insertRow(ftr)
The reason I used the json for the geometry is because this version accounts for multipart features and curves just by changing the "rings" to "paths" and the "curveRings" to "curvePaths".
One other thing is that I did a bit of substitution to change the Shape field to "SHAPE@" and the OBJECTID to "OID@". "Shape@" and "OID@" are keywords for the different cursors and it's easier to work with those so you don't have to worry about the actual name, which can be apparent in the beginning where we get the names from the Describe object.
