From 4093cb603c70de9af083cd1667e7fba6cc04edc2 Mon Sep 17 00:00:00 2001
From: flavis <christian@flavigny.de>
Date: Tue, 10 Aug 2021 11:22:29 +0200
Subject: [PATCH] add further checks for invalid records

---
 R/parse.R | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/R/parse.R b/R/parse.R
index 4d59ffc..58cc3df 100644
--- a/R/parse.R
+++ b/R/parse.R
@@ -76,6 +76,13 @@ read_one <- function(name, path) {
     verlauf <- xml_find_first(x, "sitzungsverlauf")
     speakerl <- xml_find_first(x, "rednerliste")
 
+    # check if record is invalid or empty (every record should have at least
+    # one speech, a speaker and a date
+    if (is.na(date) || length(verlauf) == 0 || length(speakerl) == 0) {
+        warning("Invalid record found. Skipping.")
+        return(NULL)
+    }
+
     xml_children(speakerl) %>%
         parse_speakerlist() ->
         speaker