From 308e2d271fb2636e0f8100605691deede9713857 Mon Sep 17 00:00:00 2001
From: tastytea <tastytea@tastytea.de>
Date: Fri, 28 May 2021 13:55:11 +0200
Subject: [PATCH] Skip rest of file if encoding of files in EPUB is broken.

Standard says UTF-8. I don't want to deal with weird Windows-encodings or
whatever this is.

Closes: https://schlomp.space/tastytea/epubgrep/issues/7
---
 src/zip.cpp | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

diff --git a/src/zip.cpp b/src/zip.cpp
index e4765d6..28be71f 100644
--- a/src/zip.cpp
+++ b/src/zip.cpp
@@ -27,6 +27,7 @@
 #include <cstdlib>
 #include <cstring>
 #include <fstream>
+#include <iostream>
 #include <stdexcept>
 #include <string>
 #include <vector>
@@ -45,7 +46,18 @@ std::vector<std::string> list(const fs::path &filepath)
     std::vector<std::string> toc;
     while (archive_read_next_header(zipfile, &entry) == ARCHIVE_OK)
     {
-        toc.emplace_back(archive_entry_pathname_utf8(entry));
+        const auto *in_epub_filepath{archive_entry_pathname_utf8(entry)};
+        if (in_epub_filepath == nullptr)
+        { // If the encoding is broken, we skip the file.
+            std::cerr << translate("WARNING: ")
+                      << format(translate("{0:s} is damaged. "
+                                          "Skipping rest of file.\n")
+                                    .str()
+                                    .data(),
+                                filepath);
+            break;
+        }
+        toc.emplace_back(in_epub_filepath);
         archive_read_data_skip(zipfile);
     }