process_url: strip illegal characters in filename

There are edge cases when a retrived title contains illegal characters
(e.g. '/').
Replace it with a space to ensure a valid filename.

Reproducible case:
https://www.bilibili.com/video/av20827366/
(title="【2K/60fps】这可能是我做过最美的miku了【boomclap布料解算版】")

3 files changed, 22 insertions, 0 deletions

diff --git a/src/process_url.c b/src/process_url.c
index 4bfce8d..7a92013 100644
--- a/src/process_url.c
+++ b/src/process_url.c
@@ -29,6 +29,9 @@ extern Site_map site_map;
 Options options;
 static queue_t dl_queue;
 
+const char illegal_char[] = {'/', '\\', '|', '<', '>',
+                             ':', '"',  '?', '*', '\0'};
+
 thrd_t tid[MAX_THREAD];
 mtx_t mtx;
 cnd_t cnd;
@@ -516,6 +519,11 @@ void add_url(const char *URL, const char *outdir, const char *fn,
   } else {
     filename = malloc(strlen(fn) + 1);
     strcpy(filename, fn);
+    for (unsigned char i = 0; illegal_char[i] != '\0'; i++) {
+      if (repchr(filename, illegal_char[i], ' '))
+        DEBUG_PRINT("Found illegal character '%c' in filename, replacing ...\n",
+                    illegal_char[i]);
+    }
   }
 
   // Pass our cache (outdir_g) to parse_url()
diff --git a/src/utils.c b/src/utils.c
index 1132a94..ee41735 100644
--- a/src/utils.c
+++ b/src/utils.c
@@ -1,4 +1,5 @@
 #include <pcre2.h>
+#include <stddef.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -83,6 +84,17 @@ int regex_match(const char *subject, str_array_t patterns,
   return 0;
 }
 
+int repchr(char *str, char t, char r) {
+  int c = 0;
+  for (size_t i = 0; str[i] != '\0'; i++) {
+    if (str[i] == t) {
+      str[i] = r;
+      c++;
+    }
+  }
+  return c;
+}
+
 generic_array_t create_array(size_t elem_size, size_t n) {
   generic_array_t array;
   array.data = n ? malloc(elem_size * n) : NULL;
diff --git a/src/utils.h b/src/utils.h
index e4140a6..79c58ca 100644
--- a/src/utils.h
+++ b/src/utils.h
@@ -28,6 +28,8 @@ typedef struct queue {
 
 int regex_match(const char *, str_array_t, str_array_t *);
 
+int repchr(char *str, char t, char r);
+
 generic_array_t create_array(size_t elem_size, size_t n);
 
 void free_array(generic_array_t *array);