Fixed handling of update URLs which mention a port

This appears to have been a regression in
f405849b49614254f42eb3ee6147434458978623, which for some reason erased
the port (and any trailing parts) from not just the update directory but
also from the update URL.

Unfortunately we can't access Mana-Mantis #381 at the moment, but
presumably the port was removed due to the colon being a problematic
character. Instead, now the colon (and other special characters) are
replaced by _ when determining the update directory.

Closes https://git.themanaworld.org/mana/mana/-/issues/80

2 files changed, 32 insertions, 54 deletions

diff --git a/src/utils/stringutils.cpp b/src/utils/stringutils.cpp
index 747243fd..0244dd8c 100644
--- a/src/utils/stringutils.cpp
+++ b/src/utils/stringutils.cpp
@@ -101,16 +101,15 @@ std::string strprintf(char const *format, ...)
     return res;
 }
 
-std::string &removeBadChars(std::string &str)
+std::string &replaceCharacters(std::string &str,
+                               std::string_view chars,
+                               char replacement)
 {
-    std::string::size_type pos;
-    do
+    for (auto &c : str)
     {
-        pos = str.find_first_of("@#[]");
-        if (pos != std::string::npos)
-            str.erase(pos, 1);
-    } while (pos != std::string::npos);
-
+        if (chars.find(c) != std::string::npos)
+            c = replacement;
+    }
     return str;
 }
 
@@ -220,49 +219,27 @@ std::string normalize(const std::string &name)
     return toLower(trim(normalized));
 }
 
-std::string removeTrailingSymbol(const std::string &s, const char c)
+std::string getDirectoryFromURL(const std::string &url)
 {
-    // Remove the trailing symblol at the end of the string
-    if (!s.empty() && s.at(s.size() - 1) == c)
-        return s.substr(0, s.size() - 1);
-    return std::string(s);
-}
-
-std::string getHostNameFromURL(const std::string &url)
-{
-    std::string myHostName;
-
-    // Don't go out of range in the next check
-    if (url.length() < 2)
-        return myHostName;
+    std::string directory = url;
 
-    // Remove any trailing slash at the end of the update host
-    myHostName = removeTrailingSymbol(url, '/');
+    // Parse out any "http://", "ftp://", etc...
+    size_t pos = directory.find("://");
+    if (pos != std::string::npos)
+        directory.erase(0, pos + 3);
 
-    // Parse out any "http://", "ftp://", ect...
-    size_t pos = myHostName.find("://");
-    if (pos == std::string::npos)
-    {
-        logger->log("Warning: no protocol was specified for the url: %s",
-                    url.c_str());
-    }
+    // Replace characters which are not valid or difficult in file system paths
+    replaceCharacters(directory, ":*?\"<>| ", '_');
 
-    if (myHostName.empty() || pos + 3 >= myHostName.length())
+    // Replace ".." (double dots) with "_" to avoid directory traversal.
+    pos = directory.find("..");
+    while (pos != std::string::npos)
     {
-        logger->log("Error: Invalid url: %s", url.c_str());
-        return myHostName;
+        directory.replace(pos, 2, "_");
+        pos = directory.find("..");
     }
-    myHostName = myHostName.substr(pos + 3);
-
-    // Remove possible trailing port (i.e.: localhost:8000 -> localhost)
-    pos = myHostName.find(":");
-    if (pos != std::string::npos)
-        myHostName = myHostName.substr(0, pos);
-
-    // remove possible other junk
-    removeBadChars(myHostName);
 
-    return myHostName;
+    return directory;
 }
 
 std::string join(const std::vector<std::string> &strings, const char *separator)
diff --git a/src/utils/stringutils.h b/src/utils/stringutils.h
index ca54803d..349d2a66 100644
--- a/src/utils/stringutils.h
+++ b/src/utils/stringutils.h
@@ -95,12 +95,16 @@ std::string strprintf(char const *, ...)
 ;
 
 /**
- * Removes bad characters from a string
+ * Replaces a set of characters with another character.
  *
  * @param str the string to remove the bad chars from
+ * @param chars the bad characters to remove
+ * @param replacement the character to replace the bad chars with
  * @return a reference to the string without bad chars
  */
-std::string &removeBadChars(std::string &str);
+std::string &replaceCharacters(std::string &str,
+                               std::string_view chars,
+                               char replacement = '_');
 
 /**
  * Removes colors from a string
@@ -210,15 +214,12 @@ std::string autocomplete(const std::vector<std::string> &candidates,
 std::string normalize(const std::string &name);
 
 /**
- * Remove a potential trailing symbol from a string.
- */
-std::string removeTrailingSymbol(const std::string &s, const char c);
-
-/**
- * Gets the hostname out of the given URL.
- * i.e.: http://www.manasource.org:9601 -> www.manasource.org
+ * Derives a directory from the given URL, stripping the schema and replacing
+ * certain invalid characters.
+ *
+ * i.e.: http://www.manasource.org:9601/updates/ -> www.manasource.org_9601/updates/
  */
-std::string getHostNameFromURL(const std::string &url);
+std::string getDirectoryFromURL(const std::string &url);
 
 /**
  * Joins a vector of strings into one string, separated by the given