Fierelier/notepad-plus-plus-legacy
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

URL parser improvements

Browse Source 
- Removed special handling for file scheme.
- Improved handling of curly brackets.

Close #8975, close #8994
This commit is contained in:
Udo Hoffmann 2020-10-12 11:16:28 +02:00 committed by Don HO
parent 469fa62c0e
commit e32462188d
No known key found for this signature in database
GPG Key ID: 6C429F1D8D84F46E
1 changed files with 9 additions and 94 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

103
PowerEditor/src/Notepad_plus.cpp
View File

@ -2539,6 +2539,8 @@ bool isUrlTextChar(TCHAR const c)
case '\'': case '\'':
case '<': case '<':
case '>': case '>':
case '{':
case '}':
case '?': case '?':
case '\0x7f': case '\0x7f':
return false; return false;
@ -2670,6 +2672,11 @@ void scanToUrlEnd(TCHAR *text, int textLen, int start, int* distance)
q = text [p]; q = text [p];
s = sQueryQuotes; s = sQueryQuotes;
} }
else if (text [p] == '{')
{
q = '}';
s = sQueryQuotes;
}
else if (isUrlTextChar(text [p])) else if (isUrlTextChar(text [p]))
s = sQuery; s = sQuery;
else else
@ -2728,8 +2735,8 @@ bool removeUnwantedTrailingCharFromUrl (TCHAR const *text, int* length)
} }
} }
{ // remove unwanted closing parenthesis { // remove unwanted closing parenthesis
const TCHAR *closingParenthesis = L")]}>"; const TCHAR *closingParenthesis = L")]>";
const TCHAR *openingParenthesis = L"([{<"; const TCHAR *openingParenthesis = L"([<";
for (int i = 0; closingParenthesis [i]; i++) for (int i = 0; closingParenthesis [i]; i++)
if (text [l] == closingParenthesis [i]) if (text [l] == closingParenthesis [i])
{ {
@ -2768,93 +2775,6 @@ bool removeUnwantedTrailingCharFromUrl (TCHAR const *text, int* length)
return false; return false;
} }
bool isSlashOrBackslash(TCHAR const c)
{
return (c == '/') || (c == '\\');
}
bool isFilenameChar(TCHAR const c, bool const quoted)
{
if (c < ' ')
return false;
if ((c == ' ') && (!quoted))
return false;
switch (c)
{
case '"':
case '%':
case '*':
case '/':
case '<':
case '>':
case ':':
case '?':
case '|':
case '\\':
case '\0x7f':
return false;
}
return true;
}
// scanToFileEnd searches the end of an Filename, coarsly parsing it into prefix and name.
// The prefix parsing is done to avoid multiple colons.
// The <quoted> parameter specifies, whether spaces are allowed.
void scanToFileEnd(TCHAR *text, int textLen, int start, bool quoted, int* distance)
{
int p = start;
enum {sStart, sPrefix, sColon, sName} s = sStart;
while (p < textLen)
{
switch (s)
{
case sStart:
if (isFilenameChar (text [p], false))
s = sPrefix;
else if (!isSlashOrBackslash(text [p]))
{
*distance = p - start;
return;
}
break;
case sPrefix:
if (isSlashOrBackslash(text [p]) || isFilenameChar(text [p], quoted))
s = sName;
else if (text [p] == ':')
s = sColon;
else
{
*distance = p - start;
return;
}
break;
case sColon:
if (isSlashOrBackslash(text[p]))
s = sName;
else
{
*distance = p - start;
return;
}
break;
case sName:
if (! (isSlashOrBackslash(text [p]) || isFilenameChar(text [p], quoted)))
{
*distance = p - start;
return;
}
break;
}
p++;
}
*distance = p - start;
}
// isUrl checks, whether there is a valid URL at <text [start]>. // isUrl checks, whether there is a valid URL at <text [start]>.
// If yes: // If yes:
// - True is returned. // - True is returned.
@ -2885,11 +2805,6 @@ bool isUrl(TCHAR * text, int textLen, int start, int* segmentLen)
if (r) if (r)
{ {
while (removeUnwantedTrailingCharFromUrl (& text [start], & len)); while (removeUnwantedTrailingCharFromUrl (& text [start], & len));
if (url.nScheme == INTERNET_SCHEME_FILE)
{
scanToFileEnd (text, textLen, start + schemeLen, (start > 0) && (text [start - 1] == '"'), & len);
len += schemeLen;
}
*segmentLen = len; *segmentLen = len;
return true; return true;
} }