block signals before fork()ing the pipestream child

[libi2ncommon] / src / restricted_html.cpp

/*
The software in this package is distributed under the GNU General
Public License version 2 (with a special exception described below).

A copy of GNU General Public License (GPL) is included in this distribution,
in the file COPYING.GPL.

As a special exception, if other files instantiate templates or use macros
or inline functions from this file, or you compile this file and link it
with other works to produce a work based on this file, this file
does not by itself cause the resulting work to be covered
by the GNU General Public License.

However the source code for this file must still be made available
in accordance with section (3) of the GNU General Public License.

This exception does not invalidate any other reasons why a work based
on this file might be covered by the GNU General Public License.
*/
/** @file
 * @brief restricts html messages to an allowed group of tags.
 *
 * @copyright © Copyright 2017 Intra2net AG
 *
 */

#include <string>
#include <sstream>
#include <iomanip>

#include <stringfunc.hxx>
#include <restricted_html.hpp>

using namespace std;


namespace I2n
{



/**
 * @brief Replace all "+" characters found in s to spaces (" ").
 *
 * @param s string that will be modified.
 */
static void unescape_space(string &s)
{
    string::size_type pos;
    while ((pos=s.find('+')) != string::npos)
        s[pos]=' ';
}

/**
 * @brief Converts a hexadecimal sequence to its respective character.
 *
 * @param s string of size 2. Example: "77"
 * @return respective character represented by the hex sequence.
 */
static char x2c(const string& s)
{
    char digit;
    digit=(s[0]>='A' ? ((s[0] & 0xdf)-'A')+10 : (s[0]-'0'));
    digit*=16;
    digit+=(s[1]>='A' ? ((s[1] & 0xdf)-'A')+10 : (s[1]-'0'));
    return digit;
}

/**
 * @brief Scan a string to find escaped hex chars in the format "%HH" and replace
 * for their respective character.
 * Example: "www%2E" becomes "www."
 *
 * @param s String that will be modified.
 */
static void unescape_hex(string& s)
{
    static char hex_escape='%';
    string::size_type escape_pos;
    string hex_seq;
    string rest=s;
    for (s=""; ((escape_pos=rest.find(hex_escape)) != string::npos);)
    {
        if (escape_pos+2<static_cast<unsigned int>(rest.length())
            && ::isalnum(rest[escape_pos+1]) && ::isalnum(rest[escape_pos+2]))
        {
            hex_seq=rest.substr(escape_pos+1,2);
            s=s+rest.substr(0,escape_pos)+x2c(hex_seq);
            rest=rest.erase(0,escape_pos+3);
        }
        else
        {
            s=s+rest.substr(0,escape_pos+1);
            rest=rest.erase(0,escape_pos+1);
        }
    }
    s+=rest;
}

/**
* @brief Decode url that contains percent-encoding. Replace space " " with "+".
* Example: "%77%77%77%2E" becomes "www."
*
* @param s url string.
* @return the decoded string.
*/
string decode_url(string s)
{
    unescape_space (s);
    unescape_hex (s);
    return (s);
}

/**
 * @brief Verify if the parameter character requires encoding, If it is non
 * alphanumeric or valid ascii signs.
 *
 * @param c character to be verified.
 * @return true if the character should be encoded.
 */
bool needs_encoding (const char &c)
{
    // some valid ascii signs
    if (c == '_' || c == '-')
        return false;

    // is digit?
    if (c > 47 && c < 58)
        return false;

    // is uppercase letter?
    if (c > 64 && c < 91)
        return false;

    // is lowercase letter?
    if (c > 96 && c < 123)
        return false;

    return true;
}

/**
* @brief Encode url with percent-encoding. Any non-alphanumeric character is
* converted to its hex value with the percent character (%) as prefix, except "_"
* and "-". Replace space " " with "+".
*
* @param s url string.
* @return the encoded url string.
*/
string encode_url(string s)
{
    // convert non-alphanumeric characters to hex, convert space to +
    ostringstream out;
    for (string::iterator pos2=s.begin(); pos2 != s.end(); pos2++)
    {
        if (*pos2 == ' ')
            out << '+';
        else if (needs_encoding (*pos2))
            out << '%' << std::uppercase << setw(2) << setfill('0') << \
                std::hex << (int)(unsigned char)*pos2;
        else
            out << (*pos2);
    }

    return out.str();
}
} // eo namespace I2n