libpurple/util.c

/*
 * Purple - Internet Messaging Library
 * Copyright (C) Pidgin Developers <devel@pidgin.im>
 *
 * Purple is the legal property of its developers, whose names are too numerous
 * to list here. Please refer to the COPYRIGHT file distributed with this
 * source distribution.
 *
 * This library is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License as published by the Free
 * Software Foundation; either version 2 of the License, or (at your option)
 * any later version.
 *
 * This library is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this library; if not, see <https://www.gnu.org/licenses/>.
 */

#include <purpleconfig.h>

#include <glib/gi18n-lib.h>

#include <ctype.h>

#include "purpleprivate.h"

#include "core.h"
#include "debug.h"
#include "prefs.h"
#include "purpleaccountmanager.h"
#include "purpleconversation.h"
#include "purplepath.h"
#include "purpleprotocol.h"
#include "signals.h"
#include "util.h"

#ifdef _WIN32
# include "win32/libc_interface.h"
#endif

#include <json-glib/json-glib.h>

#define BUF_LEN 2048

void
purple_util_init(void) {
}

void
purple_util_uninit(void) {
	purple_util_set_user_dir(NULL);
}

/**************************************************************************
 * Path/Filename Functions
 **************************************************************************/
static gboolean
purple_util_write_data_to_file_common(const char *dir, const char *filename, const char *data, gssize size)
{
	gchar *filename_full;
	gboolean ret = FALSE;

	g_return_val_if_fail(dir != NULL, FALSE);

	purple_debug_misc("util", "Writing file %s to directory %s",
			  filename, dir);

	/* Ensure the directory exists */
	if (!g_file_test(dir, G_FILE_TEST_IS_DIR))
	{
		if (g_mkdir(dir, S_IRUSR | S_IWUSR | S_IXUSR) == -1)
		{
			purple_debug_error("util", "Error creating directory %s: %s\n",
					   dir, g_strerror(errno));
			return FALSE;
		}
	}

	filename_full = g_build_filename(dir, filename, NULL);
	ret = g_file_set_contents(filename_full, data, size, NULL);
	g_free(filename_full);

	return ret;
}

gboolean
purple_util_write_data_to_config_file(const char *filename, const char *data, gssize size)
{
	const char *config_dir = purple_config_dir();
	gboolean ret = purple_util_write_data_to_file_common(config_dir, filename, data, size);

	return ret;
}

PurpleXmlNode *
purple_util_read_xml_from_config_file(const char *filename, const char *description)
{
	return purple_xmlnode_from_file(purple_config_dir(), filename, description, "util");
}

gboolean
purple_running_gnome(void)
{
#ifndef _WIN32
	gchar *tmp = g_find_program_in_path("gvfs-open");

	if (tmp == NULL) {
		tmp = g_find_program_in_path("gnome-open");

		if (tmp == NULL) {
			return FALSE;
		}
	}

	g_free(tmp);

	tmp = (gchar *)g_getenv("GNOME_DESKTOP_SESSION_ID");

	return ((tmp != NULL) && (*tmp != '\0'));
#else
	return FALSE;
#endif
}

/**************************************************************************
 * String Functions
 **************************************************************************/
void
purple_str_strip_char(char *text, char thechar)
{
	int i, j;

	g_return_if_fail(text != NULL);

	for (i = 0, j = 0; text[i]; i++)
		if (text[i] != thechar)
			text[j++] = text[i];

	text[j] = '\0';
}

gchar *
purple_strreplace(const char *string, const char *delimiter,
				const char *replacement)
{
	gchar **split;
	gchar *ret;

	g_return_val_if_fail(string      != NULL, NULL);
	g_return_val_if_fail(delimiter   != NULL, NULL);
	g_return_val_if_fail(replacement != NULL, NULL);

	split = g_strsplit(string, delimiter, 0);
	ret = g_strjoinv(replacement, split);
	g_strfreev(split);

	return ret;
}

void
purple_str_wipe(gchar *str)
{
	if (str == NULL)
		return;
	memset(str, 0, strlen(str));
	g_free(str);
}

gboolean
purple_strmatches(const char *pattern, const char *str) {
	char *normal_pattern = NULL;
	char *normal_str = NULL;
	char *cmp_pattern = NULL;
	char *cmp_str = NULL;
	char *idx_pattern = NULL;
	char *idx_str = NULL;

	g_return_val_if_fail(pattern != NULL, FALSE);

	/* Short circuit on NULL and empty string. */
	if(purple_strempty(str)) {
		return FALSE;
	}

	normal_pattern = g_utf8_normalize(pattern, -1, G_NORMALIZE_ALL);
	cmp_pattern = g_utf8_casefold(normal_pattern, -1);
	g_free(normal_pattern);

	normal_str = g_utf8_normalize(str, -1, G_NORMALIZE_ALL);
	cmp_str = g_utf8_casefold(normal_str, -1);
	g_free(normal_str);

	idx_pattern = cmp_pattern;
	idx_str = cmp_str;

	/* I know while(TRUE)'s suck, but the alternative would be a multi-line for
	 * loop that wouldn't have the additional comments, which is much better
	 * IMHO. -- GK 2023-01-24.
	 */
	while(TRUE) {
		gunichar character = g_utf8_get_char(idx_pattern);

		/* If we've reached the end of the pattern, we're done. */
		if(character == 0) {
			break;
		}

		idx_str = g_utf8_strchr(idx_str, -1, character);
		if(idx_str == NULL) {
			g_free(cmp_pattern);
			g_free(cmp_str);

			return FALSE;
		}

		/* We found the current character in pattern, so move to the next. */
		idx_pattern = g_utf8_next_char(idx_pattern);

		/* move idx_str to the next character as well, because the current
		 * character has already been matched.
		 */
		idx_str = g_utf8_next_char(idx_str);
	};

	g_free(cmp_pattern);
	g_free(cmp_str);

	return TRUE;
}

/**************************************************************************
 * URI/URL Functions
 **************************************************************************/

void purple_got_protocol_handler_uri(const char *uri)
{
	char proto[11];
	char delimiter;
	const char *tmp, *param_string;
	char *cmd;
	GHashTable *params = NULL;
	gsize len;
	if (!(tmp = strchr(uri, ':')) || tmp == uri) {
		purple_debug_error("util", "Malformed protocol handler message - missing protocol.\n");
		return;
	}

	len = MIN(sizeof(proto) - 1, (gsize)(tmp - uri));

	strncpy(proto, uri, len);
	proto[len] = '\0';

	tmp++;

	if (purple_strequal(proto, "xmpp"))
		delimiter = ';';
	else
		delimiter = '&';

	purple_debug_info("util", "Processing message '%s' for protocol '%s' using delimiter '%c'.\n", tmp, proto, delimiter);

	if ((param_string = strchr(tmp, '?'))) {
		const char *keyend = NULL, *pairstart;
		char *key, *value = NULL;

		cmd = g_strndup(tmp, (param_string - tmp));
		param_string++;

		params = g_hash_table_new_full(g_str_hash, g_str_equal, g_free, g_free);
		pairstart = tmp = param_string;

		while (*tmp || *pairstart) {
			if (*tmp == delimiter || !(*tmp)) {
				/* If there is no explicit value */
				if (keyend == NULL) {
					keyend = tmp;
				}
				/* without these brackets, clang won't
				 * recognize tmp as a non-NULL
				 */

				if (keyend && keyend != pairstart) {
					char *p;
					key = g_strndup(pairstart, (keyend - pairstart));
					/* If there is an explicit value */
					if (keyend != tmp && keyend != (tmp - 1))
						value = g_strndup(keyend + 1, (tmp - keyend - 1));
					for (p = key; *p; ++p)
						*p = g_ascii_tolower(*p);
					g_hash_table_insert(params, key, value);
				}
				keyend = value = NULL;
				pairstart = (*tmp) ? tmp + 1 : tmp;
			} else if (*tmp == '=')
				keyend = tmp;

			if (*tmp)
				tmp++;
		}
	} else
		cmd = g_strdup(tmp);

	purple_signal_emit_return_1(purple_get_core(), "uri-handler", proto, cmd, params);

	g_free(cmd);
	g_clear_pointer(&params, g_hash_table_destroy);
}

/* Originally lifted from
 * http://www.oreillynet.com/pub/a/network/excerpt/spcookbook_chap03/index3.html
 * ... and slightly modified to be a bit more rfc822 compliant
 * ... and modified a bit more to make domain checking rfc1035 compliant
 *     with the exception permitted in rfc1101 for domains to start with digit
 *     but not completely checking to avoid conflicts with IP addresses
 */
gboolean
purple_email_is_valid(const char *address)
{
	const char *c, *domain;
	static char *rfc822_specials = "()<>@,;:\\\"[]";

	g_return_val_if_fail(address != NULL, FALSE);

	if (*address == '.') return FALSE;

	/* first we validate the name portion (name@domain) (rfc822)*/
	for (c = address;  *c;  c++) {
		if (*c == '\"' && (c == address || *(c - 1) == '.' || *(c - 1) == '\"')) {
			while (*++c) {
				if (*c == '\\') {
					if (*c++ && *c < 127 && *c > 0 && *c != '\n' && *c != '\r') continue;
					else return FALSE;
				}
				if (*c == '\"') break;
				if (*c < ' ' || *c >= 127) return FALSE;
			}
			if (!*c++) return FALSE;
			if (*c == '@') break;
			if (*c != '.') return FALSE;
			continue;
		}
		if (*c == '@') break;
		if (*c <= ' ' || *c >= 127) return FALSE;
		if (strchr(rfc822_specials, *c)) return FALSE;
	}

	/* It's obviously not an email address if we didn't find an '@' above */
	if (*c == '\0') return FALSE;

	/* strictly we should return false if (*(c - 1) == '.') too, but I think
	 * we should permit user.@domain type addresses - they do work :) */
	if (c == address) return FALSE;

	/* next we validate the domain portion (name@domain) (rfc1035 & rfc1011) */
	if (!*(domain = ++c)) return FALSE;
	do {
		if (*c == '.' && (c == domain || *(c - 1) == '.' || *(c - 1) == '-'))
			return FALSE;
		if (*c == '-' && (*(c - 1) == '.' || *(c - 1) == '@')) return FALSE;
		if ((*c < '0' && *c != '-' && *c != '.') || (*c > '9' && *c < 'A') ||
			(*c > 'Z' && *c < 'a') || (*c > 'z')) return FALSE;
	} while (*++c);

	if (*(c - 1) == '-') return FALSE;

	return ((c - domain) > 3 ? TRUE : FALSE);
}

/**************************************************************************
 * UTF8 String Functions
 **************************************************************************/
gchar *
purple_utf8_try_convert(const char *str)
{
	gsize converted;
	gchar *utf8;

	g_return_val_if_fail(str != NULL, NULL);

	if (g_utf8_validate(str, -1, NULL)) {
		return g_strdup(str);
	}

	utf8 = g_locale_to_utf8(str, -1, &converted, NULL, NULL);
	if (utf8 != NULL)
		return utf8;

	utf8 = g_convert(str, -1, "UTF-8", "ISO-8859-15", &converted, NULL, NULL);
	if ((utf8 != NULL) && (converted == strlen(str)))
		return utf8;

	g_free(utf8);

	return NULL;
}

int
purple_utf8_strcasecmp(const char *a, const char *b)
{
	char *a_norm = NULL;
	char *b_norm = NULL;
	int ret = -1;

	if(!a && b)
		return -1;
	else if(!b && a)
		return 1;
	else if(!a && !b)
		return 0;

	if(!g_utf8_validate(a, -1, NULL) || !g_utf8_validate(b, -1, NULL))
	{
		purple_debug_error("purple_utf8_strcasecmp",
						 "One or both parameters are invalid UTF8\n");
		return ret;
	}

	a_norm = g_utf8_casefold(a, -1);
	b_norm = g_utf8_casefold(b, -1);
	ret = g_utf8_collate(a_norm, b_norm);
	g_free(a_norm);
	g_free(b_norm);

	return ret;
}

/* this is almost identical to purple_url_encode (hence purple_url_decode
 * being used above), but we want to keep certain characters unescaped
 * for compat reasons */
const char *
purple_escape_filename(const char *str)
{
	const char *iter;
	static char buf[BUF_LEN];
	char utf_char[6];
	guint i, j = 0;

	g_return_val_if_fail(str != NULL, NULL);
	g_return_val_if_fail(g_utf8_validate(str, -1, NULL), NULL);

	iter = str;
	for (; *iter && j < (BUF_LEN - 1) ; iter = g_utf8_next_char(iter)) {
		gunichar c = g_utf8_get_char(iter);
		/* If the character is an ASCII character and is alphanumeric,
		 * or one of the specified values, no need to escape */
		if (c < 128 && (g_ascii_isalnum(c) || c == '@' || c == '-' ||
				c == '_' || c == '.' || c == '#')) {
			buf[j++] = c;
		} else {
			int bytes = g_unichar_to_utf8(c, utf_char);
			for (i = 0; (int)i < bytes; i++) {
				if (j > (BUF_LEN - 4))
					break;
				if (i >= sizeof(utf_char)) {
					g_warn_if_reached();
					break;
				}
				sprintf(buf + j, "%%%02x", utf_char[i] & 0xff);
				j += 3;
			}
		}
	}
#ifdef _WIN32
	/* File/Directory names in windows cannot end in periods/spaces.
	 * http://msdn.microsoft.com/en-us/library/aa365247%28VS.85%29.aspx
	 */
	while (j > 0 && (buf[j - 1] == '.' || buf[j - 1] == ' '))
		j--;
#endif
	buf[j] = '\0';

	return buf;
}