<?php

class blcPlaintextUrlBase extends blcParser {
	var $supported_formats = array( 'html', 'plaintext' );

	//Regexp for detecting plaintext URLs lifted from make_clickable()
	var $url_regexp = '#(?<=[\s>\]]|^)(\()?([\w]+?://(?:[\w\\x80-\\xff\#$%&~/=?@\[\](+-]|[.,;:](?![\s<]|(\))?([\s]|$))|(?(1)\)(?![\s<.,;:]|$)|\)))+)#is';

	//Used by the edit and unlink callbacks
	var $old_url = '';
	var $new_url = '';

	/**
	 * Parse a string for plaintext URLs
	 *
	 * @param string $content The text to parse.
	 * @param string $base_url The base URL. Ignored.
	 * @param string $default_link_text Default link text.
	 * @return array An array of new blcLinkInstance objects.
	 */
	function parse( $content, $base_url = '', $default_link_text = '' ) {
		//Don't want to detect URLs inside links or tag attributes -
		//there are already other parsers for that.

		//Avoid <a href="http://...">http://...</a>
		$content = preg_replace( '#<a[^>]*>.*?</a>#si', '', $content );
		//HTML tags are treated as natural boundaries for plaintext URLs
		//(since we strip tags, we must place another boundary char where they were).
		//The closing tag of [shortcodes] is also treated as a boundary.
		$content = str_replace( array( '<', '>', '[/' ), array( "\n<", ">\n", "\n[/" ), $content );
		//Finally, kill all tags.
		$content = strip_tags( $content );

		//Find all URLs
		$found = preg_match_all(
			$this->url_regexp,
			$content,
			$matches
		);

		$instances = array();

		if ( $found ) {
			//Create a new instance for each match
			foreach ( $matches[2] as $match ) {
				$url = $this->validate_url( trim( $match ) );
				if ( false == $url ) {
					continue;
				}

				//Create a new link instance.
				$instance = new blcLinkInstance();

				$instance->set_parser( $this );
				$instance->raw_url   = $match;
				$instance->link_text = $match;

				$link_obj = new blcLink( $url ); //Creates or loads the link
				$instance->set_link( $link_obj );

				$instances[] = $instance;
			}
		}

		return $instances;
	}

	/**
	 * Validate and sanitize a URL.
	 *
	 * @param string $url
	 * @return bool|string A valid URL, or false if the URL is not valid.
	 */
	protected function validate_url( $url ) {
		//Do a little bit of validation
		$url = esc_url_raw( $url );
		if ( empty( $url ) ) {
			return false;
		}
		if ( function_exists( 'filter_var' ) ) {
			//Note: filter_var() is no panacea as it accepts many invalid URLs
			if ( ! filter_var( $url, FILTER_VALIDATE_URL ) ) {
				return false;
			}
		}
		$parts = @parse_url( $url );
		if ( empty( $parts['host'] ) || ! strpos( $parts['host'], '.' ) ) {
			return false;
		}
		return $url;
	}

	/**
	 * Change all occurrences of a given plaintext URLs to a new URL.
	 *
	 * @param string $content Look for URLs in this string.
	 * @param string $new_url Change them to this URL.
	 * @param string $old_url The URL to look for.
	 * @param string $old_raw_url The raw, not-normalized URL. Optional.
	 *
	 * @return array|WP_Error If successful, the return value will be an associative array with two
	 * keys : 'content' - the modified content, and 'raw_url' - the new raw, non-normalized URL used
	 * for the modified links. In most cases, the returned raw_url will be equal to the new_url.
	 */
	function edit( $content, $new_url, $old_url, $old_raw_url = '' ) {
		$this->new_url = $new_url;
		if ( empty( $old_raw_url ) ) {
			$this->old_url = $old_url;
		} else {
			$this->old_url = $old_raw_url;
		}

		return array(
			'content'   => preg_replace_callback( $this->url_regexp, array( &$this, 'edit_callback' ), $content ),
			'raw_url'   => $new_url,
			'link_text' => $new_url,
		);
	}

	function edit_callback( $match ) {
		if ( $match[2] == $this->old_url ) {
			return $this->new_url;
		} else {
			return $match[0];
		}
	}


	/**
	 * Remove all occurrences of a specific plaintext URL.
	 *
	 * @param string $content   Look for URLs in this string.
	 * @param string $url The URL to look for.
	 * @param string $raw_url The raw, non-normalized version of the URL to look for. Optional.
	 * @return string Input string with all matching plaintext URLs removed.
	 */
	function unlink( $content, $url, $raw_url = '' ) {
		if ( empty( $raw_url ) ) {
			$this->old_url = $url;
		} else {
			$this->old_url = $raw_url;
		}

		return preg_replace_callback( $this->url_regexp, array( &$this, 'unlink_callback' ), $content );
	}

	function unlink_callback( $match ) {
		if ( $match[2] == $this->old_url ) {
			return '';
		} else {
			return $match[0];
		}
	}
}