parseCSV 0.4.1 beta

git-svn-id: http://parsecsv-for-php.googlecode.com/svn/trunk@31 339761fc-0c37-0410-822d-8b8cac1f6a97

ChangeLog.txt

@@ -1,3 +1,81 @@
+parseCSV 0.4.1 beta
+-----------------------------------
+Date: 29-May-2008
+
+- Fixed a small bug in how the output() method
+  handles input data.
+
+-----------------------------------
+
+
+parseCSV 0.4 beta
+-----------------------------------
+Date: 11-Apr-2008
+
+- Error reporting for files/data which is corrupt
+  or has formatting errors like using double
+  quotes in a field without enclosing quotes. Or
+  not escaping double quotes with a second one.
+
+- parse() method does not require input anymore
+  if the "$object->file" property has been set.
+
+I'm calling this a beta release due to the heavy
+modifications to the core parsing logic required
+for error reporting to work. I have tested the
+new code quite extensively, I'm fairly confident
+that it still parses exactly as it always has.
+
+The second reason I'm calling it a beta release
+is cause I'm sure the error reporting code will
+need more refinements and tweaks to detect more
+types of errors, as it's only picking two types
+or syntax errors right now. However, it seems
+these two are the most common errors that you
+would be likely to come across.
+
+-----------------------------------
+
+
+parseCSV 0.3.2
+-----------------------------------
+Date: 1-Apr-2008
+
+This is primarily a bug-fix release for a critical
+bug which was brought to my attention.
+
+- Fixed a critical bug in conditions parsing which
+  would generate corrupt matching patterns causing
+  the condition(s) to not work at all in some
+  situations.
+
+- Fixed a small code error which would cause PHP to
+  generate a invalid offset notice when zero length
+  values were fed into the unparse() method to
+  generate CSV data from an array.
+
+Notice: If you have been using the "parsecsv-stable"
+branch as an external in any of your projects,
+please use the "stable/parsecsv" branch from this
+point on as I will eventually remove the former due
+to it's stupid naming.
+
+-----------------------------------
+
+
+parseCSV 0.3.1
+-----------------------------------
+Date: 1-Sep-2007
+
+- Small change to default output settings to
+  conform with RFC 4180 (http://rfc.net/rfc4180.html).
+  Only the LF (line feed) character was used
+  by default to separate rows, rather than
+  CRLF (carriage return & line feed).
+
+-----------------------------------
+
+
 parseCSV 0.3.0
 -----------------------------------
 Date: 9-Aug-2007
@@ -18,6 +96,9 @@ Date: 9-Aug-2007
 - Minor changes and optimizations, and a few
   spelling corrections. Oops :)
 
+- Included more complex code examples in the
+  parseCSV download.
+
 -----------------------------------
 
 

examples/_books.csv

@@ -1,15 +1,15 @@
-rating,title,author,type,asin,tags,review
+rating,title,author,type,asin,tags,review
-0,The Killing Kind,John Connolly,Book,0340771224,,i still haven't had time to read this one...
+0,The Killing Kind,John Connolly,Book,0340771224,,i still haven't had time to read this one...
-0,The Third Secret,Steve Berry,Book,0340899263,,need to find time to read this book
+0,The Third Secret,Steve Berry,Book,0340899263,,need to find time to read this book
-3,The Last Templar,Raymond Khoury,Book,0752880705,,
+3,The Last Templar,Raymond Khoury,Book,0752880705,,
-5,The Traveller,John Twelve Hawks,Book,059305430X,,
+5,The Traveller,John Twelve Hawks,Book,059305430X,,
-4,Crisis Four,Andy Mcnab,Book,0345428080,,
+4,Crisis Four,Andy Mcnab,Book,0345428080,,
-5,Prey,Michael Crichton,Book,0007154534,,
+5,Prey,Michael Crichton,Book,0007154534,,
-3,The Broker (Paperback),John Grisham,Book,0440241588,book johngrisham,"good book, but is slow in the middle"
+3,The Broker (Paperback),John Grisham,Book,0440241588,book johngrisham,"good book, but is slow in the middle"
-3,Without Blood (Paperback),Alessandro Baricco,Book,1841955744,,
+3,Without Blood (Paperback),Alessandro Baricco,Book,1841955744,,
-5,State of Fear (Paperback),Michael Crichton,Book,0061015733,,
+5,State of Fear (Paperback),Michael Crichton,Book,0061015733,,
-4,The Rule of Four (Paperback),Ian Caldwell,Book,0099451956,book bestseller,
+4,The Rule of Four (Paperback),Ian Caldwell,Book,0099451956,book bestseller,
-4,Deception Point (Paperback),Dan Brown,Book,0671027387,book danbrown bestseller,
+4,Deception Point (Paperback),Dan Brown,Book,0671027387,book danbrown bestseller,
-5,Digital Fortress : A Thriller (Mass Market Paperback),Dan Brown,Book,0312995423,book danbrown bestseller,
+5,Digital Fortress : A Thriller (Mass Market Paperback),Dan Brown,Book,0312995423,book danbrown bestseller,
-5,Angels & Demons (Mass Market Paperback),Dan Brown,Book,0671027360,book danbrown bestseller,
+5,Angels & Demons (Mass Market Paperback),Dan Brown,Book,0671027360,book danbrown bestseller,
 4,The Da Vinci Code (Hardcover),Dan Brown,"   Book   ",0385504209,book movie danbrown bestseller davinci,

examples/basic.php

@@ -13,7 +13,11 @@ $csv = new parseCSV();
 # Parse '_books.csv' using automatic delimiter detection...
 $csv->auto('_books.csv');
 
-# ...or if you know the delimiter, use the parse() function.
+# ...or if you know the delimiter, set the delimiter character
+# if its not the default comma...
+// $csv->delimiter = "\t";   # tab delimited
+
+# ...and then use the parse() function.
 // $csv->parse('_books.csv');
 
 

parsecsv.lib.php

@@ -4,7 +4,7 @@ class parseCSV {
 	
 /*
 
-	Class: parseCSV v0.3.0
+	Class: parseCSV v0.4.1 beta
 	http://code.google.com/p/parsecsv-for-php/
 	
 	
@@ -123,7 +123,7 @@ class parseCSV {
 	var $output_encoding = 'ISO-8859-1';
 	
 	# used by unparse(), save(), and output() functions
-	var $linefeed = "\n";
+	var $linefeed = "\r\n";
 	
 	# only used by output() function
 	var $output_delimiter = ',';
@@ -140,6 +140,19 @@ class parseCSV {
 	# loaded file contents
 	var $file_data;
 	
+	# error while parsing input data
+	#  0 = No errors found. Everything should be fine :)
+	#  1 = Hopefully correctable syntax error was found.
+	#  2 = Enclosure character (double quote by default)
+	#      was found in non-enclosed field. This means
+	#      the file is either corrupt, or does not
+	#      standard CSV formatting. Please validate
+	#      the parsed data yourself.
+	var $error = 0;
+	
+	# detailed error info
+	var $error_info = array();
+	
 	# array of field values in data parsed
 	var $titles = array();
 	
@@ -170,6 +183,7 @@ class parseCSV {
 	 * @return  nothing
 	 */
 	function parse ($input = null, $offset = null, $limit = null, $conditions = null) {
+		if ( $input === null ) $input = $this->file;
 		if ( !empty($input) ) {
 			if ( $offset !== null ) $this->offset = $offset;
 			if ( $limit !== null ) $this->limit = $limit;
@@ -272,13 +286,13 @@ class parseCSV {
 			$pch = ( isset($data{$i-1}) ) ? $data{$i-1} : false ;
 			
 			// open and closing quotes
-			if ( $ch == $enclosure && (!$enclosed || $nch != $enclosure) ) {
+			if ( $ch == $enclosure ) {
-				$enclosed = ( $enclosed ) ? false : true ;
+				if ( !$enclosed || $nch != $enclosure ) {
-			
+					$enclosed = ( $enclosed ) ? false : true ;
-			// inline quotes	
+				} elseif ( $enclosed ) {
-			} elseif ( $ch == $enclosure && $enclosed ) {
+					$i++;
-				$i++;
+				}
-
+				
 			// end of row
 			} elseif ( ($ch == "\n" && $pch != "\r" || $ch == "\r") && !$enclosed ) {
 				if ( $n >= $search_depth ) {
@@ -311,13 +325,12 @@ class parseCSV {
 		
 		// capture most probable delimiter
 		ksort($filtered);
-		$delimiter = reset($filtered);
+		$this->delimiter = reset($filtered);
-		$this->delimiter = $delimiter;
 		
 		// parse data
 		if ( $parse ) $this->data = $this->parse_string();
 		
-		return $delimiter;
+		return $this->delimiter;
 		
 	}
 	
@@ -349,6 +362,8 @@ class parseCSV {
 			} else return false;
 		}
 		
+		$white_spaces = str_replace($this->delimiter, '', " \t\x0B\0");
+		
 		$rows = array();
 		$row = array();
 		$row_count = 0;
@@ -365,24 +380,68 @@ class parseCSV {
 			$nch = ( isset($data{$i+1}) ) ? $data{$i+1} : false ;
 			$pch = ( isset($data{$i-1}) ) ? $data{$i-1} : false ;
 			
-			// open and closing quotes
+			// open/close quotes, and inline quotes
-			if ( $ch == $this->enclosure && (!$enclosed || $nch != $this->enclosure) ) {
+			if ( $ch == $this->enclosure ) {
-				$enclosed = ( $enclosed ) ? false : true ;
+				if ( !$enclosed ) {
-				if ( $enclosed ) $was_enclosed = true;
+					if ( ltrim($current, $white_spaces) == '' ) {
-			
+						$enclosed = true;
-			// inline quotes	
+						$was_enclosed = true;
-			} elseif ( $ch == $this->enclosure && $enclosed ) {
+					} else {
-				$current .= $ch;
+						$this->error = 2;
-				$i++;
+						$error_row = count($rows) + 1;
-
+						$error_col = $col + 1;
+						if ( !isset($this->error_info[$error_row.'-'.$error_col]) ) {
+							$this->error_info[$error_row.'-'.$error_col] = array(
+								'type' => 2,
+								'info' => 'Syntax error found on row '.$error_row.'. Non-enclosed fields can not contain double-quotes.',
+								'row' => $error_row,
+								'field' => $error_col,
+								'field_name' => (!empty($head[$col])) ? $head[$col] : null,
+							);
+						}
+						$current .= $ch;
+					}
+				} elseif ($nch == $this->enclosure) {
+					$current .= $ch;
+					$i++;
+				} elseif ( $nch != $this->delimiter && $nch != "\r" && $nch != "\n" ) {
+					for ( $x=($i+1); isset($data{$x}) && ltrim($data{$x}, $white_spaces) == ''; $x++ ) {}
+					if ( $data{$x} == $this->delimiter ) {
+						$enclosed = false;
+						$i = $x;
+					} else {
+						if ( $this->error < 1 ) {
+							$this->error = 1;
+						}
+						$error_row = count($rows) + 1;
+						$error_col = $col + 1;
+						if ( !isset($this->error_info[$error_row.'-'.$error_col]) ) {
+							$this->error_info[$error_row.'-'.$error_col] = array(
+								'type' => 1,
+								'info' =>
+									'Syntax error found on row '.(count($rows) + 1).'. '.
+									'A single double-quote was found within an enclosed string. '.
+									'Enclosed double-quotes must be escaped with a second double-quote.',
+								'row' => count($rows) + 1,
+								'field' => $col + 1,
+								'field_name' => (!empty($head[$col])) ? $head[$col] : null,
+							);
+						}
+						$current .= $ch;
+						$enclosed = false;
+					}
+				} else {
+					$enclosed = false;
+				}
+				
 			// end of field/row
-			} elseif ( ($ch == $this->delimiter || ($ch == "\n" && $pch != "\r") || $ch == "\r") && !$enclosed ) {
+			} elseif ( ($ch == $this->delimiter || $ch == "\n" || $ch == "\r") && !$enclosed ) {
-				if ( !$was_enclosed ) $current = trim($current);
 				$key = ( !empty($head[$col]) ) ? $head[$col] : $col ;
-				$row[$key] = $current;
+				$row[$key] = ( $was_enclosed ) ? $current : trim($current) ;
 				$current = '';
+				$was_enclosed = false;
 				$col++;
-			
+				
 				// end of row
 				if ( $ch == "\n" || $ch == "\r" ) {
 					if ( $this->_validate_offset($row_count) && $this->_validate_row_conditions($row, $this->conditions) ) {
@@ -405,6 +464,7 @@ class parseCSV {
 					if ( $this->sort_by === null && $this->limit !== null && count($rows) == $this->limit ) {
 						$i = $strlen;
 					}
+					if ( $ch == "\r" && $nch == "\n" ) $i++;
 				}
 				
 			// append character to current field
@@ -441,7 +501,7 @@ class parseCSV {
 		$entry = array();
 		
 		// create heading
-		if ( $this->heading && !$append ) {
+		if ( $this->heading && !$append && !empty($fields) ) {
 			foreach( $fields as $key => $value ) {
 				$entry[] = $this->_enclose_value($value);
 			}
@@ -503,11 +563,11 @@ class parseCSV {
 	function _validate_row_conditions ($row = array(), $conditions = null) {
 		if ( !empty($row) ) {
 			if ( !empty($conditions) ) {
-				$conditions = (strpos($conditions, 'OR') !== false) ? explode('OR', $conditions) : array($conditions) ;
+				$conditions = (strpos($conditions, ' OR ') !== false) ? explode(' OR ', $conditions) : array($conditions) ;
 				$or = '';
 				foreach( $conditions as $key => $value ) {
-					if ( strpos($value, 'AND') !== false ) {
+					if ( strpos($value, ' AND ') !== false ) {
-						$value = explode('AND', $value);
+						$value = explode(' AND ', $value);
 						$and = '';
 						foreach( $value as $k => $v ) {
 							$and .= $this->_validate_row_condition($row, $v);
@@ -601,11 +661,13 @@ class parseCSV {
 	 * @return  Processed value
 	 */
 	function _enclose_value ($value = null) {
-		$delimiter = preg_quote($this->delimiter, '/');
+		if ( $value !== null && $value != '' ) {
-		$enclosure = preg_quote($this->enclosure, '/');
+			$delimiter = preg_quote($this->delimiter, '/');
-		if ( preg_match("/".$delimiter."|".$enclosure."|\n|\r/i", $value) || ($value{0} == ' ' || substr($value, -1) == ' ') ) {
+			$enclosure = preg_quote($this->enclosure, '/');
-			$value = str_replace($this->enclosure, $this->enclosure.$this->enclosure, $value);
+			if ( preg_match("/".$delimiter."|".$enclosure."|\n|\r/i", $value) || ($value{0} == ' ' || substr($value, -1) == ' ') ) {
-			$value = $this->enclosure.$value.$this->enclosure;
+				$value = str_replace($this->enclosure, $this->enclosure.$this->enclosure, $value);
+				$value = $this->enclosure.$value.$this->enclosure;
+			}
 		}
 		return $value;
 	}