parseCSV 0.4 beta

- 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.

git-svn-id: http://parsecsv-for-php.googlecode.com/svn/trunk@28 339761fc-0c37-0410-822d-8b8cac1f6a97

ChangeLog.txt

@@ -1,3 +1,32 @@
+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

examples/_books.csv

@@ -1,15 +1,15 @@
-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 Third Secret,Steve Berry,Book,0340899263,,need to find time to read this book
-3,The Last Templar,Raymond Khoury,Book,0752880705,,
-5,The Traveller,John Twelve Hawks,Book,059305430X,,
-4,Crisis Four,Andy Mcnab,Book,0345428080,,
-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,Without Blood (Paperback),Alessandro Baricco,Book,1841955744,,
-5,State of Fear (Paperback),Michael Crichton,Book,0061015733,,
-4,The Rule of Four (Paperback),Ian Caldwell,Book,0099451956,book 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,Angels & Demons (Mass Market Paperback),Dan Brown,Book,0671027360,book danbrown bestseller,
+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 Third Secret,Steve Berry,Book,0340899263,,need to find time to read this book
+3,The Last Templar,Raymond Khoury,Book,0752880705,,
+5,The Traveller,John Twelve Hawks,Book,059305430X,,
+4,Crisis Four,Andy Mcnab,Book,0345428080,,
+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,Without Blood (Paperback),Alessandro Baricco,Book,1841955744,,
+5,State of Fear (Paperback),Michael Crichton,Book,0061015733,,
+4,The Rule of Four (Paperback),Ian Caldwell,Book,0099451956,book 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,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,

parsecsv.lib.php

@@ -4,7 +4,7 @@ class parseCSV {
 	
 /*
 
-	Class: parseCSV v0.3.2
+	Class: parseCSV v0.4 beta
 	http://code.google.com/p/parsecsv-for-php/
 	
 	
@@ -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) ) {
-				$enclosed = ( $enclosed ) ? false : true ;
-			
-			// inline quotes	
-			} elseif ( $ch == $enclosure && $enclosed ) {
-				$i++;
-
+			if ( $ch == $enclosure ) {
+				if ( !$enclosed || $nch != $enclosure ) {
+					$enclosed = ( $enclosed ) ? false : true ;
+				} elseif ( $enclosed ) {
+					$i++;
+				}
+				
 			// end of row
 			} elseif ( ($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 = $delimiter;
+		$this->delimiter = reset($filtered);
 		
 		// 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
-			if ( $ch == $this->enclosure && (!$enclosed || $nch != $this->enclosure) ) {
-				$enclosed = ( $enclosed ) ? false : true ;
-				if ( $enclosed ) $was_enclosed = true;
-			
-			// inline quotes	
-			} elseif ( $ch == $this->enclosure && $enclosed ) {
-				$current .= $ch;
-				$i++;
-
+			// open/close quotes, and inline quotes
+			if ( $ch == $this->enclosure ) {
+				if ( !$enclosed ) {
+					if ( ltrim($current, $white_spaces) == '' ) {
+						$enclosed = true;
+						$was_enclosed = true;
+					} else {
+						$this->error = 2;
+						$error_row = count($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 ) {
-				if ( !$was_enclosed ) $current = trim($current);
+			} elseif ( ($ch == $this->delimiter || $ch == "\n" || $ch == "\r") && !$enclosed ) {
 				$key = ( !empty($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