Merge pull request #48 from box/add_support_for_more_cell_types

Add support for more cell types

README.md

@@ -1,11 +1,13 @@
 # Spout
 
+[![Latest Stable Version](https://poser.pugx.org/box/spout/v/stable)](https://packagist.org/packages/box/spout)
+[![Project Status](http://opensource.box.com/badges/active.svg)](http://opensource.box.com/badges)
+[![Build Status](https://travis-ci.org/box/spout.png?branch=master)](http://travis-ci.org/box/spout)
+[![Total Downloads](https://poser.pugx.org/box/spout/downloads)](https://packagist.org/packages/box/spout) [![License](https://poser.pugx.org/box/spout/license)](https://packagist.org/packages/box/spout)
+
 Spout is a PHP library to read and write CSV and XLSX files, in a fast and scalable way.
 Contrary to other file readers or writers, it is capable of processing very large files while keeping the memory usage really low (less than 10MB).
 
-[![Build Status](https://travis-ci.org/box/spout.png?branch=master)](http://travis-ci.org/box/spout)
-[![Project Status](http://opensource.box.com/badges/active.svg)](http://opensource.box.com/badges)
-
 
 ## Installation
 
@@ -248,6 +250,11 @@ Here are a few numbers regarding the performance of Spout:
 | Read XLSX (using shared strings) | 1 second                 | 1-2 minutes                  | 35-40 minutes                    |
 | Write XLSX                       | 1 second                 | 20-25 seconds                | 8-10 minutes                     |
 
+#### Does Spout support charts or formulas?
+
+No. This is a compromise to keep memory usage low. Charts and formulas requires data to be kept in memory in order to be used.
+So the larger the file would be, the more memory would be consumed, preventing your code to scale well.
+
 
 ## Support
 

composer.json

@@ -2,7 +2,6 @@
     "name": "box/spout",
     "description": "PHP Library to read and write CSV and XLSX files, in a fast and scalable way",
     "type": "library",
-    "version": "1.0.7",
     "keywords": ["php","read","write","csv","xlsx","excel","spreadsheet","scale","memory","stream","ooxml"],
     "license": "Apache-2.0",
     "homepage": "https://www.github.com/box/spout",

src/Spout/Reader/Helper/XLSX/SharedStringsHelper.php

@@ -71,6 +71,24 @@ class SharedStringsHelper
         $this->tempFolder = $this->fileSystemHelper->createFolder($rootTempFolder, uniqid('sharedstrings'));
     }
 
+    /**
+     * Returns whether the XLSX file contains a shared strings XML file
+     *
+     * @return bool
+     */
+    public function hasSharedStrings()
+    {
+        $hasSharedStrings = false;
+        $zip = new \ZipArchive();
+
+        if ($zip->open($this->filePath) === true) {
+            $hasSharedStrings = ($zip->locateName(self::SHARED_STRINGS_XML_FILE_PATH) !== false);
+            $zip->close();
+        }
+
+        return $hasSharedStrings;
+    }
+
     /**
      * Builds an in-memory array containing all the shared strings of the worksheet.
      * All the strings are stored in a XML file, located at 'xl/sharedStrings.xml'.
@@ -93,7 +111,7 @@ class SharedStringsHelper
         $this->tempFilePointer = null;
         $escaper = new \Box\Spout\Common\Escaper\XLSX();
 
-        $sharedStringsFilePath = 'zip://' . $this->filePath . '#' . self::SHARED_STRINGS_XML_FILE_PATH;
+        $sharedStringsFilePath = $this->getSharedStringsFilePath();
         if ($xmlReader->open($sharedStringsFilePath, null, LIBXML_NONET) === false) {
             throw new IOException('Could not open "' . self::SHARED_STRINGS_XML_FILE_PATH . '".');
         }
@@ -143,6 +161,15 @@ class SharedStringsHelper
         $xmlReader->close();
     }
 
+    /**
+     * @return string The path to the shared strings XML file, working cross-platforms
+     */
+    protected function getSharedStringsFilePath()
+    {
+        $sharedStringsXmlFilePath = str_replace('/', DIRECTORY_SEPARATOR, self::SHARED_STRINGS_XML_FILE_PATH);
+        return 'zip://' . $this->filePath . '#' . $sharedStringsXmlFilePath;
+    }
+
     /**
      * Removes nodes that should not be read, like the pronunciation of the Kanji characters.
      * By keeping them, their text content would be added to the read string.

src/Spout/Reader/Helper/XLSX/WorksheetHelper.php

@@ -186,7 +186,8 @@ class WorksheetHelper
      */
     protected function getFileAsXMLElementWithNamespace($xmlFilePath, $mainNamespace)
     {
-        $xmlContents = $this->globalFunctionsHelper->file_get_contents('zip://' . $this->filePath . '#' . $xmlFilePath);
+        $normalizedXmlFilePath = str_replace('/', DIRECTORY_SEPARATOR, $xmlFilePath);
+        $xmlContents = $this->globalFunctionsHelper->file_get_contents('zip://' . $this->filePath . '#' . $normalizedXmlFilePath);
 
         $xmlElement = new \SimpleXMLElement($xmlContents);
         $xmlElement->registerXPathNamespace('ns', $mainNamespace);

src/Spout/Reader/Internal/XLSX/Worksheet.php

@@ -32,11 +32,13 @@ class Worksheet
     }
 
     /**
-     * @return string Path of the XML file containing the worksheet data, without the leading slash
+     * @return string Path of the XML file containing the worksheet data,
+     *                without the leading slash and working cross-platforms.
      */
     public function getDataXmlFilePath()
     {
-        return ltrim($this->dataXmlFilePath, '/');
+        $dataXmlFilePath = ltrim($this->dataXmlFilePath, '/');
+        return str_replace('/', DIRECTORY_SEPARATOR, $dataXmlFilePath);
     }
 
     /**

src/Spout/Reader/XLSX.php

@@ -72,8 +72,12 @@ class XLSX extends AbstractReader
         $this->zip = new \ZipArchive();
 
         if ($this->zip->open($filePath) === true) {
+            $this->sharedStringsHelper = new SharedStringsHelper($filePath, $this->tempFolder);
+
+            if ($this->sharedStringsHelper->hasSharedStrings()) {
                 // Extracts all the strings from the worksheets for easy access in the future
-            $this->extractSharedStrings($filePath);
+                $this->sharedStringsHelper->extractSharedStrings();
+            }
 
             // Fetch all available worksheets
             $this->worksheetHelper = new WorksheetHelper($filePath, $this->globalFunctionsHelper);
@@ -87,19 +91,6 @@ class XLSX extends AbstractReader
         }
     }
 
-    /**
-     * Builds an in-memory array containing all the shared strings of the worksheets.
-     *
-     * @param  string $filePath Path of the XLSX file to be read
-     * @return void
-     * @throws \Box\Spout\Common\Exception\IOException If sharedStrings XML file can't be read
-     */
-    protected function extractSharedStrings($filePath)
-    {
-        $this->sharedStringsHelper = new SharedStringsHelper($filePath, $this->tempFolder);
-        $this->sharedStringsHelper->extractSharedStrings();
-    }
-
     /**
      * Returns whether another worksheet exists after the current worksheet.
      *
@@ -186,6 +177,7 @@ class XLSX extends AbstractReader
             throw new BadUsageException('You must call nextSheet() before calling hasNextRow() or nextRow()');
         }
 
+        $escaper = new \Box\Spout\Common\Escaper\XLSX();
         $isInsideRowTag = false;
         $rowData = [];
 
@@ -197,6 +189,7 @@ class XLSX extends AbstractReader
                     $lastCellIndex = $matches[1];
                     $this->numberOfColumns = CellHelper::getColumnIndexFromCellIndex($lastCellIndex) + 1;
                 }
+
             } else if ($this->xmlReader->nodeType == \XMLReader::ELEMENT && $this->xmlReader->name === 'row') {
                 // Start of the row description
                 $isInsideRowTag = true;
@@ -209,20 +202,15 @@ class XLSX extends AbstractReader
                     $numberOfColumnsForRow = intval($numberOfColumnsForRow);
                 }
                 $rowData = ($numberOfColumnsForRow !== 0) ? array_fill(0, $numberOfColumnsForRow, '') : [];
+
             } else if ($isInsideRowTag && $this->xmlReader->nodeType == \XMLReader::ELEMENT && $this->xmlReader->name === 'c') {
                 // Start of a cell description
                 $currentCellIndex = $this->xmlReader->getAttribute('r');
                 $currentColumnIndex = CellHelper::getColumnIndexFromCellIndex($currentCellIndex);
-                $node = $this->xmlReader->expand();
 
-                $hasSharedString = ($this->xmlReader->getAttribute('t') === 's');
-                if ($hasSharedString) {
-                    $sharedStringIndex = intval($node->nodeValue);
-                    $rowData[$currentColumnIndex] = $this->sharedStringsHelper->getStringAtIndex($sharedStringIndex);
-                } else {
-                    // for inline strings or numbers, just get the value
-                    $rowData[$currentColumnIndex] = trim($node->nodeValue);
-                }
+                $node = $this->xmlReader->expand();
+                $rowData[$currentColumnIndex] = $this->getCellValue($node, $escaper);
+
             } else if ($this->xmlReader->nodeType == \XMLReader::END_ELEMENT && $this->xmlReader->name === 'row') {
                 // End of the row description
                 // If needed, we fill the empty cells
@@ -235,6 +223,58 @@ class XLSX extends AbstractReader
         return ($rowData !== []) ? $rowData : null;
     }
 
+    /**
+     * Returns the (unescaped) cell value associated to the given XML node.
+     *
+     * @param \DOMNode $node
+     * @param \Box\Spout\Common\Escaper\XLSX $escaper
+     * @return string|int|float|bool|null The value associated with the cell (null when the cell has an error)
+     */
+    protected function getCellValue($node, $escaper)
+    {
+        $cellValue = '';
+
+        // Default cell type is "n"
+        $cellType = $node->getAttribute('t') ?: 'n';
+
+        if ($cellType === 'inlineStr') {
+            // inline strings are formatted this way:
+            // <c r="A1" t="inlineStr"><is><t>[INLINE_STRING]</t></is></c>
+            $tNode = $node->getElementsByTagName('t')->item(0);
+            $escapedCellValue = trim($tNode->nodeValue);
+            $cellValue = $escaper->unescape($escapedCellValue);
+        } else {
+            // all other cell types should have a "v" tag containing the value.
+            // if not, the returned value should be empty string.
+            $vNode = $node->getElementsByTagName('v')->item(0);
+
+            if ($vNode !== null) {
+                if ($cellType === 's') {
+                    // shared strings are formatted this way:
+                    // <c r="A1" t="s"><v>[SHARED_STRING_INDEX]</v></c>
+                    $sharedStringIndex = intval($vNode->nodeValue);
+                    $escapedCellValue = $this->sharedStringsHelper->getStringAtIndex($sharedStringIndex);
+                    $cellValue = $escaper->unescape($escapedCellValue);
+                } else if ($cellType === 'b') {
+                    // !! is similar to boolval()
+                    $cellValue = !!$vNode->nodeValue;
+                } else if ($cellType === 'n') {
+                    $nodeValue = $vNode->nodeValue;
+                    $cellValue = is_int($nodeValue) ? intval($nodeValue) : floatval($nodeValue);
+                } else if ($cellType === 'd') {
+                    $cellValue = new \DateTime($vNode->nodeValue);
+                } else if ($cellType === 'e') {
+                    $cellValue = null;
+                } else if ($cellType === 'str') {
+                    $escapedCellValue = trim($vNode->nodeValue);
+                    $cellValue = $escaper->unescape($escapedCellValue);
+                }
+            }
+        }
+
+        return $cellValue;
+    }
+
     /**
      * Closes the reader. To be used after reading the file.
      *

tests/Spout/Reader/XLSXTest.php

@@ -100,6 +100,41 @@ class XLSXTest extends \PHPUnit_Framework_TestCase
         }
     }
 
+    /**
+     * @return void
+     */
+    public function testReadShouldSupportFilesWithoutSharedStringsFile()
+    {
+        $allRows = $this->getAllRowsForFile('sheet_with_no_shared_strings_file.xlsx');
+
+        $expectedRows = [
+            [10, 11],
+            [20, 21],
+        ];
+        $this->assertEquals($expectedRows, $allRows);
+    }
+
+    /**
+     * @return void
+     */
+    public function testReadShouldSupportAllCellTypes()
+    {
+        $allRows = $this->getAllRowsForFile('sheet_with_all_cell_types.xlsx');
+
+        $expectedRows = [
+            [
+                's1--A1', 's1--A2',
+                false, true,
+                \DateTime::createFromFormat('Y-m-d H:i:s', '2015-06-03 13:21:58'),
+                \DateTime::createFromFormat('Y-m-d H:i:s', '2015-06-01 00:00:00'),
+                10, 10.43,
+                null,
+            ],
+            ['', '', '', '', '', '', '', '', ''],
+        ];
+        $this->assertEquals($expectedRows, $allRows);
+    }
+
     /**
      * @return void
      */
@@ -231,6 +266,21 @@ class XLSXTest extends \PHPUnit_Framework_TestCase
         $this->assertEquals([], $allRows, 'Sheet with no cells should be correctly processed.');
     }
 
+    /**
+     * @return void
+     */
+    public function testReadShouldSkipFormulas()
+    {
+        $allRows = $this->getAllRowsForFile('sheet_with_formulas.xlsx');
+
+        $expectedRows = [
+            ['val1', 'val2', 'total1', 'total2'],
+            [10, 20, 30, 21],
+            [11, 21, 32, 41],
+        ];
+        $this->assertEquals($expectedRows, $allRows);
+    }
+
     /**
      * @param string $fileName
      * @return array All the read rows the given file

tests/Spout/Writer/SheetTest.php

@@ -90,9 +90,18 @@ class SheetTest extends \PHPUnit_Framework_TestCase
      */
     private function assertSheetNameEquals($expectedName, $resourcePath, $message = '')
     {
-        $pathToWorkbookFile = $resourcePath . '#xl/workbook.xml';
+        $pathToWorkbookFile = $resourcePath . $this->normalizePath('#xl/workbook.xml');
         $xmlContents = file_get_contents('zip://' . $pathToWorkbookFile);
 
         $this->assertContains('<sheet name="' . $expectedName . '"', $xmlContents, $message);
     }
+
+    /**
+     * @param string $path
+     * @return string The path with the correct directory separators, as defined for the current platform
+     */
+    private function normalizePath($path)
+    {
+        return str_replace('/', DIRECTORY_SEPARATOR, $path);
+    }
 }

tests/Spout/Writer/XLSXTest.php

@@ -408,7 +408,7 @@ class XLSXTest extends \PHPUnit_Framework_TestCase
     private function assertInlineDataWasWrittenToSheet($fileName, $sheetIndex, $inlineData, $message = '')
     {
         $resourcePath = $this->getGeneratedResourcePath($fileName);
-        $pathToSheetFile = $resourcePath . '#xl/worksheets/sheet' . $sheetIndex . '.xml';
+        $pathToSheetFile = $resourcePath . $this->normalizePath('#xl/worksheets/sheet' . $sheetIndex . '.xml');
         $xmlContents = file_get_contents('zip://' . $pathToSheetFile);
 
         $this->assertContains((string)$inlineData, $xmlContents, $message);
@@ -424,7 +424,7 @@ class XLSXTest extends \PHPUnit_Framework_TestCase
     private function assertInlineDataWasNotWrittenToSheet($fileName, $sheetIndex, $inlineData, $message = '')
     {
         $resourcePath = $this->getGeneratedResourcePath($fileName);
-        $pathToSheetFile = $resourcePath . '#xl/worksheets/sheet' . $sheetIndex . '.xml';
+        $pathToSheetFile = $resourcePath . $this->normalizePath('#xl/worksheets/sheet' . $sheetIndex . '.xml');
         $xmlContents = file_get_contents('zip://' . $pathToSheetFile);
 
         $this->assertNotContains((string)$inlineData, $xmlContents, $message);
@@ -439,9 +439,18 @@ class XLSXTest extends \PHPUnit_Framework_TestCase
     private function assertSharedStringWasWritten($fileName, $sharedString, $message = '')
     {
         $resourcePath = $this->getGeneratedResourcePath($fileName);
-        $pathToSharedStringsFile = $resourcePath . '#xl/sharedStrings.xml';
+        $pathToSharedStringsFile = $resourcePath . $this->normalizePath('#xl/sharedStrings.xml');
         $xmlContents = file_get_contents('zip://' . $pathToSharedStringsFile);
 
         $this->assertContains($sharedString, $xmlContents, $message);
     }
+
+    /**
+     * @param string $path
+     * @return string The path with the correct directory separators, as defined for the current platform
+     */
+    private function normalizePath($path)
+    {
+        return str_replace('/', DIRECTORY_SEPARATOR, $path);
+    }
 }