Fix/sheets xlsx chart (#1761)

* Add support for Google Sheets Exported XLSX Charts

Google Sheets XLSX charts use oneCellAnchor positioning and the data series
do not have the *Cache elements with cached values.

* update CHANGELOG

* Add support for Google Sheets Exported XLSX Charts

Google Sheets XLSX charts use oneCellAnchor positioning and the data series
do not have the *Cache elements with cached values. Because the reader had been
assuming *Cache elements existed as children of strRef and numRef, errors about
the node being deleted were thrown when reading Xlsx exported from Google Sheets.

Co-authored-by: Darren Maczka <dkm@utk.edu>

CHANGELOG.md

@@ -58,6 +58,7 @@ and this project adheres to [Semantic Versioning](https://semver.org).
 
 ### Fixed
 
+- Resolve Google Sheets Xlsx charts issue. Google Sheets uses oneCellAnchor positioning and does not include *Cache values in the exported Xlsx.
 - Fix for Xls Reader when SST has a bad length [#1592](https://github.com/PHPOffice/PhpSpreadsheet/issues/1592)
 - Resolve Xlsx loader issue whe hyperlinks don't have a destination
 - Resolve issues when printer settings resources IDs clash with drawing IDs

src/PhpSpreadsheet/Reader/Xlsx.php

@@ -1155,13 +1155,27 @@ class Xlsx extends BaseReader
                                                     $this->readHyperLinkDrawing($objDrawing, $oneCellAnchor, $hyperlinks);
 
                                                     $objDrawing->setWorksheet($docSheet);
-                                                } else {
-                                                    //    ? Can charts be positioned with a oneCellAnchor ?
+                                                } elseif ($this->includeCharts && $oneCellAnchor->graphicFrame) {
+                                                    // Exported XLSX from Google Sheets positions charts with a oneCellAnchor
                                                     $coordinates = Coordinate::stringFromColumnIndex(((string) $oneCellAnchor->from->col) + 1) . ($oneCellAnchor->from->row + 1);
                                                     $offsetX = Drawing::EMUToPixels($oneCellAnchor->from->colOff);
                                                     $offsetY = Drawing::EMUToPixels($oneCellAnchor->from->rowOff);
                                                     $width = Drawing::EMUToPixels(self::getArrayItem($oneCellAnchor->ext->attributes(), 'cx'));
                                                     $height = Drawing::EMUToPixels(self::getArrayItem($oneCellAnchor->ext->attributes(), 'cy'));
+
+                                                    $graphic = $oneCellAnchor->graphicFrame->children('http://schemas.openxmlformats.org/drawingml/2006/main')->graphic;
+                                                    /** @var SimpleXMLElement $chartRef */
+                                                    $chartRef = $graphic->graphicData->children('http://schemas.openxmlformats.org/drawingml/2006/chart')->chart;
+                                                    $thisChart = (string) $chartRef->attributes('http://schemas.openxmlformats.org/officeDocument/2006/relationships');
+
+                                                    $chartDetails[$docSheet->getTitle() . '!' . $thisChart] = [
+                                                        'fromCoordinate' => $coordinates,
+                                                        'fromOffsetX' => $offsetX,
+                                                        'fromOffsetY' => $offsetY,
+                                                        'width' => $width,
+                                                        'height' => $height,
+                                                        'worksheetTitle' => $docSheet->getTitle(),
+                                                    ];
                                                 }
                                             }
                                         }
@@ -1508,10 +1522,13 @@ class Xlsx extends BaseReader
                                     $excel->getSheetByName($charts[$chartEntryRef]['sheet'])->addChart($objChart);
                                     $objChart->setWorksheet($excel->getSheetByName($charts[$chartEntryRef]['sheet']));
                                     $objChart->setTopLeftPosition($chartDetails[$chartPositionRef]['fromCoordinate'], $chartDetails[$chartPositionRef]['fromOffsetX'], $chartDetails[$chartPositionRef]['fromOffsetY']);
+                                    if (array_key_exists('toCoordinate', $chartDetails[$chartPositionRef])) {
+                                        // For oneCellAnchor positioned charts, toCoordinate is not in the data. Does it need to be calculated?
                                         $objChart->setBottomRightPosition($chartDetails[$chartPositionRef]['toCoordinate'], $chartDetails[$chartPositionRef]['toOffsetX'], $chartDetails[$chartPositionRef]['toOffsetY']);
                                     }
                                 }
                             }
+                        }
 
                         break;
 

src/PhpSpreadsheet/Reader/Xlsx/Chart.php

@@ -328,26 +328,51 @@ class Chart
     {
         if (isset($seriesDetail->strRef)) {
             $seriesSource = (string) $seriesDetail->strRef->f;
-            $seriesData = self::chartDataSeriesValues($seriesDetail->strRef->strCache->children($namespacesChartMeta['c']), 's');
+            $seriesValues = new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, null, null, null, $marker);
 
-            return new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, $seriesData['formatCode'], $seriesData['pointCount'], $seriesData['dataValues'], $marker);
+            if (isset($seriesDetail->strRef->strCache)) {
+                $seriesData = self::chartDataSeriesValues($seriesDetail->strRef->strCache->children($namespacesChartMeta['c']), 's');
+                $seriesValues
+                    ->setFormatCode($seriesData['formatCode'])
+                    ->setDataValues($seriesData['dataValues']);
+            }
+
+            return $seriesValues;
         } elseif (isset($seriesDetail->numRef)) {
             $seriesSource = (string) $seriesDetail->numRef->f;
+            $seriesValues = new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_NUMBER, $seriesSource, null, null, null, $marker);
+            if (isset($seriesDetail->strRef->strCache)) {
                 $seriesData = self::chartDataSeriesValues($seriesDetail->numRef->numCache->children($namespacesChartMeta['c']));
+                $seriesValues
+                    ->setFormatCode($seriesData['formatCode'])
+                    ->setDataValues($seriesData['dataValues']);
+            }
 
-            return new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_NUMBER, $seriesSource, $seriesData['formatCode'], $seriesData['pointCount'], $seriesData['dataValues'], $marker);
+            return $seriesValues;
         } elseif (isset($seriesDetail->multiLvlStrRef)) {
             $seriesSource = (string) $seriesDetail->multiLvlStrRef->f;
-            $seriesData = self::chartDataSeriesValuesMultiLevel($seriesDetail->multiLvlStrRef->multiLvlStrCache->children($namespacesChartMeta['c']), 's');
-            $seriesData['pointCount'] = count($seriesData['dataValues']);
+            $seriesValues = new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, null, null, null, $marker);
 
-            return new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, $seriesData['formatCode'], $seriesData['pointCount'], $seriesData['dataValues'], $marker);
+            if (isset($seriesDetail->multiLvlStrRef->multiLvlStrCache)) {
+                $seriesData = self::chartDataSeriesValuesMultiLevel($seriesDetail->multiLvlStrRef->multiLvlStrCache->children($namespacesChartMeta['c']), 's');
+                $seriesValues
+                    ->setFormatCode($seriesData['formatCode'])
+                    ->setDataValues($seriesData['dataValues']);
+            }
+
+            return $seriesValues;
         } elseif (isset($seriesDetail->multiLvlNumRef)) {
             $seriesSource = (string) $seriesDetail->multiLvlNumRef->f;
-            $seriesData = self::chartDataSeriesValuesMultiLevel($seriesDetail->multiLvlNumRef->multiLvlNumCache->children($namespacesChartMeta['c']), 's');
-            $seriesData['pointCount'] = count($seriesData['dataValues']);
+            $seriesValues = new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, null, null, null, $marker);
 
-            return new DataSeriesValues(DataSeriesValues::DATASERIES_TYPE_STRING, $seriesSource, $seriesData['formatCode'], $seriesData['pointCount'], $seriesData['dataValues'], $marker);
+            if (isset($seriesDetail->multiLvlNumRef->multiLvlNumCache)) {
+                $seriesData = self::chartDataSeriesValuesMultiLevel($seriesDetail->multiLvlNumRef->multiLvlNumCache->children($namespacesChartMeta['c']), 's');
+                $seriesValues
+                    ->setFormatCode($seriesData['formatCode'])
+                    ->setDataValues($seriesData['dataValues']);
+            }
+
+            return $seriesValues;
         }
 
         return null;

tests/PhpSpreadsheetTests/Reader/SheetsXlsxChartTest.php

@@ -0,0 +1,51 @@
+<?php
+
+namespace PhpOffice\PhpSpreadsheetTests\Reader;
+
+use PhpOffice\PhpSpreadsheet\Chart\DataSeries;
+use PhpOffice\PhpSpreadsheet\IOFactory;
+use PHPUnit\Framework\TestCase;
+
+class SheetsXlsxChartTest extends TestCase
+{
+    public function testLoadSheetsXlsxChart(): void
+    {
+        $filename = 'tests/data/Reader/XLSX/sheetsChartsTest.xlsx';
+        $reader = IOFactory::createReader('Xlsx')->setIncludeCharts(true);
+        $spreadsheet = $reader->load($filename);
+        $worksheet = $spreadsheet->getActiveSheet();
+
+        $charts = $worksheet->getChartCollection();
+        self::assertEquals(2, $worksheet->getChartCount());
+        self::assertCount(2, $charts);
+
+        $chart1 = $charts[0];
+        $pa1 = $chart1->getPlotArea();
+        self::assertEquals(2, $pa1->getPlotSeriesCount());
+
+        $pg1 = $pa1->getPlotGroup()[0];
+
+        self::assertEquals(DataSeries::TYPE_LINECHART, $pg1->getPlotType());
+        self::assertCount(2, $pg1->getPlotLabels());
+        self::assertCount(2, $pg1->getPlotValues());
+        self::assertCount(2, $pg1->getPlotCategories());
+
+        $chart2 = $charts[1];
+        $pa1 = $chart2->getPlotArea();
+        self::assertEquals(2, $pa1->getPlotSeriesCount());
+
+        $pg1 = $pa1->getPlotGroupByIndex(0);
+        //Before a refresh, data values are empty
+        foreach ($pg1->getPlotValues() as $dv) {
+            self::assertEmpty($dv->getPointCount());
+        }
+        $pg1->refresh($worksheet);
+        foreach ($pg1->getPlotValues() as $dv) {
+            self::assertEquals(9, $dv->getPointCount());
+        }
+        self::assertEquals(DataSeries::TYPE_SCATTERCHART, $pg1->getPlotType());
+        self::assertCount(2, $pg1->getPlotLabels());
+        self::assertCount(2, $pg1->getPlotValues());
+        self::assertCount(2, $pg1->getPlotCategories());
+    }
+}