More Extendability for CSV Reader

CHANGELOG.md

@@ -9,9 +9,9 @@ and this project adheres to [Semantic Versioning](https://semver.org). Thia is a
 
 ### Added
 
+- Make Reader/Csv Extendable, add new preferred Reader/CsvNoEscape class. [Issue #4836](https://github.com/PHPOffice/PhpSpreadsheet/issues/4836) [PR #4837](https://github.com/PHPOffice/PhpSpreadsheet/pull/4837) [PR #4845](https://github.com/PHPOffice/PhpSpreadsheet/pull/4845)
 - XLOOKUP function. [Issue #1453](https://github.com/PHPOffice/PhpSpreadsheet/issues/1453) [PR #4844](https://github.com/PHPOffice/PhpSpreadsheet/pull/4844)
 - Introduction of a benchmark test suite, independent of the default unit test suite. Users can use it as a template for experimenting and making decisions concerning performance. [PR #4824](https://github.com/PHPOffice/PhpSpreadsheet/pull/4824)
-- Make Reader/Csv Extendable, add new preferred Reader/CsvNoEscape class. [Issue #4836](https://github.com/PHPOffice/PhpSpreadsheet/issues/4836) [PR #4837](https://github.com/PHPOffice/PhpSpreadsheet/pull/4837)
 
 ### Removed
 

src/PhpSpreadsheet/Cell/Cell.php

@@ -426,7 +426,7 @@ public function getCalculatedValueString(): string
             $value = array_shift($value);
         }
 
-        return StringHelper::convertToString($value, false);
+        return StringHelper::convertToString($value, false, convertBool: true);
     }
 
     /**

src/PhpSpreadsheet/Reader/BaseReader.php

@@ -94,7 +94,7 @@ public function getReadDataOnly(): bool
         return $this->readDataOnly;
     }
 
-    public function setReadDataOnly(bool $readCellValuesOnly): self
+    public function setReadDataOnly(bool $readCellValuesOnly): static
     {
         $this->readDataOnly = $readCellValuesOnly;
 
@@ -106,7 +106,7 @@ public function getReadEmptyCells(): bool
         return $this->readEmptyCells;
     }
 
-    public function setReadEmptyCells(bool $readEmptyCells): self
+    public function setReadEmptyCells(bool $readEmptyCells): static
     {
         $this->readEmptyCells = $readEmptyCells;
 
@@ -118,7 +118,7 @@ public function getIgnoreRowsWithNoCells(): bool
         return $this->ignoreRowsWithNoCells;
     }
 
-    public function setIgnoreRowsWithNoCells(bool $ignoreRowsWithNoCells): self
+    public function setIgnoreRowsWithNoCells(bool $ignoreRowsWithNoCells): static
     {
         $this->ignoreRowsWithNoCells = $ignoreRowsWithNoCells;
 
@@ -130,7 +130,7 @@ public function getIncludeCharts(): bool
         return $this->includeCharts;
     }
 
-    public function setIncludeCharts(bool $includeCharts): self
+    public function setIncludeCharts(bool $includeCharts): static
     {
         $this->includeCharts = $includeCharts;
 
@@ -142,7 +142,7 @@ public function getEnableDrawingPassThrough(): bool
         return $this->enableDrawingPassThrough;
     }
 
-    public function setEnableDrawingPassThrough(bool $enableDrawingPassThrough): self
+    public function setEnableDrawingPassThrough(bool $enableDrawingPassThrough): static
     {
         $this->enableDrawingPassThrough = $enableDrawingPassThrough;
 
@@ -156,7 +156,7 @@ public function getLoadSheetsOnly(): ?array
     }
 
     /** @param null|string|string[] $sheetList */
-    public function setLoadSheetsOnly(string|array|null $sheetList): self
+    public function setLoadSheetsOnly(string|array|null $sheetList): static
     {
         if ($sheetList === null) {
             return $this->setLoadAllSheets();
@@ -167,7 +167,7 @@ public function setLoadSheetsOnly(string|array|null $sheetList): self
         return $this;
     }
 
-    public function setLoadAllSheets(): self
+    public function setLoadAllSheets(): static
     {
         $this->loadSheetsOnly = null;
 
@@ -179,7 +179,7 @@ public function getReadFilter(): IReadFilter
         return $this->readFilter;
     }
 
-    public function setReadFilter(IReadFilter $readFilter): self
+    public function setReadFilter(IReadFilter $readFilter): static
     {
         $this->readFilter = $readFilter;
 
@@ -192,7 +192,7 @@ public function setReadFilter(IReadFilter $readFilter): self
      * these can be specified within a spreadsheet
      * in a way that can subject the caller to security exploits.
      */
-    public function setAllowExternalImages(bool $allowExternalImages): self
+    public function setAllowExternalImages(bool $allowExternalImages): static
     {
         $this->allowExternalImages = $allowExternalImages;
 
@@ -213,7 +213,7 @@ public function getAllowExternalImages(): bool
      *
      * @param Closure(string):bool $isWhitelisted
      */
-    public function setIsWhitelisted(Closure $isWhitelisted): self
+    public function setIsWhitelisted(Closure $isWhitelisted): static
     {
         $this->isWhitelisted = $isWhitelisted;
 
@@ -224,7 +224,7 @@ public function setIsWhitelisted(Closure $isWhitelisted): self
      * Create a blank sheet if none are read,
      * possibly due to a typo when using LoadSheetsOnly.
      */
-    public function setCreateBlankSheetIfNoneRead(bool $createBlankSheetIfNoneRead): self
+    public function setCreateBlankSheetIfNoneRead(bool $createBlankSheetIfNoneRead): static
     {
         $this->createBlankSheetIfNoneRead = $createBlankSheetIfNoneRead;
 
@@ -346,7 +346,7 @@ public function getValueBinder(): ?IValueBinder
         return $this->valueBinder;
     }
 
-    public function setValueBinder(?IValueBinder $valueBinder): self
+    public function setValueBinder(?IValueBinder $valueBinder): static
     {
         $this->valueBinder = $valueBinder;
 

src/PhpSpreadsheet/Reader/Csv.php

@@ -89,17 +89,17 @@ class Csv extends BaseReader
 
     protected bool $preserveNumericFormatting = false;
 
-    private bool $preserveNullString = false;
+    protected bool $preserveNullString = false;
 
-    private bool $sheetNameIsFileName = false;
+    protected bool $sheetNameIsFileName = false;
 
-    private string $getTrue = 'true';
+    protected ?string $getTrue = null;
 
-    private string $getFalse = 'false';
+    protected ?string $getFalse = null;
 
-    private string $thousandsSeparator = ',';
+    protected ?string $thousandsSeparator = null;
 
-    private string $decimalSeparator = '.';
+    protected ?string $decimalSeparator = null;
 
     /**
      * Create a new CSV Reader instance.
@@ -301,18 +301,27 @@ private function openFileOrMemory(string $filename): void
         }
         $this->openFile($filename);
         if ($this->inputEncoding !== 'UTF-8') {
-            fclose($this->fileHandle);
-            $entireFile = file_get_contents($filename);
-            $fileHandle = fopen('php://memory', 'r+b');
-            if ($fileHandle !== false && $entireFile !== false) {
-                $this->fileHandle = $fileHandle;
-                $data = StringHelper::convertEncoding($entireFile, 'UTF-8', $this->inputEncoding);
-                fwrite($this->fileHandle, $data);
-                $this->skipBOM();
-            }
+            $this->convertNonUtf8($filename);
         }
     }
 
+    protected function convertNonUtf8(string $filename): void
+    {
+        fclose($this->fileHandle);
+        $entireFile = file_get_contents($filename);
+        if ($entireFile === false) {
+            throw new ReaderException("Unable to get contents of $filename"); // @codeCoverageIgnore
+        }
+        $fileHandle = fopen('php://memory', 'r+b');
+        if ($fileHandle === false) {
+            throw new ReaderException('Unable to open php://memory'); // @codeCoverageIgnore
+        }
+        $this->fileHandle = $fileHandle;
+        $data = StringHelper::convertEncoding($entireFile, 'UTF-8', $this->inputEncoding);
+        fwrite($this->fileHandle, $data);
+        $this->skipBOM();
+    }
+
     public function setTestAutoDetect(bool $value): self
     {
         $this->testAutodetect = $value;
@@ -418,10 +427,10 @@ private function loadStringOrFile2(string $filename, Spreadsheet $spreadsheet, b
         $rowData = self::getCsv($fileHandle, 0, $delimiter, $this->enclosure, $this->escapeCharacter);
         $valueBinder = $this->valueBinder ?? Cell::getValueBinder();
         $preserveBooleanString = method_exists($valueBinder, 'getBooleanConversion') && $valueBinder->getBooleanConversion();
-        $this->getTrue = Calculation::getTRUE();
-        $this->getFalse = Calculation::getFALSE();
-        $this->thousandsSeparator = StringHelper::getThousandsSeparator();
-        $this->decimalSeparator = StringHelper::getDecimalSeparator();
+        $this->getTrue ??= Calculation::getTRUE();
+        $this->getFalse ??= Calculation::getFALSE();
+        $this->thousandsSeparator ??= StringHelper::getThousandsSeparator();
+        $this->decimalSeparator ??= StringHelper::getDecimalSeparator();
         while (is_array($rowData)) {
             $noOutputYet = true;
             $columnLetter = 'A';
@@ -466,9 +475,9 @@ private function loadStringOrFile2(string $filename, Spreadsheet $spreadsheet, b
     private function convertBoolean(mixed &$rowDatum): void
     {
         if (is_string($rowDatum)) {
-            if (strcasecmp($this->getTrue, $rowDatum) === 0 || strcasecmp('true', $rowDatum) === 0) {
+            if (strcasecmp((string) $this->getTrue, $rowDatum) === 0 || strcasecmp('true', $rowDatum) === 0) {
                 $rowDatum = true;
-            } elseif (strcasecmp($this->getFalse, $rowDatum) === 0 || strcasecmp('false', $rowDatum) === 0) {
+            } elseif (strcasecmp((string) $this->getFalse, $rowDatum) === 0 || strcasecmp('false', $rowDatum) === 0) {
                 $rowDatum = false;
             }
         } else {
@@ -484,15 +493,15 @@ private function convertFormattedNumber(mixed &$rowDatum): string
         $numberFormatMask = '';
         if ($this->castFormattedNumberToNumeric === true && is_string($rowDatum)) {
             $numeric = str_replace(
-                [$this->thousandsSeparator, $this->decimalSeparator],
+                [(string) $this->thousandsSeparator, (string) $this->decimalSeparator],
                 ['', '.'],
                 $rowDatum
             );
 
             if (is_numeric($numeric)) {
-                $decimalPos = strpos($rowDatum, $this->decimalSeparator);
+                $decimalPos = strpos($rowDatum, (string) $this->decimalSeparator);
                 if ($this->preserveNumericFormatting === true) {
-                    $numberFormatMask = (str_contains($rowDatum, $this->thousandsSeparator))
+                    $numberFormatMask = (str_contains($rowDatum, (string) $this->thousandsSeparator))
                         ? '#,##0' : '0';
                     if ($decimalPos !== false) {
                         $decimals = strlen($rowDatum) - $decimalPos - 1;
@@ -760,4 +769,32 @@ private static function getDefaultEscapeCharacter(int $version = PHP_VERSION_ID)
     {
         return $version < 90000 ? '\\' : '';
     }
+
+    public function setGetTrue(?string $getTrue): self
+    {
+        $this->getTrue = $getTrue;
+
+        return $this;
+    }
+
+    public function setGetFalse(?string $getFalse): self
+    {
+        $this->getFalse = $getFalse;
+
+        return $this;
+    }
+
+    public function setDecimalSeparator(?string $decimalSeparator): self
+    {
+        $this->decimalSeparator = $decimalSeparator;
+
+        return $this;
+    }
+
+    public function setThousandsSeparator(?string $thousandsSeparator): self
+    {
+        $this->thousandsSeparator = $thousandsSeparator;
+
+        return $this;
+    }
 }

src/PhpSpreadsheet/Writer/Xls/Worksheet.php

@@ -432,7 +432,7 @@ public function close(): void
                             match ($calctype) {
                                 'integer', 'double' => $this->writeNumber($row, $column, is_numeric($calculatedValue) ? ((float) $calculatedValue) : 0.0, $xfIndex),
                                 'string' => $this->writeString($row, $column, $calculatedValueString, $xfIndex),
-                                'boolean' => $this->writeBoolErr($row, $column, (int) $calculatedValueString, 0, $xfIndex),
+                                'boolean' => $this->writeBoolErr($row, $column, (int) $calculatedValue, 0, $xfIndex), // @phpstan-ignore-line
                                 default => $this->writeString($row, $column, $cell->getValueString(), $xfIndex),
                             };
                         }

tests/PhpSpreadsheetTests/Reader/Csv/CsvIconv2.php

@@ -0,0 +1,32 @@
+<?php
+
+declare(strict_types=1);
+
+namespace PhpOffice\PhpSpreadsheetTests\Reader\Csv;
+
+use PhpOffice\PhpSpreadsheet\Reader\Csv;
+use PhpOffice\PhpSpreadsheet\Reader\Exception as ReaderException;
+
+class CsvIconv2 extends Csv
+{
+    protected function convertNonUtf8(string $filename): void
+    {
+        fclose($this->fileHandle);
+        $filtered = "php://filter/convert.iconv.{$this->inputEncoding}.utf-8/resource=$filename";
+        $sourceStream = fopen($filtered, 'rb');
+        if ($sourceStream === false) {
+            throw new ReaderException("Unable to get contents of $filename"); // @codeCoverageIgnore
+        }
+        $fileHandle = fopen('php://memory', 'r+b');
+        if ($fileHandle === false) {
+            throw new ReaderException('Unable to open php://memory'); // @codeCoverageIgnore
+        }
+        $bytes = stream_copy_to_stream($sourceStream, $fileHandle);
+        fclose($sourceStream);
+        if ($bytes === false) {
+            throw new ReaderException('Unable to copy stream'); // @codeCoverageIgnore
+        }
+        $this->fileHandle = $fileHandle;
+        $this->skipBOM();
+    }
+}